diff --git "a/exp/log/log-train-2023-03-20-17-19-56-0" "b/exp/log/log-train-2023-03-20-17-19-56-0" new file mode 100644--- /dev/null +++ "b/exp/log/log-train-2023-03-20-17-19-56-0" @@ -0,0 +1,25523 @@ +2023-03-20 17:19:56,410 INFO [train.py:971] (0/2) Training started +2023-03-20 17:19:56,413 INFO [train.py:981] (0/2) Device: cuda:0 +2023-03-20 17:19:56,653 INFO [lexicon.py:168] (0/2) Loading pre-compiled data/lang_char/Linv.pt +2023-03-20 17:19:56,681 INFO [train.py:993] (0/2) {'frame_shift_ms': 10.0, 'allowed_excess_duration_ratio': 0.1, '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': '1.22', 'k2-build-type': 'Release', 'k2-with-cuda': True, 'k2-git-sha1': '96c9a2aece2a3a7633da07740e24fa3d96f5498c', 'k2-git-date': 'Thu Nov 10 08:14:02 2022', 'lhotse-version': '1.13.0.dev+git.527d964.clean', 'torch-version': '1.12.1', 'torch-cuda-available': True, 'torch-cuda-version': '11.6', 'python-version': '3.8', 'icefall-git-branch': 'aishell_zipformer', 'icefall-git-sha1': '8337628-dirty', 'icefall-git-date': 'Mon Mar 20 16:20:05 2023', 'icefall-path': '/ceph-data4/yangxiaoyu/softwares/icefall_development/icefall_aishell_zipformer', 'k2-path': '/ceph-data4/yangxiaoyu/softwares/anaconda3/envs/k2_latest/lib/python3.8/site-packages/k2/__init__.py', 'lhotse-path': '/ceph-data4/yangxiaoyu/softwares/lhotse_development/lhotse_random_padding_left/lhotse/__init__.py', 'hostname': 'de-74279-k2-train-7-1218101249-5d97868c7c-v8ngc', 'IP address': '10.177.77.18'}, 'world_size': 2, 'master_port': 12354, 'tensorboard': True, 'num_epochs': 50, 'start_epoch': 1, 'start_batch': 0, 'exp_dir': PosixPath('pruned_transducer_stateless7/exp'), 'lang_dir': PosixPath('data/lang_char'), 'base_lr': 0.05, 'lr_batches': 5000, 'lr_epochs': 6, 'context_size': 1, 'prune_range': 5, 'lm_scale': 0.25, 'am_scale': 0.0, 'simple_loss_scale': 0.5, 'seed': 42, 'print_diagnostics': False, 'inf_check': False, 'save_every_n': 4000, 'keep_last_k': 30, 'average_period': 200, 'use_fp16': True, 'num_encoder_layers': '2,4,3,2,4', 'feedforward_dims': '1024,1024,2048,2048,1024', 'nhead': '8,8,8,8,8', 'encoder_dims': '384,384,384,384,384', 'attention_dims': '192,192,192,192,192', 'encoder_unmasked_dims': '256,256,256,256,256', 'zipformer_downsampling_factors': '1,2,4,8,2', 'cnn_module_kernels': '31,31,31,31,31', 'decoder_dim': 512, 'joiner_dim': 512, 'max_duration': 300, 'bucketing_sampler': True, 'num_buckets': 30, 'shuffle': True, 'return_cuts': True, 'num_workers': 2, 'enable_spec_aug': True, 'spec_aug_time_warp_factor': 80, 'enable_musan': True, 'manifest_dir': PosixPath('data/fbank'), 'on_the_fly_feats': False, 'blank_id': 0, 'vocab_size': 4336} +2023-03-20 17:19:56,681 INFO [train.py:995] (0/2) About to create model +2023-03-20 17:19:57,356 INFO [zipformer.py:178] (0/2) 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. +2023-03-20 17:19:57,459 INFO [train.py:999] (0/2) Number of model parameters: 77741923 +2023-03-20 17:20:00,307 INFO [train.py:1014] (0/2) Using DDP +2023-03-20 17:20:00,493 INFO [aishell.py:39] (0/2) About to get train cuts from data/fbank/aishell_cuts_train.jsonl.gz +2023-03-20 17:20:02,677 INFO [asr_datamodule.py:161] (0/2) Enable MUSAN +2023-03-20 17:20:02,677 INFO [asr_datamodule.py:171] (0/2) Enable SpecAugment +2023-03-20 17:20:02,677 INFO [asr_datamodule.py:172] (0/2) Time warp factor: 80 +2023-03-20 17:20:02,677 INFO [asr_datamodule.py:182] (0/2) Num frame mask: 10 +2023-03-20 17:20:02,677 INFO [asr_datamodule.py:195] (0/2) About to create train dataset +2023-03-20 17:20:02,677 INFO [asr_datamodule.py:223] (0/2) Using DynamicBucketingSampler. +2023-03-20 17:20:02,707 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-20 17:20:02,755 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-20 17:20:02,778 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-20 17:20:02,795 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-20 17:20:02,955 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-20 17:20:02,971 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-20 17:20:02,981 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-20 17:20:03,076 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-20 17:20:03,160 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-20 17:20:03,181 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-20 17:20:03,201 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-20 17:20:03,838 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-20 17:20:03,843 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-20 17:20:03,851 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-20 17:20:03,990 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-20 17:20:03,993 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-20 17:20:04,052 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-20 17:20:04,962 INFO [asr_datamodule.py:236] (0/2) About to create train dataloader +2023-03-20 17:20:04,962 INFO [aishell.py:45] (0/2) About to get valid cuts from data/fbank/aishell_cuts_dev.jsonl.gz +2023-03-20 17:20:04,964 INFO [asr_datamodule.py:249] (0/2) About to create dev dataset +2023-03-20 17:20:05,437 INFO [asr_datamodule.py:266] (0/2) About to create dev dataloader +2023-03-20 17:20:05,437 INFO [train.py:1129] (0/2) start training from epoch 1 +2023-03-20 17:20:14,029 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-20 17:20:14,087 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-20 17:20:14,115 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-20 17:20:14,134 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-20 17:20:14,335 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-20 17:20:14,871 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-20 17:20:14,882 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-20 17:20:14,985 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-20 17:20:15,060 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-20 17:20:15,076 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-20 17:20:15,093 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-20 17:20:15,261 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-20 17:20:15,266 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-20 17:20:15,274 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-20 17:20:15,413 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-20 17:20:15,416 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-20 17:20:15,474 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-20 17:20:19,862 INFO [train.py:901] (0/2) Epoch 1, batch 0, loss[loss=5.306, simple_loss=8.781, pruned_loss=9.113, over 7196.00 frames. ], tot_loss[loss=5.306, simple_loss=8.781, pruned_loss=9.113, over 7196.00 frames. ], batch size: 39, lr: 2.50e-02, grad_scale: 2.0 +2023-03-20 17:20:19,864 INFO [train.py:926] (0/2) Computing validation loss +2023-03-20 17:20:45,995 INFO [train.py:935] (0/2) Epoch 1, validation: loss=4.989, simple_loss=8.202, pruned_loss=8.839, over 1622729.00 frames. +2023-03-20 17:20:45,996 INFO [train.py:936] (0/2) Maximum memory allocated so far is 8242MB +2023-03-20 17:20:47,592 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=5.0, num_to_drop=2, layers_to_drop={1, 3} +2023-03-20 17:20:51,647 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-20 17:20:54,309 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=23.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:21:00,287 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-20 17:21:05,508 INFO [train.py:901] (0/2) Epoch 1, batch 50, loss[loss=0.4301, simple_loss=0.6621, pruned_loss=0.8059, over 5705.00 frames. ], tot_loss[loss=1.274, simple_loss=2.145, pruned_loss=1.876, over 320433.25 frames. ], batch size: 25, lr: 2.75e-02, grad_scale: 1.0 +2023-03-20 17:21:05,773 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=5.50 vs. limit=2.0 +2023-03-20 17:21:06,266 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-20 17:21:07,820 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-20 17:21:09,372 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3759, 3.4878, 3.5232, 3.5261, 3.4107, 3.4176, 2.5900, 3.5219], + device='cuda:0'), covar=tensor([0.0026, 0.0015, 0.0013, 0.0006, 0.0017, 0.0013, 0.0040, 0.0015], + device='cuda:0'), in_proj_covar=tensor([0.0009, 0.0009, 0.0009, 0.0009, 0.0009, 0.0009, 0.0009, 0.0009], + device='cuda:0'), out_proj_covar=tensor([9.2471e-06, 9.2057e-06, 9.3445e-06, 9.1950e-06, 9.3865e-06, 9.1799e-06, + 9.2620e-06, 9.2446e-06], device='cuda:0') +2023-03-20 17:21:09,646 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-20 17:21:17,556 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=83.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:21:23,052 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-20 17:21:23,438 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-20 17:21:24,936 INFO [train.py:901] (0/2) Epoch 1, batch 100, loss[loss=0.5086, simple_loss=0.7411, pruned_loss=0.9491, over 7186.00 frames. ], tot_loss[loss=0.8559, simple_loss=1.387, pruned_loss=1.37, over 569445.52 frames. ], batch size: 41, lr: 3.00e-02, grad_scale: 1.0 +2023-03-20 17:21:25,320 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.002e+01 5.426e+01 1.031e+02 2.944e+02 4.647e+03, threshold=2.062e+02, percent-clipped=0.0 +2023-03-20 17:21:39,817 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=13.14 vs. limit=2.0 +2023-03-20 17:21:41,984 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=144.0, num_to_drop=2, layers_to_drop={0, 2} +2023-03-20 17:21:42,917 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=5.35 vs. limit=2.0 +2023-03-20 17:21:44,523 INFO [train.py:901] (0/2) Epoch 1, batch 150, loss[loss=0.5134, simple_loss=0.7228, pruned_loss=0.9047, over 7303.00 frames. ], tot_loss[loss=0.7066, simple_loss=1.106, pruned_loss=1.174, over 762840.61 frames. ], batch size: 68, lr: 3.25e-02, grad_scale: 1.0 +2023-03-20 17:21:45,909 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=5.04 vs. limit=2.0 +2023-03-20 17:21:57,432 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=4.28 vs. limit=2.0 +2023-03-20 17:22:06,003 INFO [train.py:901] (0/2) Epoch 1, batch 200, loss[loss=0.5184, simple_loss=0.7274, pruned_loss=0.8122, over 6813.00 frames. ], tot_loss[loss=0.6298, simple_loss=0.9572, pruned_loss=1.049, over 912343.19 frames. ], batch size: 107, lr: 3.50e-02, grad_scale: 1.0 +2023-03-20 17:22:06,368 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.567e+01 2.396e+01 3.102e+01 4.606e+01 1.207e+02, threshold=6.203e+01, percent-clipped=0.0 +2023-03-20 17:22:10,624 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-20 17:22:13,965 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-20 17:22:18,522 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-20 17:22:25,478 INFO [train.py:901] (0/2) Epoch 1, batch 250, loss[loss=0.4933, simple_loss=0.6705, pruned_loss=0.7424, over 7335.00 frames. ], tot_loss[loss=0.5854, simple_loss=0.8675, pruned_loss=0.9571, over 1031364.41 frames. ], batch size: 63, lr: 3.75e-02, grad_scale: 1.0 +2023-03-20 17:22:27,739 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-20 17:22:31,934 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4036, 3.4050, 3.4054, 3.4053, 3.4054, 3.4051, 3.4053, 3.4049], + device='cuda:0'), covar=tensor([0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0'), in_proj_covar=tensor([0.0009, 0.0009, 0.0009, 0.0009, 0.0009, 0.0009, 0.0009, 0.0009], + device='cuda:0'), out_proj_covar=tensor([8.8813e-06, 8.7014e-06, 9.0176e-06, 8.7719e-06, 9.0519e-06, 8.8105e-06, + 9.0172e-06, 8.8638e-06], device='cuda:0') +2023-03-20 17:22:35,380 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.59 vs. limit=2.0 +2023-03-20 17:22:37,217 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=7.80 vs. limit=2.0 +2023-03-20 17:22:38,385 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=4.69 vs. limit=2.0 +2023-03-20 17:22:41,810 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=3.66 vs. limit=2.0 +2023-03-20 17:22:43,150 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=296.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:22:43,827 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-20 17:22:44,627 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=300.0, num_to_drop=2, layers_to_drop={0, 2} +2023-03-20 17:22:44,924 INFO [train.py:901] (0/2) Epoch 1, batch 300, loss[loss=0.4921, simple_loss=0.6553, pruned_loss=0.6985, over 7317.00 frames. ], tot_loss[loss=0.5584, simple_loss=0.8083, pruned_loss=0.8885, over 1121649.38 frames. ], batch size: 75, lr: 4.00e-02, grad_scale: 1.0 +2023-03-20 17:22:45,279 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.093e+01 3.024e+01 4.407e+01 5.736e+01 1.326e+02, threshold=8.814e+01, percent-clipped=19.0 +2023-03-20 17:22:50,399 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-20 17:22:52,451 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.20 vs. limit=2.0 +2023-03-20 17:23:03,914 INFO [train.py:901] (0/2) Epoch 1, batch 350, loss[loss=0.5107, simple_loss=0.6553, pruned_loss=0.7096, over 7265.00 frames. ], tot_loss[loss=0.5397, simple_loss=0.7626, pruned_loss=0.8342, over 1192768.70 frames. ], batch size: 52, lr: 4.25e-02, grad_scale: 1.0 +2023-03-20 17:23:04,463 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=15.24 vs. limit=5.0 +2023-03-20 17:23:04,860 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=12.83 vs. limit=5.0 +2023-03-20 17:23:05,984 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=3.20 vs. limit=2.0 +2023-03-20 17:23:06,274 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=357.0, num_to_drop=2, layers_to_drop={0, 1} +2023-03-20 17:23:06,437 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=13.20 vs. limit=5.0 +2023-03-20 17:23:16,940 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-20 17:23:18,075 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=387.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:23:23,745 INFO [train.py:901] (0/2) Epoch 1, batch 400, loss[loss=0.4845, simple_loss=0.6076, pruned_loss=0.6442, over 7294.00 frames. ], tot_loss[loss=0.5299, simple_loss=0.7307, pruned_loss=0.7935, over 1249240.49 frames. ], batch size: 42, lr: 4.50e-02, grad_scale: 2.0 +2023-03-20 17:23:24,107 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.730e+01 3.441e+01 4.515e+01 6.284e+01 2.215e+02, threshold=9.029e+01, percent-clipped=10.0 +2023-03-20 17:23:26,285 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-03-20 17:23:35,776 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.75 vs. limit=5.0 +2023-03-20 17:23:38,542 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=439.0, num_to_drop=2, layers_to_drop={0, 2} +2023-03-20 17:23:41,860 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=448.0, num_to_drop=2, layers_to_drop={0, 1} +2023-03-20 17:23:42,914 INFO [train.py:901] (0/2) Epoch 1, batch 450, loss[loss=0.5043, simple_loss=0.6187, pruned_loss=0.6433, over 7303.00 frames. ], tot_loss[loss=0.5265, simple_loss=0.7086, pruned_loss=0.7627, over 1290415.22 frames. ], batch size: 83, lr: 4.75e-02, grad_scale: 2.0 +2023-03-20 17:23:47,414 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-20 17:23:48,340 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-20 17:23:55,810 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3681, 4.4014, 4.3769, 4.3823, 4.3668, 4.3482, 4.3389, 4.3652], + device='cuda:0'), covar=tensor([0.0095, 0.0072, 0.0077, 0.0081, 0.0082, 0.0120, 0.0114, 0.0119], + device='cuda:0'), in_proj_covar=tensor([0.0010, 0.0011, 0.0010, 0.0010, 0.0010, 0.0010, 0.0010, 0.0010], + device='cuda:0'), out_proj_covar=tensor([9.1818e-06, 9.4909e-06, 9.1386e-06, 9.6156e-06, 9.6520e-06, 9.8047e-06, + 9.2184e-06, 9.3970e-06], device='cuda:0') +2023-03-20 17:24:00,214 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7132, 3.9324, 3.8745, 3.8485, 3.6465, 3.7932, 3.9875, 3.6691], + device='cuda:0'), covar=tensor([0.0602, 0.0225, 0.0215, 0.0360, 0.0465, 0.0386, 0.0178, 0.0635], + device='cuda:0'), in_proj_covar=tensor([0.0011, 0.0010, 0.0010, 0.0010, 0.0010, 0.0009, 0.0011, 0.0010], + device='cuda:0'), out_proj_covar=tensor([9.8786e-06, 9.2900e-06, 9.4879e-06, 9.5815e-06, 9.5622e-06, 9.4508e-06, + 9.5893e-06, 9.6932e-06], device='cuda:0') +2023-03-20 17:24:01,953 INFO [train.py:901] (0/2) Epoch 1, batch 500, loss[loss=0.5238, simple_loss=0.634, pruned_loss=0.6355, over 7287.00 frames. ], tot_loss[loss=0.525, simple_loss=0.6906, pruned_loss=0.7345, over 1324475.18 frames. ], batch size: 77, lr: 4.99e-02, grad_scale: 2.0 +2023-03-20 17:24:02,311 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.353e+01 4.338e+01 5.607e+01 7.790e+01 1.498e+02, threshold=1.121e+02, percent-clipped=18.0 +2023-03-20 17:24:12,203 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-20 17:24:13,319 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-20 17:24:13,666 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-20 17:24:15,146 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-20 17:24:18,439 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-20 17:24:21,052 INFO [train.py:901] (0/2) Epoch 1, batch 550, loss[loss=0.4428, simple_loss=0.5289, pruned_loss=0.5125, over 7023.00 frames. ], tot_loss[loss=0.5204, simple_loss=0.6716, pruned_loss=0.7014, over 1350134.40 frames. ], batch size: 35, lr: 4.98e-02, grad_scale: 2.0 +2023-03-20 17:24:22,239 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1729, 2.4760, 2.0852, 2.3854, 2.0622, 2.2022, 2.1715, 2.1143], + device='cuda:0'), covar=tensor([1.8006, 0.5703, 1.1742, 0.7650, 0.9964, 0.8646, 0.6347, 0.6602], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0014, 0.0016, 0.0015, 0.0015, 0.0013, 0.0015, 0.0014], + device='cuda:0'), out_proj_covar=tensor([1.6185e-05, 1.2105e-05, 1.4123e-05, 1.3509e-05, 1.4016e-05, 1.2495e-05, + 1.2263e-05, 1.2646e-05], device='cuda:0') +2023-03-20 17:24:25,655 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=562.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:24:26,086 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=14.43 vs. limit=5.0 +2023-03-20 17:24:27,753 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-20 17:24:27,850 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=568.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:24:29,833 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=7.89 vs. limit=5.0 +2023-03-20 17:24:33,226 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-20 17:24:35,823 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-20 17:24:35,909 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=590.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:24:40,198 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=600.0, num_to_drop=2, layers_to_drop={1, 2} +2023-03-20 17:24:40,469 INFO [train.py:901] (0/2) Epoch 1, batch 600, loss[loss=0.4807, simple_loss=0.5813, pruned_loss=0.513, over 7255.00 frames. ], tot_loss[loss=0.5133, simple_loss=0.653, pruned_loss=0.6636, over 1370751.31 frames. ], batch size: 64, lr: 4.98e-02, grad_scale: 2.0 +2023-03-20 17:24:40,841 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 5.139e+01 1.040e+02 1.577e+02 2.315e+02 8.829e+02, threshold=3.155e+02, percent-clipped=69.0 +2023-03-20 17:24:41,621 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-20 17:24:42,254 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=3.70 vs. limit=2.0 +2023-03-20 17:24:49,329 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=623.0, num_to_drop=2, layers_to_drop={1, 2} +2023-03-20 17:24:51,605 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=629.0, num_to_drop=2, layers_to_drop={0, 1} +2023-03-20 17:24:53,779 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-20 17:24:59,229 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=648.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:25:00,341 INFO [train.py:901] (0/2) Epoch 1, batch 650, loss[loss=0.4406, simple_loss=0.5356, pruned_loss=0.4397, over 7136.00 frames. ], tot_loss[loss=0.5024, simple_loss=0.6324, pruned_loss=0.6217, over 1386157.32 frames. ], batch size: 41, lr: 4.98e-02, grad_scale: 2.0 +2023-03-20 17:25:00,455 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=651.0, num_to_drop=2, layers_to_drop={0, 2} +2023-03-20 17:25:00,712 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-20 17:25:00,762 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=652.0, num_to_drop=2, layers_to_drop={2, 3} +2023-03-20 17:25:12,942 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-20 17:25:18,781 INFO [train.py:901] (0/2) Epoch 1, batch 700, loss[loss=0.4069, simple_loss=0.496, pruned_loss=0.3826, over 7349.00 frames. ], tot_loss[loss=0.4897, simple_loss=0.6112, pruned_loss=0.5794, over 1400457.96 frames. ], batch size: 44, lr: 4.98e-02, grad_scale: 2.0 +2023-03-20 17:25:19,156 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.124e+02 1.970e+02 2.813e+02 3.918e+02 8.895e+02, threshold=5.626e+02, percent-clipped=42.0 +2023-03-20 17:25:19,186 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-20 17:25:33,330 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=739.0, num_to_drop=2, layers_to_drop={0, 2} +2023-03-20 17:25:35,293 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=743.0, num_to_drop=2, layers_to_drop={0, 2} +2023-03-20 17:25:37,991 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-20 17:25:38,688 INFO [train.py:901] (0/2) Epoch 1, batch 750, loss[loss=0.415, simple_loss=0.5023, pruned_loss=0.3742, over 7303.00 frames. ], tot_loss[loss=0.4748, simple_loss=0.5887, pruned_loss=0.5373, over 1408979.06 frames. ], batch size: 68, lr: 4.97e-02, grad_scale: 2.0 +2023-03-20 17:25:38,697 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-20 17:25:48,255 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-20 17:25:51,942 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-20 17:25:51,994 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=787.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:25:55,997 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-20 17:25:58,329 INFO [train.py:901] (0/2) Epoch 1, batch 800, loss[loss=0.4225, simple_loss=0.499, pruned_loss=0.3757, over 7254.00 frames. ], tot_loss[loss=0.462, simple_loss=0.569, pruned_loss=0.5002, over 1413236.97 frames. ], batch size: 89, lr: 4.97e-02, grad_scale: 4.0 +2023-03-20 17:25:58,339 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-20 17:25:58,689 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.542e+02 2.725e+02 3.452e+02 4.919e+02 8.420e+02, threshold=6.904e+02, percent-clipped=17.0 +2023-03-20 17:26:01,060 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=7.43 vs. limit=5.0 +2023-03-20 17:26:02,529 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=7.67 vs. limit=5.0 +2023-03-20 17:26:03,000 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.27 vs. limit=2.0 +2023-03-20 17:26:06,088 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-20 17:26:09,284 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.60 vs. limit=5.0 +2023-03-20 17:26:13,194 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.12 vs. limit=2.0 +2023-03-20 17:26:16,069 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=847.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:26:17,480 INFO [train.py:901] (0/2) Epoch 1, batch 850, loss[loss=0.4211, simple_loss=0.5011, pruned_loss=0.3532, over 7346.00 frames. ], tot_loss[loss=0.4486, simple_loss=0.5496, pruned_loss=0.4644, over 1421409.35 frames. ], batch size: 54, lr: 4.96e-02, grad_scale: 4.0 +2023-03-20 17:26:20,168 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-20 17:26:20,541 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-20 17:26:24,596 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-20 17:26:27,194 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-20 17:26:36,616 INFO [train.py:901] (0/2) Epoch 1, batch 900, loss[loss=0.3712, simple_loss=0.4436, pruned_loss=0.2956, over 7331.00 frames. ], tot_loss[loss=0.4373, simple_loss=0.5326, pruned_loss=0.4337, over 1422458.11 frames. ], batch size: 49, lr: 4.96e-02, grad_scale: 4.0 +2023-03-20 17:26:36,975 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.109e+02 2.661e+02 3.954e+02 5.359e+02 9.011e+02, threshold=7.908e+02, percent-clipped=11.0 +2023-03-20 17:26:39,313 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=908.0, num_to_drop=2, layers_to_drop={1, 3} +2023-03-20 17:26:43,020 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=918.0, num_to_drop=2, layers_to_drop={1, 3} +2023-03-20 17:26:45,209 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=924.0, num_to_drop=2, layers_to_drop={2, 3} +2023-03-20 17:26:53,855 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=946.0, num_to_drop=2, layers_to_drop={1, 3} +2023-03-20 17:26:55,133 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-03-20 17:26:55,317 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-20 17:26:55,652 INFO [train.py:901] (0/2) Epoch 1, batch 950, loss[loss=0.4333, simple_loss=0.4939, pruned_loss=0.353, over 7308.00 frames. ], tot_loss[loss=0.4283, simple_loss=0.5187, pruned_loss=0.4063, over 1428218.04 frames. ], batch size: 83, lr: 4.96e-02, grad_scale: 4.0 +2023-03-20 17:26:56,092 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=952.0, num_to_drop=2, layers_to_drop={1, 2} +2023-03-20 17:27:07,404 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.12 vs. limit=5.0 +2023-03-20 17:27:12,620 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-20 17:27:15,033 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=1000.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:27:15,359 INFO [train.py:901] (0/2) Epoch 1, batch 1000, loss[loss=0.4148, simple_loss=0.4733, pruned_loss=0.3235, over 7309.00 frames. ], tot_loss[loss=0.4199, simple_loss=0.5059, pruned_loss=0.3819, over 1430120.72 frames. ], batch size: 49, lr: 4.95e-02, grad_scale: 4.0 +2023-03-20 17:27:15,712 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.866e+02 3.326e+02 4.317e+02 5.927e+02 1.638e+03, threshold=8.634e+02, percent-clipped=6.0 +2023-03-20 17:27:28,666 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-20 17:27:29,221 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.99 vs. limit=2.0 +2023-03-20 17:27:32,076 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=1043.0, num_to_drop=2, layers_to_drop={0, 1} +2023-03-20 17:27:35,076 INFO [train.py:901] (0/2) Epoch 1, batch 1050, loss[loss=0.3658, simple_loss=0.4319, pruned_loss=0.2616, over 7303.00 frames. ], tot_loss[loss=0.4122, simple_loss=0.4939, pruned_loss=0.36, over 1431175.99 frames. ], batch size: 49, lr: 4.95e-02, grad_scale: 4.0 +2023-03-20 17:27:43,871 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-20 17:27:47,609 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-20 17:27:51,125 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=1091.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:27:55,031 INFO [train.py:901] (0/2) Epoch 1, batch 1100, loss[loss=0.3194, simple_loss=0.3757, pruned_loss=0.2209, over 7037.00 frames. ], tot_loss[loss=0.4047, simple_loss=0.4828, pruned_loss=0.3394, over 1434315.47 frames. ], batch size: 35, lr: 4.94e-02, grad_scale: 4.0 +2023-03-20 17:27:55,397 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.549e+02 3.735e+02 4.896e+02 6.513e+02 1.819e+03, threshold=9.792e+02, percent-clipped=10.0 +2023-03-20 17:28:09,542 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-20 17:28:09,887 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 17:28:14,560 INFO [train.py:901] (0/2) Epoch 1, batch 1150, loss[loss=0.3617, simple_loss=0.4203, pruned_loss=0.2453, over 7152.00 frames. ], tot_loss[loss=0.4003, simple_loss=0.4748, pruned_loss=0.323, over 1437412.95 frames. ], batch size: 41, lr: 4.94e-02, grad_scale: 4.0 +2023-03-20 17:28:15,422 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=1153.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:28:19,110 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-20 17:28:19,877 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-20 17:28:34,655 INFO [train.py:901] (0/2) Epoch 1, batch 1200, loss[loss=0.3344, simple_loss=0.3903, pruned_loss=0.2174, over 7262.00 frames. ], tot_loss[loss=0.3931, simple_loss=0.4642, pruned_loss=0.3055, over 1439099.64 frames. ], batch size: 47, lr: 4.93e-02, grad_scale: 8.0 +2023-03-20 17:28:35,019 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.042e+02 3.708e+02 4.369e+02 5.950e+02 1.977e+03, threshold=8.739e+02, percent-clipped=4.0 +2023-03-20 17:28:35,471 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=1203.0, num_to_drop=2, layers_to_drop={0, 1} +2023-03-20 17:28:40,247 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=1214.0, num_to_drop=2, layers_to_drop={0, 1} +2023-03-20 17:28:40,711 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.45 vs. limit=5.0 +2023-03-20 17:28:41,757 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=1218.0, num_to_drop=2, layers_to_drop={0, 1} +2023-03-20 17:28:43,964 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=1224.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 17:28:45,713 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-20 17:28:52,363 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=1246.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:28:54,256 INFO [train.py:901] (0/2) Epoch 1, batch 1250, loss[loss=0.3847, simple_loss=0.4299, pruned_loss=0.256, over 7328.00 frames. ], tot_loss[loss=0.3908, simple_loss=0.4581, pruned_loss=0.2937, over 1440729.12 frames. ], batch size: 75, lr: 4.92e-02, grad_scale: 8.0 +2023-03-20 17:28:56,523 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0631, 3.3146, 2.8472, 2.8482, 2.6558, 2.8701, 2.8577, 3.0366], + device='cuda:0'), covar=tensor([0.1397, 0.0922, 0.1408, 0.1475, 0.2131, 0.1410, 0.2085, 0.0926], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0027, 0.0030, 0.0031, 0.0033, 0.0031, 0.0031, 0.0029], + device='cuda:0'), out_proj_covar=tensor([2.1451e-05, 2.1666e-05, 2.4835e-05, 2.3659e-05, 2.8102e-05, 2.5499e-05, + 2.2998e-05, 2.2696e-05], device='cuda:0') +2023-03-20 17:28:59,902 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=1266.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:29:02,718 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=1272.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:29:04,305 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-20 17:29:04,490 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.97 vs. limit=5.0 +2023-03-20 17:29:07,527 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-20 17:29:08,278 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-20 17:29:10,778 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8454, 3.8532, 3.9775, 3.8161, 3.8993, 3.6815, 3.7979, 3.8262], + device='cuda:0'), covar=tensor([0.0655, 0.0618, 0.0570, 0.0964, 0.0737, 0.1190, 0.0764, 0.1061], + device='cuda:0'), in_proj_covar=tensor([0.0013, 0.0012, 0.0013, 0.0014, 0.0012, 0.0015, 0.0013, 0.0015], + device='cuda:0'), out_proj_covar=tensor([9.0250e-06, 8.1316e-06, 1.0131e-05, 9.9989e-06, 8.4353e-06, 1.0912e-05, + 8.8401e-06, 1.1342e-05], device='cuda:0') +2023-03-20 17:29:11,476 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=1294.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:29:14,205 INFO [train.py:901] (0/2) Epoch 1, batch 1300, loss[loss=0.4594, simple_loss=0.491, pruned_loss=0.312, over 6703.00 frames. ], tot_loss[loss=0.387, simple_loss=0.4507, pruned_loss=0.2816, over 1441510.46 frames. ], batch size: 106, lr: 4.92e-02, grad_scale: 8.0 +2023-03-20 17:29:14,565 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.819e+02 3.657e+02 4.833e+02 5.566e+02 9.665e+02, threshold=9.665e+02, percent-clipped=1.0 +2023-03-20 17:29:25,630 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-20 17:29:26,218 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=6.04 vs. limit=5.0 +2023-03-20 17:29:27,268 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-20 17:29:30,086 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-20 17:29:34,434 INFO [train.py:901] (0/2) Epoch 1, batch 1350, loss[loss=0.2751, simple_loss=0.3121, pruned_loss=0.1682, over 6406.00 frames. ], tot_loss[loss=0.3838, simple_loss=0.4444, pruned_loss=0.2704, over 1441161.31 frames. ], batch size: 28, lr: 4.91e-02, grad_scale: 8.0 +2023-03-20 17:29:38,742 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-20 17:29:55,967 INFO [train.py:901] (0/2) Epoch 1, batch 1400, loss[loss=0.3441, simple_loss=0.3919, pruned_loss=0.2029, over 7354.00 frames. ], tot_loss[loss=0.3803, simple_loss=0.4378, pruned_loss=0.26, over 1440469.38 frames. ], batch size: 63, lr: 4.91e-02, grad_scale: 8.0 +2023-03-20 17:29:56,337 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.053e+02 4.073e+02 4.817e+02 5.677e+02 9.535e+02, threshold=9.634e+02, percent-clipped=0.0 +2023-03-20 17:30:06,475 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-20 17:30:09,425 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=1434.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:30:16,031 INFO [train.py:901] (0/2) Epoch 1, batch 1450, loss[loss=0.4369, simple_loss=0.4666, pruned_loss=0.2703, over 7322.00 frames. ], tot_loss[loss=0.3783, simple_loss=0.4321, pruned_loss=0.2518, over 1439790.58 frames. ], batch size: 75, lr: 4.90e-02, grad_scale: 8.0 +2023-03-20 17:30:25,502 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-20 17:30:35,454 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=1495.0, num_to_drop=2, layers_to_drop={1, 2} +2023-03-20 17:30:37,665 INFO [train.py:901] (0/2) Epoch 1, batch 1500, loss[loss=0.3445, simple_loss=0.3906, pruned_loss=0.1925, over 7295.00 frames. ], tot_loss[loss=0.3784, simple_loss=0.4287, pruned_loss=0.2452, over 1440205.84 frames. ], batch size: 80, lr: 4.89e-02, grad_scale: 8.0 +2023-03-20 17:30:38,104 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.828e+02 4.191e+02 5.686e+02 6.667e+02 2.069e+03, threshold=1.137e+03, percent-clipped=7.0 +2023-03-20 17:30:38,615 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=1503.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:30:39,398 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-20 17:30:41,134 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=1509.0, num_to_drop=2, layers_to_drop={0, 3} +2023-03-20 17:30:46,349 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.12 vs. limit=2.0 +2023-03-20 17:30:48,736 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.76 vs. limit=2.0 +2023-03-20 17:30:58,650 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-20 17:30:59,044 INFO [train.py:901] (0/2) Epoch 1, batch 1550, loss[loss=0.3403, simple_loss=0.3782, pruned_loss=0.1895, over 7220.00 frames. ], tot_loss[loss=0.3758, simple_loss=0.4231, pruned_loss=0.2374, over 1439364.50 frames. ], batch size: 45, lr: 4.89e-02, grad_scale: 8.0 +2023-03-20 17:30:59,104 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=1551.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:31:05,361 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=1566.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:31:20,769 INFO [train.py:901] (0/2) Epoch 1, batch 1600, loss[loss=0.435, simple_loss=0.4553, pruned_loss=0.2527, over 7269.00 frames. ], tot_loss[loss=0.3741, simple_loss=0.4185, pruned_loss=0.2305, over 1440676.87 frames. ], batch size: 52, lr: 4.88e-02, grad_scale: 8.0 +2023-03-20 17:31:21,139 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.008e+02 4.402e+02 5.060e+02 6.395e+02 1.208e+03, threshold=1.012e+03, percent-clipped=1.0 +2023-03-20 17:31:24,907 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-20 17:31:24,990 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=1611.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:31:25,283 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-20 17:31:28,038 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-20 17:31:31,354 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=1627.0, num_to_drop=2, layers_to_drop={1, 3} +2023-03-20 17:31:36,414 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-20 17:31:39,311 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-20 17:31:41,599 INFO [train.py:901] (0/2) Epoch 1, batch 1650, loss[loss=0.4179, simple_loss=0.4404, pruned_loss=0.2345, over 7232.00 frames. ], tot_loss[loss=0.3745, simple_loss=0.4155, pruned_loss=0.2255, over 1441765.53 frames. ], batch size: 93, lr: 4.87e-02, grad_scale: 8.0 +2023-03-20 17:31:43,005 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5847, 3.5410, 3.7068, 3.6140, 3.7259, 2.9804, 3.7370, 3.4184], + device='cuda:0'), covar=tensor([0.0450, 0.0532, 0.0447, 0.0352, 0.0251, 0.0814, 0.0460, 0.0520], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0021, 0.0024, 0.0021, 0.0020, 0.0023, 0.0024, 0.0021], + device='cuda:0'), out_proj_covar=tensor([1.6849e-05, 1.7814e-05, 2.1824e-05, 1.6536e-05, 1.5879e-05, 1.9348e-05, + 2.0199e-05, 1.8105e-05], device='cuda:0') +2023-03-20 17:31:46,675 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-20 17:31:50,558 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=1672.0, num_to_drop=2, layers_to_drop={2, 3} +2023-03-20 17:31:52,758 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.96 vs. limit=5.0 +2023-03-20 17:32:01,074 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 17:32:03,118 INFO [train.py:901] (0/2) Epoch 1, batch 1700, loss[loss=0.3483, simple_loss=0.3671, pruned_loss=0.1904, over 6929.00 frames. ], tot_loss[loss=0.3758, simple_loss=0.4133, pruned_loss=0.2215, over 1441836.39 frames. ], batch size: 35, lr: 4.86e-02, grad_scale: 8.0 +2023-03-20 17:32:03,545 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.196e+02 4.136e+02 5.429e+02 6.545e+02 1.463e+03, threshold=1.086e+03, percent-clipped=5.0 +2023-03-20 17:32:04,768 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-20 17:32:09,942 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6940, 3.5746, 3.9804, 3.7650, 3.9164, 3.4303, 3.8449, 3.7093], + device='cuda:0'), covar=tensor([0.0386, 0.0761, 0.0340, 0.0329, 0.0308, 0.0732, 0.0468, 0.0463], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0020, 0.0023, 0.0020, 0.0020, 0.0023, 0.0023, 0.0021], + device='cuda:0'), out_proj_covar=tensor([1.6186e-05, 1.6900e-05, 2.0077e-05, 1.5934e-05, 1.5690e-05, 1.9491e-05, + 1.9481e-05, 1.8094e-05], device='cuda:0') +2023-03-20 17:32:12,575 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.11 vs. limit=5.0 +2023-03-20 17:32:13,617 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-20 17:32:24,814 INFO [train.py:901] (0/2) Epoch 1, batch 1750, loss[loss=0.351, simple_loss=0.386, pruned_loss=0.1779, over 7270.00 frames. ], tot_loss[loss=0.3754, simple_loss=0.4104, pruned_loss=0.2164, over 1441003.46 frames. ], batch size: 70, lr: 4.86e-02, grad_scale: 8.0 +2023-03-20 17:32:26,309 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.17 vs. limit=2.0 +2023-03-20 17:32:33,819 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-20 17:32:34,667 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-20 17:32:41,812 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=1790.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 17:32:46,789 INFO [train.py:901] (0/2) Epoch 1, batch 1800, loss[loss=0.3414, simple_loss=0.3799, pruned_loss=0.1663, over 7278.00 frames. ], tot_loss[loss=0.3749, simple_loss=0.4076, pruned_loss=0.2113, over 1442805.97 frames. ], batch size: 77, lr: 4.85e-02, grad_scale: 8.0 +2023-03-20 17:32:47,179 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.610e+02 3.716e+02 4.744e+02 6.070e+02 1.750e+03, threshold=9.488e+02, percent-clipped=2.0 +2023-03-20 17:32:50,290 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=1809.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:32:53,907 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-20 17:33:03,014 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=1838.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:33:06,314 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-20 17:33:07,354 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.79 vs. limit=5.0 +2023-03-20 17:33:08,391 INFO [train.py:901] (0/2) Epoch 1, batch 1850, loss[loss=0.4375, simple_loss=0.4369, pruned_loss=0.2348, over 7296.00 frames. ], tot_loss[loss=0.3734, simple_loss=0.4035, pruned_loss=0.2062, over 1443760.47 frames. ], batch size: 86, lr: 4.84e-02, grad_scale: 8.0 +2023-03-20 17:33:11,059 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=1857.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:33:11,134 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=1857.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:33:11,277 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.91 vs. limit=2.0 +2023-03-20 17:33:14,518 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-20 17:33:16,736 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=1870.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:33:29,455 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-20 17:33:29,570 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=1899.0, num_to_drop=2, layers_to_drop={0, 1} +2023-03-20 17:33:30,369 INFO [train.py:901] (0/2) Epoch 1, batch 1900, loss[loss=0.3617, simple_loss=0.3875, pruned_loss=0.1758, over 7341.00 frames. ], tot_loss[loss=0.3731, simple_loss=0.3999, pruned_loss=0.2023, over 1442307.30 frames. ], batch size: 73, lr: 4.83e-02, grad_scale: 8.0 +2023-03-20 17:33:30,778 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.126e+02 4.333e+02 5.813e+02 7.332e+02 1.551e+03, threshold=1.163e+03, percent-clipped=10.0 +2023-03-20 17:33:37,833 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=1918.0, num_to_drop=2, layers_to_drop={1, 3} +2023-03-20 17:33:39,470 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=1922.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:33:44,027 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=1931.0, num_to_drop=2, layers_to_drop={0, 3} +2023-03-20 17:33:45,062 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.89 vs. limit=2.0 +2023-03-20 17:33:50,276 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-20 17:33:51,759 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-20 17:33:52,353 INFO [train.py:901] (0/2) Epoch 1, batch 1950, loss[loss=0.3658, simple_loss=0.3839, pruned_loss=0.1778, over 7309.00 frames. ], tot_loss[loss=0.3755, simple_loss=0.3987, pruned_loss=0.2003, over 1442673.23 frames. ], batch size: 59, lr: 4.83e-02, grad_scale: 8.0 +2023-03-20 17:33:56,316 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3677, 2.6820, 3.0717, 3.1049, 3.5018, 3.2198, 3.1225, 3.2335], + device='cuda:0'), covar=tensor([0.0305, 0.0601, 0.0403, 0.0412, 0.0272, 0.0414, 0.0419, 0.0376], + device='cuda:0'), in_proj_covar=tensor([0.0032, 0.0036, 0.0036, 0.0035, 0.0034, 0.0034, 0.0033, 0.0037], + device='cuda:0'), out_proj_covar=tensor([2.2845e-05, 2.5535e-05, 2.5840e-05, 2.5545e-05, 2.5167e-05, 2.4359e-05, + 2.2630e-05, 2.7382e-05], device='cuda:0') +2023-03-20 17:33:59,781 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=1967.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:34:01,016 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-20 17:34:04,645 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-20 17:34:05,027 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-20 17:34:14,559 INFO [train.py:901] (0/2) Epoch 1, batch 2000, loss[loss=0.3522, simple_loss=0.3695, pruned_loss=0.1675, over 7289.00 frames. ], tot_loss[loss=0.3733, simple_loss=0.3944, pruned_loss=0.1953, over 1444802.45 frames. ], batch size: 66, lr: 4.82e-02, grad_scale: 8.0 +2023-03-20 17:34:15,035 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.389e+02 4.542e+02 5.662e+02 7.090e+02 1.677e+03, threshold=1.132e+03, percent-clipped=2.0 +2023-03-20 17:34:20,158 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-20 17:34:30,877 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-20 17:34:36,209 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.97 vs. limit=2.0 +2023-03-20 17:34:36,586 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.14 vs. limit=2.0 +2023-03-20 17:34:38,084 INFO [train.py:901] (0/2) Epoch 1, batch 2050, loss[loss=0.3544, simple_loss=0.3738, pruned_loss=0.1675, over 7236.00 frames. ], tot_loss[loss=0.3703, simple_loss=0.3902, pruned_loss=0.1901, over 1442528.51 frames. ], batch size: 55, lr: 4.81e-02, grad_scale: 16.0 +2023-03-20 17:34:38,095 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-20 17:34:50,893 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8677, 3.3184, 3.1234, 3.0235, 3.0664, 2.9529, 2.7334, 3.2067], + device='cuda:0'), covar=tensor([0.0421, 0.0150, 0.0339, 0.0436, 0.0482, 0.0439, 0.0627, 0.0155], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0035, 0.0040, 0.0043, 0.0037, 0.0041, 0.0044, 0.0029], + device='cuda:0'), out_proj_covar=tensor([3.3295e-05, 2.5880e-05, 3.0299e-05, 3.4593e-05, 2.7711e-05, 3.1997e-05, + 3.5436e-05, 2.2161e-05], device='cuda:0') +2023-03-20 17:34:56,920 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=2090.0, num_to_drop=2, layers_to_drop={1, 2} +2023-03-20 17:35:01,733 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.04 vs. limit=2.0 +2023-03-20 17:35:01,815 INFO [train.py:901] (0/2) Epoch 1, batch 2100, loss[loss=0.3054, simple_loss=0.3091, pruned_loss=0.1509, over 6065.00 frames. ], tot_loss[loss=0.3671, simple_loss=0.3862, pruned_loss=0.1856, over 1440950.72 frames. ], batch size: 26, lr: 4.80e-02, grad_scale: 16.0 +2023-03-20 17:35:02,251 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.029e+02 3.619e+02 4.625e+02 6.076e+02 9.529e+02, threshold=9.250e+02, percent-clipped=0.0 +2023-03-20 17:35:08,958 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-20 17:35:12,346 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-20 17:35:19,196 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=2138.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:35:21,119 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2077, 3.3259, 3.1241, 3.3926, 3.2887, 3.4912, 3.0917, 3.2310], + device='cuda:0'), covar=tensor([0.0434, 0.0628, 0.0995, 0.0837, 0.0422, 0.0382, 0.1284, 0.0708], + device='cuda:0'), in_proj_covar=tensor([0.0011, 0.0013, 0.0016, 0.0015, 0.0012, 0.0012, 0.0016, 0.0014], + device='cuda:0'), out_proj_covar=tensor([6.3705e-06, 8.6981e-06, 1.1880e-05, 1.0346e-05, 6.8551e-06, 8.2737e-06, + 1.2742e-05, 9.2395e-06], device='cuda:0') +2023-03-20 17:35:25,653 INFO [train.py:901] (0/2) Epoch 1, batch 2150, loss[loss=0.3486, simple_loss=0.3755, pruned_loss=0.1608, over 7363.00 frames. ], tot_loss[loss=0.365, simple_loss=0.3837, pruned_loss=0.1822, over 1440950.04 frames. ], batch size: 63, lr: 4.79e-02, grad_scale: 16.0 +2023-03-20 17:35:39,792 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-20 17:35:45,429 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=2194.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:35:49,118 INFO [train.py:901] (0/2) Epoch 1, batch 2200, loss[loss=0.3851, simple_loss=0.3913, pruned_loss=0.1894, over 7291.00 frames. ], tot_loss[loss=0.3648, simple_loss=0.3831, pruned_loss=0.1803, over 1437980.79 frames. ], batch size: 86, lr: 4.78e-02, grad_scale: 16.0 +2023-03-20 17:35:49,560 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.340e+02 3.853e+02 4.847e+02 6.366e+02 1.558e+03, threshold=9.694e+02, percent-clipped=3.0 +2023-03-20 17:35:52,393 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-20 17:35:55,367 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=2213.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:35:59,518 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=2222.0, num_to_drop=2, layers_to_drop={0, 2} +2023-03-20 17:36:01,330 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=2226.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:36:13,514 INFO [train.py:901] (0/2) Epoch 1, batch 2250, loss[loss=0.3542, simple_loss=0.3778, pruned_loss=0.1653, over 7312.00 frames. ], tot_loss[loss=0.3608, simple_loss=0.3798, pruned_loss=0.1764, over 1439548.70 frames. ], batch size: 80, lr: 4.77e-02, grad_scale: 16.0 +2023-03-20 17:36:21,035 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=2267.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:36:22,415 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=2270.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:36:23,854 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4404, 3.8852, 3.3332, 3.9729, 3.6538, 4.6438, 4.4890, 4.3514], + device='cuda:0'), covar=tensor([0.0092, 0.0376, 0.1163, 0.0251, 0.0979, 0.0081, 0.0147, 0.0236], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0030, 0.0039, 0.0024, 0.0039, 0.0025, 0.0032, 0.0024], + device='cuda:0'), out_proj_covar=tensor([1.3787e-05, 2.2855e-05, 3.1989e-05, 1.7681e-05, 3.3792e-05, 1.7144e-05, + 2.4822e-05, 1.7054e-05], device='cuda:0') +2023-03-20 17:36:25,098 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-20 17:36:25,562 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-20 17:36:36,314 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-20 17:36:36,771 INFO [train.py:901] (0/2) Epoch 1, batch 2300, loss[loss=0.3409, simple_loss=0.363, pruned_loss=0.1594, over 7247.00 frames. ], tot_loss[loss=0.3591, simple_loss=0.3783, pruned_loss=0.1742, over 1441303.30 frames. ], batch size: 47, lr: 4.77e-02, grad_scale: 16.0 +2023-03-20 17:36:37,857 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.965e+02 3.824e+02 4.896e+02 6.606e+02 1.192e+03, threshold=9.792e+02, percent-clipped=5.0 +2023-03-20 17:36:44,034 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=2315.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:36:50,176 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.17 vs. limit=5.0 +2023-03-20 17:36:56,561 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=2340.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:37:01,661 INFO [train.py:901] (0/2) Epoch 1, batch 2350, loss[loss=0.3551, simple_loss=0.3776, pruned_loss=0.1663, over 7329.00 frames. ], tot_loss[loss=0.3557, simple_loss=0.3755, pruned_loss=0.1713, over 1438929.75 frames. ], batch size: 61, lr: 4.76e-02, grad_scale: 16.0 +2023-03-20 17:37:16,844 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.00 vs. limit=2.0 +2023-03-20 17:37:19,481 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-20 17:37:25,982 INFO [train.py:901] (0/2) Epoch 1, batch 2400, loss[loss=0.3449, simple_loss=0.3646, pruned_loss=0.1626, over 7247.00 frames. ], tot_loss[loss=0.3528, simple_loss=0.3731, pruned_loss=0.1688, over 1439928.89 frames. ], batch size: 55, lr: 4.75e-02, grad_scale: 16.0 +2023-03-20 17:37:25,994 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-20 17:37:26,133 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=2401.0, num_to_drop=2, layers_to_drop={0, 3} +2023-03-20 17:37:26,428 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.851e+02 3.554e+02 4.568e+02 5.565e+02 8.265e+02, threshold=9.137e+02, percent-clipped=0.0 +2023-03-20 17:37:36,025 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-20 17:37:38,775 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-20 17:37:49,956 INFO [train.py:901] (0/2) Epoch 1, batch 2450, loss[loss=0.3466, simple_loss=0.3734, pruned_loss=0.1599, over 7336.00 frames. ], tot_loss[loss=0.3496, simple_loss=0.3703, pruned_loss=0.1665, over 1440137.68 frames. ], batch size: 54, lr: 4.74e-02, grad_scale: 16.0 +2023-03-20 17:37:50,551 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9677, 3.4735, 3.2764, 3.4836, 3.6789, 3.7822, 3.2154, 3.3696], + device='cuda:0'), covar=tensor([0.0609, 0.0685, 0.1450, 0.0720, 0.0371, 0.0478, 0.2182, 0.1014], + device='cuda:0'), in_proj_covar=tensor([0.0013, 0.0016, 0.0021, 0.0018, 0.0014, 0.0015, 0.0025, 0.0018], + device='cuda:0'), out_proj_covar=tensor([7.5336e-06, 1.0468e-05, 1.6583e-05, 1.2516e-05, 8.0203e-06, 9.5057e-06, + 2.0393e-05, 1.2391e-05], device='cuda:0') +2023-03-20 17:38:04,296 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-20 17:38:08,953 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.95 vs. limit=2.0 +2023-03-20 17:38:08,958 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.40 vs. limit=5.0 +2023-03-20 17:38:11,562 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=2494.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:38:14,768 INFO [train.py:901] (0/2) Epoch 1, batch 2500, loss[loss=0.3196, simple_loss=0.3469, pruned_loss=0.1461, over 7278.00 frames. ], tot_loss[loss=0.347, simple_loss=0.3685, pruned_loss=0.1643, over 1439913.61 frames. ], batch size: 70, lr: 4.73e-02, grad_scale: 16.0 +2023-03-20 17:38:15,212 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.011e+02 3.688e+02 4.749e+02 6.089e+02 1.180e+03, threshold=9.499e+02, percent-clipped=5.0 +2023-03-20 17:38:20,416 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=2513.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:38:26,511 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=2526.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:38:27,835 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-20 17:38:33,882 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-03-20 17:38:34,028 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=2542.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:38:38,171 INFO [train.py:901] (0/2) Epoch 1, batch 2550, loss[loss=0.3295, simple_loss=0.3621, pruned_loss=0.1485, over 7280.00 frames. ], tot_loss[loss=0.3449, simple_loss=0.3673, pruned_loss=0.1625, over 1439772.28 frames. ], batch size: 77, lr: 4.72e-02, grad_scale: 16.0 +2023-03-20 17:38:43,518 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=2561.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:38:44,544 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6891, 2.9645, 2.7537, 2.3578, 2.9195, 2.5525, 2.1246, 2.8242], + device='cuda:0'), covar=tensor([0.0352, 0.0306, 0.0416, 0.0801, 0.0287, 0.0401, 0.0798, 0.0168], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0032, 0.0036, 0.0038, 0.0031, 0.0036, 0.0039, 0.0026], + device='cuda:0'), out_proj_covar=tensor([2.8902e-05, 2.4375e-05, 2.7773e-05, 3.0855e-05, 2.3575e-05, 2.8013e-05, + 3.1943e-05, 1.9837e-05], device='cuda:0') +2023-03-20 17:38:49,572 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=2574.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:38:57,952 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1599, 4.1231, 4.0764, 4.4968, 4.6414, 4.4759, 3.9777, 4.1721], + device='cuda:0'), covar=tensor([0.0561, 0.0773, 0.1022, 0.0828, 0.0441, 0.0498, 0.0632, 0.0618], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0073, 0.0089, 0.0059, 0.0069, 0.0064, 0.0055, 0.0066], + device='cuda:0'), out_proj_covar=tensor([4.9999e-05, 7.2226e-05, 9.5225e-05, 6.0313e-05, 6.8928e-05, 5.9353e-05, + 5.2435e-05, 6.3910e-05], device='cuda:0') +2023-03-20 17:39:03,553 INFO [train.py:901] (0/2) Epoch 1, batch 2600, loss[loss=0.333, simple_loss=0.3592, pruned_loss=0.1534, over 7232.00 frames. ], tot_loss[loss=0.3421, simple_loss=0.3649, pruned_loss=0.1606, over 1441457.06 frames. ], batch size: 55, lr: 4.71e-02, grad_scale: 16.0 +2023-03-20 17:39:04,020 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.100e+02 3.577e+02 4.452e+02 5.636e+02 1.247e+03, threshold=8.904e+02, percent-clipped=2.0 +2023-03-20 17:39:19,225 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0647, 3.3605, 3.2304, 3.1047, 3.1140, 2.7432, 2.7089, 3.2645], + device='cuda:0'), covar=tensor([0.0307, 0.0170, 0.0239, 0.0285, 0.0355, 0.0387, 0.0509, 0.0116], + device='cuda:0'), in_proj_covar=tensor([0.0034, 0.0031, 0.0032, 0.0032, 0.0029, 0.0032, 0.0035, 0.0025], + device='cuda:0'), out_proj_covar=tensor([2.7488e-05, 2.2730e-05, 2.4516e-05, 2.6597e-05, 2.1875e-05, 2.5517e-05, + 2.8099e-05, 1.8227e-05], device='cuda:0') +2023-03-20 17:39:26,895 INFO [train.py:901] (0/2) Epoch 1, batch 2650, loss[loss=0.3521, simple_loss=0.3656, pruned_loss=0.1693, over 7288.00 frames. ], tot_loss[loss=0.339, simple_loss=0.3627, pruned_loss=0.1584, over 1442461.65 frames. ], batch size: 66, lr: 4.70e-02, grad_scale: 16.0 +2023-03-20 17:39:35,135 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6734, 3.8968, 3.7097, 3.5706, 3.5654, 2.9414, 3.4071, 3.9907], + device='cuda:0'), covar=tensor([0.0136, 0.0096, 0.0200, 0.0150, 0.0261, 0.0554, 0.0325, 0.0066], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0020, 0.0022, 0.0024, 0.0025, 0.0023, 0.0023, 0.0021], + device='cuda:0'), out_proj_covar=tensor([1.2426e-05, 1.3882e-05, 1.5104e-05, 1.6569e-05, 1.9618e-05, 1.7556e-05, + 1.5012e-05, 1.2859e-05], device='cuda:0') +2023-03-20 17:39:43,050 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=2685.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:39:48,166 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=2696.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:39:50,388 INFO [train.py:901] (0/2) Epoch 1, batch 2700, loss[loss=0.3545, simple_loss=0.3634, pruned_loss=0.1727, over 7290.00 frames. ], tot_loss[loss=0.3367, simple_loss=0.3606, pruned_loss=0.157, over 1440954.71 frames. ], batch size: 47, lr: 4.69e-02, grad_scale: 16.0 +2023-03-20 17:39:50,813 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.158e+02 3.938e+02 4.931e+02 6.424e+02 1.116e+03, threshold=9.861e+02, percent-clipped=3.0 +2023-03-20 17:39:56,178 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.01 vs. limit=2.0 +2023-03-20 17:40:04,203 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.94 vs. limit=2.0 +2023-03-20 17:40:08,576 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.26 vs. limit=2.0 +2023-03-20 17:40:11,657 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=2746.0, num_to_drop=2, layers_to_drop={2, 3} +2023-03-20 17:40:13,953 INFO [train.py:901] (0/2) Epoch 1, batch 2750, loss[loss=0.3098, simple_loss=0.3431, pruned_loss=0.1382, over 7283.00 frames. ], tot_loss[loss=0.3354, simple_loss=0.36, pruned_loss=0.1558, over 1443013.91 frames. ], batch size: 66, lr: 4.68e-02, grad_scale: 16.0 +2023-03-20 17:40:20,781 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.06 vs. limit=2.0 +2023-03-20 17:40:32,842 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.98 vs. limit=5.0 +2023-03-20 17:40:37,983 INFO [train.py:901] (0/2) Epoch 1, batch 2800, loss[loss=0.3119, simple_loss=0.342, pruned_loss=0.1409, over 7288.00 frames. ], tot_loss[loss=0.3341, simple_loss=0.3586, pruned_loss=0.1551, over 1441818.82 frames. ], batch size: 68, lr: 4.67e-02, grad_scale: 16.0 +2023-03-20 17:40:38,405 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.290e+02 4.051e+02 5.024e+02 6.754e+02 1.289e+03, threshold=1.005e+03, percent-clipped=6.0 +2023-03-20 17:40:45,184 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.10 vs. limit=2.0 +2023-03-20 17:40:50,013 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-1.pt +2023-03-20 17:41:07,605 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-20 17:41:07,661 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-20 17:41:07,686 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-20 17:41:07,703 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-20 17:41:07,877 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-20 17:41:07,893 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-20 17:41:07,904 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-20 17:41:08,002 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-20 17:41:08,078 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-20 17:41:08,092 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-20 17:41:08,110 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-20 17:41:08,621 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-20 17:41:08,626 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-20 17:41:08,634 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-20 17:41:08,773 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-20 17:41:08,775 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-20 17:41:08,834 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-20 17:41:10,776 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=2824.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:41:11,120 INFO [train.py:901] (0/2) Epoch 2, batch 0, loss[loss=0.3633, simple_loss=0.3769, pruned_loss=0.1748, over 7289.00 frames. ], tot_loss[loss=0.3633, simple_loss=0.3769, pruned_loss=0.1748, over 7289.00 frames. ], batch size: 86, lr: 4.63e-02, grad_scale: 16.0 +2023-03-20 17:41:11,121 INFO [train.py:926] (0/2) Computing validation loss +2023-03-20 17:41:36,989 INFO [train.py:935] (0/2) Epoch 2, validation: loss=0.3661, simple_loss=0.5483, pruned_loss=0.09198, over 1622729.00 frames. +2023-03-20 17:41:36,990 INFO [train.py:936] (0/2) Maximum memory allocated so far is 11706MB +2023-03-20 17:41:43,597 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-20 17:41:49,980 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-03-20 17:41:52,963 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-20 17:41:59,965 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-20 17:42:00,428 INFO [train.py:901] (0/2) Epoch 2, batch 50, loss[loss=0.3214, simple_loss=0.3459, pruned_loss=0.1484, over 7265.00 frames. ], tot_loss[loss=0.3285, simple_loss=0.3549, pruned_loss=0.1511, over 324766.69 frames. ], batch size: 47, lr: 4.62e-02, grad_scale: 16.0 +2023-03-20 17:42:02,342 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-20 17:42:04,746 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-20 17:42:05,401 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=2885.0, num_to_drop=2, layers_to_drop={0, 1} +2023-03-20 17:42:13,254 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.499e+02 3.706e+02 4.761e+02 5.953e+02 9.825e+02, threshold=9.522e+02, percent-clipped=0.0 +2023-03-20 17:42:22,613 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-20 17:42:23,088 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-20 17:42:24,673 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3144, 4.2215, 3.6912, 3.5845, 3.1950, 4.0214, 4.3710, 4.0769], + device='cuda:0'), covar=tensor([0.0765, 0.0183, 0.1179, 0.0192, 0.0290, 0.0301, 0.0063, 0.0148], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0032, 0.0049, 0.0021, 0.0022, 0.0028, 0.0022, 0.0024], + device='cuda:0'), out_proj_covar=tensor([3.0874e-05, 2.2346e-05, 4.4056e-05, 1.3117e-05, 1.3165e-05, 1.9936e-05, + 1.1429e-05, 1.3574e-05], device='cuda:0') +2023-03-20 17:42:25,471 INFO [train.py:901] (0/2) Epoch 2, batch 100, loss[loss=0.3417, simple_loss=0.3624, pruned_loss=0.1605, over 7333.00 frames. ], tot_loss[loss=0.324, simple_loss=0.3506, pruned_loss=0.1488, over 572686.99 frames. ], batch size: 49, lr: 4.61e-02, grad_scale: 16.0 +2023-03-20 17:42:31,310 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=2937.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:42:42,649 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2877, 2.2791, 3.8380, 3.4217, 2.7554, 3.2356, 2.9314, 3.5777], + device='cuda:0'), covar=tensor([0.0208, 0.0951, 0.0075, 0.0199, 0.0202, 0.0215, 0.0824, 0.0175], + device='cuda:0'), in_proj_covar=tensor([0.0012, 0.0014, 0.0012, 0.0012, 0.0013, 0.0014, 0.0020, 0.0013], + device='cuda:0'), out_proj_covar=tensor([1.1449e-05, 1.4624e-05, 1.1626e-05, 1.1406e-05, 1.0902e-05, 1.2779e-05, + 2.1525e-05, 1.2566e-05], device='cuda:0') +2023-03-20 17:42:49,382 INFO [train.py:901] (0/2) Epoch 2, batch 150, loss[loss=0.3166, simple_loss=0.3505, pruned_loss=0.1414, over 7327.00 frames. ], tot_loss[loss=0.3238, simple_loss=0.3513, pruned_loss=0.1481, over 767177.97 frames. ], batch size: 83, lr: 4.60e-02, grad_scale: 16.0 +2023-03-20 17:42:59,667 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=2996.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 17:43:01,294 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=2998.0, num_to_drop=2, layers_to_drop={1, 2} +2023-03-20 17:43:03,379 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.879e+02 3.800e+02 4.704e+02 5.909e+02 1.183e+03, threshold=9.409e+02, percent-clipped=3.0 +2023-03-20 17:43:04,472 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7040, 4.4477, 4.5565, 4.0626, 4.0262, 4.1224, 4.7810, 4.6238], + device='cuda:0'), covar=tensor([0.0489, 0.0196, 0.0120, 0.0282, 0.0385, 0.0323, 0.0148, 0.0193], + device='cuda:0'), in_proj_covar=tensor([0.0060, 0.0055, 0.0049, 0.0068, 0.0054, 0.0051, 0.0044, 0.0051], + device='cuda:0'), out_proj_covar=tensor([6.0762e-05, 5.3529e-05, 5.1277e-05, 7.2159e-05, 5.6333e-05, 5.4596e-05, + 4.4862e-05, 5.2935e-05], device='cuda:0') +2023-03-20 17:43:11,842 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=3019.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:43:12,907 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.84 vs. limit=5.0 +2023-03-20 17:43:14,528 INFO [train.py:901] (0/2) Epoch 2, batch 200, loss[loss=0.3013, simple_loss=0.3381, pruned_loss=0.1322, over 7346.00 frames. ], tot_loss[loss=0.3224, simple_loss=0.3508, pruned_loss=0.147, over 916829.71 frames. ], batch size: 73, lr: 4.59e-02, grad_scale: 16.0 +2023-03-20 17:43:19,436 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-20 17:43:22,462 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=3041.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:43:23,880 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-20 17:43:23,923 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=3044.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:43:29,228 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-20 17:43:38,975 INFO [train.py:901] (0/2) Epoch 2, batch 250, loss[loss=0.3158, simple_loss=0.3528, pruned_loss=0.1394, over 7215.00 frames. ], tot_loss[loss=0.3187, simple_loss=0.3481, pruned_loss=0.1446, over 1032889.04 frames. ], batch size: 93, lr: 4.58e-02, grad_scale: 16.0 +2023-03-20 17:43:40,573 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6019, 2.6855, 2.7130, 2.5129, 1.7243, 2.7296, 2.8232, 2.6465], + device='cuda:0'), covar=tensor([0.0476, 0.0425, 0.0331, 0.0442, 0.1364, 0.0399, 0.0236, 0.0355], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0036, 0.0036, 0.0039, 0.0041, 0.0035, 0.0034, 0.0034], + device='cuda:0'), out_proj_covar=tensor([3.4903e-05, 3.0241e-05, 3.5717e-05, 3.4100e-05, 3.4758e-05, 2.8722e-05, + 2.6538e-05, 2.6464e-05], device='cuda:0') +2023-03-20 17:43:41,976 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-20 17:43:42,101 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=3080.0, num_to_drop=2, layers_to_drop={0, 1} +2023-03-20 17:43:52,949 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.532e+02 4.005e+02 4.790e+02 6.030e+02 1.284e+03, threshold=9.580e+02, percent-clipped=2.0 +2023-03-20 17:43:53,798 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.16 vs. limit=2.0 +2023-03-20 17:43:59,907 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8379, 4.1456, 4.1836, 4.3683, 4.6450, 4.4843, 4.0684, 3.8809], + device='cuda:0'), covar=tensor([0.0867, 0.0989, 0.1272, 0.1119, 0.0562, 0.0963, 0.0649, 0.0740], + device='cuda:0'), in_proj_covar=tensor([0.0055, 0.0086, 0.0105, 0.0070, 0.0078, 0.0076, 0.0060, 0.0075], + device='cuda:0'), out_proj_covar=tensor([5.7528e-05, 8.8488e-05, 1.1535e-04, 7.6998e-05, 7.9113e-05, 7.6136e-05, + 5.9121e-05, 7.2549e-05], device='cuda:0') +2023-03-20 17:44:01,003 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.05 vs. limit=2.0 +2023-03-20 17:44:01,234 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-20 17:44:04,155 INFO [train.py:901] (0/2) Epoch 2, batch 300, loss[loss=0.2673, simple_loss=0.2996, pruned_loss=0.1175, over 7163.00 frames. ], tot_loss[loss=0.3197, simple_loss=0.3488, pruned_loss=0.1453, over 1122284.12 frames. ], batch size: 39, lr: 4.57e-02, grad_scale: 16.0 +2023-03-20 17:44:07,169 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5636, 3.1673, 2.6902, 3.2076, 2.8431, 1.9232, 2.5211, 2.7611], + device='cuda:0'), covar=tensor([0.0768, 0.0250, 0.0449, 0.0177, 0.0337, 0.1348, 0.0711, 0.0315], + device='cuda:0'), in_proj_covar=tensor([0.0022, 0.0022, 0.0023, 0.0022, 0.0025, 0.0024, 0.0027, 0.0022], + device='cuda:0'), out_proj_covar=tensor([1.6555e-05, 1.5034e-05, 1.6180e-05, 1.7007e-05, 2.1256e-05, 2.0030e-05, + 2.0255e-05, 1.5442e-05], device='cuda:0') +2023-03-20 17:44:10,548 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-20 17:44:26,314 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3955, 3.6329, 3.3663, 3.4428, 2.7359, 4.0460, 4.2127, 3.8923], + device='cuda:0'), covar=tensor([0.1781, 0.0267, 0.1825, 0.0287, 0.0442, 0.0239, 0.0103, 0.0208], + device='cuda:0'), in_proj_covar=tensor([0.0052, 0.0038, 0.0062, 0.0026, 0.0026, 0.0033, 0.0026, 0.0027], + device='cuda:0'), out_proj_covar=tensor([3.9456e-05, 2.6033e-05, 5.3726e-05, 1.5735e-05, 1.6228e-05, 2.3344e-05, + 1.3550e-05, 1.6298e-05], device='cuda:0') +2023-03-20 17:44:29,661 INFO [train.py:901] (0/2) Epoch 2, batch 350, loss[loss=0.3225, simple_loss=0.3455, pruned_loss=0.1498, over 7263.00 frames. ], tot_loss[loss=0.319, simple_loss=0.3481, pruned_loss=0.1449, over 1192630.91 frames. ], batch size: 64, lr: 4.56e-02, grad_scale: 16.0 +2023-03-20 17:44:32,086 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=3180.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:44:42,793 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.966e+02 3.735e+02 4.987e+02 6.061e+02 9.476e+02, threshold=9.975e+02, percent-clipped=1.0 +2023-03-20 17:44:45,143 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-20 17:44:53,759 INFO [train.py:901] (0/2) Epoch 2, batch 400, loss[loss=0.3244, simple_loss=0.3514, pruned_loss=0.1488, over 7285.00 frames. ], tot_loss[loss=0.3176, simple_loss=0.3468, pruned_loss=0.1442, over 1246517.16 frames. ], batch size: 77, lr: 4.55e-02, grad_scale: 16.0 +2023-03-20 17:44:54,593 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-03-20 17:45:11,982 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9898, 2.8755, 2.1179, 2.8516, 2.7000, 3.0764, 2.9505, 2.7213], + device='cuda:0'), covar=tensor([0.3058, 0.0781, 0.4316, 0.0551, 0.0502, 0.0576, 0.0617, 0.0729], + device='cuda:0'), in_proj_covar=tensor([0.0056, 0.0039, 0.0066, 0.0026, 0.0026, 0.0034, 0.0029, 0.0030], + device='cuda:0'), out_proj_covar=tensor([4.2141e-05, 2.6588e-05, 5.6000e-05, 1.6374e-05, 1.5951e-05, 2.3373e-05, + 1.6018e-05, 1.7962e-05], device='cuda:0') +2023-03-20 17:45:18,910 INFO [train.py:901] (0/2) Epoch 2, batch 450, loss[loss=0.3506, simple_loss=0.3728, pruned_loss=0.1642, over 7355.00 frames. ], tot_loss[loss=0.3175, simple_loss=0.3469, pruned_loss=0.1441, over 1290865.24 frames. ], batch size: 73, lr: 4.54e-02, grad_scale: 16.0 +2023-03-20 17:45:25,252 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=3288.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:45:26,079 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-20 17:45:26,395 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.01 vs. limit=2.0 +2023-03-20 17:45:26,546 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-20 17:45:27,611 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=3293.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:45:31,876 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.013e+02 3.842e+02 4.912e+02 6.234e+02 9.651e+02, threshold=9.824e+02, percent-clipped=0.0 +2023-03-20 17:45:43,020 INFO [train.py:901] (0/2) Epoch 2, batch 500, loss[loss=0.3288, simple_loss=0.3581, pruned_loss=0.1497, over 7282.00 frames. ], tot_loss[loss=0.3178, simple_loss=0.3473, pruned_loss=0.1442, over 1325119.07 frames. ], batch size: 52, lr: 4.53e-02, grad_scale: 16.0 +2023-03-20 17:45:46,712 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.00 vs. limit=2.0 +2023-03-20 17:45:50,956 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=3341.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:45:54,873 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=3349.0, num_to_drop=2, layers_to_drop={2, 3} +2023-03-20 17:45:59,367 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-20 17:46:00,383 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-20 17:46:00,855 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-20 17:46:02,381 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3951, 3.9415, 4.0072, 3.7075, 3.3194, 3.8723, 4.2503, 4.2250], + device='cuda:0'), covar=tensor([0.0452, 0.0248, 0.0263, 0.0317, 0.0576, 0.0332, 0.0168, 0.0228], + device='cuda:0'), in_proj_covar=tensor([0.0063, 0.0060, 0.0055, 0.0074, 0.0060, 0.0052, 0.0045, 0.0055], + device='cuda:0'), out_proj_covar=tensor([6.5913e-05, 6.2321e-05, 6.0738e-05, 8.4850e-05, 6.7700e-05, 5.9126e-05, + 4.7821e-05, 6.0702e-05], device='cuda:0') +2023-03-20 17:46:02,777 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-20 17:46:07,523 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-20 17:46:08,475 INFO [train.py:901] (0/2) Epoch 2, batch 550, loss[loss=0.3285, simple_loss=0.3561, pruned_loss=0.1505, over 7342.00 frames. ], tot_loss[loss=0.3153, simple_loss=0.3453, pruned_loss=0.1427, over 1353386.49 frames. ], batch size: 75, lr: 4.52e-02, grad_scale: 16.0 +2023-03-20 17:46:08,565 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=3375.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:46:11,510 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5793, 3.4714, 3.2517, 3.2766, 3.3443, 3.1378, 3.2669, 3.1360], + device='cuda:0'), covar=tensor([0.0069, 0.0132, 0.0149, 0.0271, 0.0127, 0.0196, 0.0186, 0.0237], + device='cuda:0'), in_proj_covar=tensor([0.0012, 0.0014, 0.0015, 0.0016, 0.0013, 0.0016, 0.0017, 0.0016], + device='cuda:0'), out_proj_covar=tensor([1.1512e-05, 1.6215e-05, 1.4635e-05, 1.9649e-05, 1.4135e-05, 1.7042e-05, + 1.8298e-05, 1.8925e-05], device='cuda:0') +2023-03-20 17:46:15,464 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=3389.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:46:17,840 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-20 17:46:22,176 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.129e+02 3.792e+02 4.639e+02 5.385e+02 1.138e+03, threshold=9.277e+02, percent-clipped=1.0 +2023-03-20 17:46:25,115 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-20 17:46:28,098 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-20 17:46:31,102 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9318, 2.9928, 2.6933, 3.2246, 1.9107, 2.2536, 3.0094, 2.6165], + device='cuda:0'), covar=tensor([0.0430, 0.0214, 0.0379, 0.0141, 0.1152, 0.1059, 0.0210, 0.0781], + device='cuda:0'), in_proj_covar=tensor([0.0024, 0.0023, 0.0026, 0.0024, 0.0027, 0.0025, 0.0025, 0.0022], + device='cuda:0'), out_proj_covar=tensor([2.2404e-05, 1.8710e-05, 2.1498e-05, 1.9788e-05, 2.6237e-05, 2.5886e-05, + 2.0546e-05, 2.1838e-05], device='cuda:0') +2023-03-20 17:46:33,412 INFO [train.py:901] (0/2) Epoch 2, batch 600, loss[loss=0.3404, simple_loss=0.3707, pruned_loss=0.155, over 7214.00 frames. ], tot_loss[loss=0.3154, simple_loss=0.3457, pruned_loss=0.1426, over 1374450.18 frames. ], batch size: 99, lr: 4.51e-02, grad_scale: 8.0 +2023-03-20 17:46:34,891 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-20 17:46:51,291 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-20 17:46:59,230 INFO [train.py:901] (0/2) Epoch 2, batch 650, loss[loss=0.3166, simple_loss=0.34, pruned_loss=0.1466, over 7166.00 frames. ], tot_loss[loss=0.3137, simple_loss=0.3442, pruned_loss=0.1415, over 1387049.80 frames. ], batch size: 39, lr: 4.50e-02, grad_scale: 8.0 +2023-03-20 17:47:01,115 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-20 17:47:01,686 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=3480.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:47:10,107 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.07 vs. limit=2.0 +2023-03-20 17:47:12,720 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.910e+02 3.942e+02 5.016e+02 7.183e+02 2.296e+03, threshold=1.003e+03, percent-clipped=6.0 +2023-03-20 17:47:17,261 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-20 17:47:19,348 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3581, 4.5029, 4.1904, 4.7166, 4.3197, 4.1527, 4.6755, 4.7155], + device='cuda:0'), covar=tensor([0.0405, 0.0264, 0.0430, 0.0369, 0.0385, 0.0411, 0.0333, 0.0241], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0043, 0.0047, 0.0044, 0.0042, 0.0049, 0.0042, 0.0046], + device='cuda:0'), out_proj_covar=tensor([5.2937e-05, 4.3691e-05, 5.0213e-05, 4.8128e-05, 4.4301e-05, 5.4260e-05, + 4.4249e-05, 5.5406e-05], device='cuda:0') +2023-03-20 17:47:24,907 INFO [train.py:901] (0/2) Epoch 2, batch 700, loss[loss=0.2924, simple_loss=0.3271, pruned_loss=0.1288, over 7308.00 frames. ], tot_loss[loss=0.3114, simple_loss=0.3428, pruned_loss=0.14, over 1400350.91 frames. ], batch size: 49, lr: 4.49e-02, grad_scale: 8.0 +2023-03-20 17:47:26,029 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8181, 4.0884, 3.8567, 3.7688, 3.8283, 3.5392, 4.0405, 3.7859], + device='cuda:0'), covar=tensor([0.0064, 0.0104, 0.0110, 0.0169, 0.0124, 0.0182, 0.0107, 0.0131], + device='cuda:0'), in_proj_covar=tensor([0.0012, 0.0015, 0.0015, 0.0017, 0.0014, 0.0017, 0.0018, 0.0016], + device='cuda:0'), out_proj_covar=tensor([1.2101e-05, 1.7652e-05, 1.5772e-05, 2.0905e-05, 1.5183e-05, 1.9395e-05, + 2.0414e-05, 1.9559e-05], device='cuda:0') +2023-03-20 17:47:26,460 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=3528.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:47:26,916 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-20 17:47:35,255 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=3546.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:47:48,342 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.42 vs. limit=2.0 +2023-03-20 17:47:49,467 INFO [train.py:901] (0/2) Epoch 2, batch 750, loss[loss=0.2794, simple_loss=0.3153, pruned_loss=0.1217, over 7271.00 frames. ], tot_loss[loss=0.3114, simple_loss=0.3425, pruned_loss=0.1402, over 1408871.11 frames. ], batch size: 64, lr: 4.48e-02, grad_scale: 8.0 +2023-03-20 17:47:50,007 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-20 17:47:50,948 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-20 17:47:58,451 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=3593.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:48:03,621 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.081e+02 4.174e+02 5.233e+02 6.266e+02 1.011e+03, threshold=1.047e+03, percent-clipped=1.0 +2023-03-20 17:48:04,658 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-20 17:48:06,395 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6849, 3.2803, 2.9900, 2.8772, 3.1768, 2.5048, 3.2681, 3.0096], + device='cuda:0'), covar=tensor([0.0440, 0.0069, 0.0296, 0.0427, 0.0227, 0.0629, 0.0121, 0.0275], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0022, 0.0025, 0.0028, 0.0026, 0.0025, 0.0022, 0.0029], + device='cuda:0'), out_proj_covar=tensor([2.3552e-05, 1.6947e-05, 2.0899e-05, 2.6895e-05, 2.3623e-05, 2.4222e-05, + 1.7713e-05, 2.9185e-05], device='cuda:0') +2023-03-20 17:48:06,429 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=3607.0, num_to_drop=2, layers_to_drop={2, 3} +2023-03-20 17:48:10,803 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-20 17:48:15,790 INFO [train.py:901] (0/2) Epoch 2, batch 800, loss[loss=0.334, simple_loss=0.3623, pruned_loss=0.1528, over 7293.00 frames. ], tot_loss[loss=0.3112, simple_loss=0.3428, pruned_loss=0.1398, over 1417543.01 frames. ], batch size: 86, lr: 4.47e-02, grad_scale: 8.0 +2023-03-20 17:48:15,810 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-20 17:48:17,346 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-20 17:48:23,802 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=3641.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:48:25,316 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=3644.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:48:28,188 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-20 17:48:40,931 INFO [train.py:901] (0/2) Epoch 2, batch 850, loss[loss=0.31, simple_loss=0.3523, pruned_loss=0.1338, over 7259.00 frames. ], tot_loss[loss=0.3114, simple_loss=0.3428, pruned_loss=0.14, over 1421785.48 frames. ], batch size: 55, lr: 4.46e-02, grad_scale: 8.0 +2023-03-20 17:48:41,070 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=3675.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:48:41,578 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=3676.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:48:47,020 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-20 17:48:47,029 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-20 17:48:50,865 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-20 17:48:53,100 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-20 17:48:55,581 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.17 vs. limit=2.0 +2023-03-20 17:48:56,250 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.669e+02 3.969e+02 4.736e+02 6.049e+02 1.256e+03, threshold=9.472e+02, percent-clipped=4.0 +2023-03-20 17:48:57,286 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-20 17:49:06,508 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=3723.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:49:07,476 INFO [train.py:901] (0/2) Epoch 2, batch 900, loss[loss=0.2931, simple_loss=0.3418, pruned_loss=0.1223, over 7299.00 frames. ], tot_loss[loss=0.3081, simple_loss=0.3406, pruned_loss=0.1378, over 1427392.62 frames. ], batch size: 68, lr: 4.45e-02, grad_scale: 8.0 +2023-03-20 17:49:13,774 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=3737.0, num_to_drop=2, layers_to_drop={2, 3} +2023-03-20 17:49:20,125 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.09 vs. limit=2.0 +2023-03-20 17:49:32,678 INFO [train.py:901] (0/2) Epoch 2, batch 950, loss[loss=0.292, simple_loss=0.3263, pruned_loss=0.1288, over 7326.00 frames. ], tot_loss[loss=0.3082, simple_loss=0.3407, pruned_loss=0.1379, over 1430966.10 frames. ], batch size: 49, lr: 4.44e-02, grad_scale: 8.0 +2023-03-20 17:49:34,196 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-20 17:49:45,771 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.15 vs. limit=2.0 +2023-03-20 17:49:48,231 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.101e+02 3.953e+02 4.768e+02 6.187e+02 9.797e+02, threshold=9.537e+02, percent-clipped=2.0 +2023-03-20 17:49:54,343 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3597, 4.6729, 4.4330, 4.7990, 4.4224, 4.3546, 4.7692, 4.7191], + device='cuda:0'), covar=tensor([0.0347, 0.0189, 0.0290, 0.0323, 0.0318, 0.0276, 0.0235, 0.0210], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0041, 0.0045, 0.0045, 0.0043, 0.0047, 0.0042, 0.0044], + device='cuda:0'), out_proj_covar=tensor([5.4051e-05, 4.1822e-05, 4.9450e-05, 5.1527e-05, 4.6827e-05, 5.3211e-05, + 4.6695e-05, 5.2193e-05], device='cuda:0') +2023-03-20 17:49:59,319 INFO [train.py:901] (0/2) Epoch 2, batch 1000, loss[loss=0.282, simple_loss=0.3205, pruned_loss=0.1218, over 7135.00 frames. ], tot_loss[loss=0.307, simple_loss=0.3394, pruned_loss=0.1373, over 1430234.42 frames. ], batch size: 41, lr: 4.43e-02, grad_scale: 8.0 +2023-03-20 17:49:59,338 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-20 17:50:19,357 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2952, 3.5506, 3.6361, 3.3234, 3.3004, 3.5552, 3.6547, 3.6773], + device='cuda:0'), covar=tensor([0.0500, 0.0350, 0.0347, 0.0471, 0.0432, 0.0470, 0.0589, 0.0581], + device='cuda:0'), in_proj_covar=tensor([0.0061, 0.0061, 0.0056, 0.0074, 0.0061, 0.0052, 0.0050, 0.0055], + device='cuda:0'), out_proj_covar=tensor([7.0852e-05, 7.0247e-05, 6.5602e-05, 9.4053e-05, 7.2009e-05, 6.2044e-05, + 5.6037e-05, 6.3441e-05], device='cuda:0') +2023-03-20 17:50:19,948 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5471, 3.5930, 3.1692, 2.8007, 3.2875, 3.3108, 3.2349, 3.5074], + device='cuda:0'), covar=tensor([0.1643, 0.0351, 0.1749, 0.0467, 0.0372, 0.0232, 0.0354, 0.0161], + device='cuda:0'), in_proj_covar=tensor([0.0075, 0.0051, 0.0080, 0.0035, 0.0036, 0.0044, 0.0037, 0.0037], + device='cuda:0'), out_proj_covar=tensor([5.4971e-05, 3.3598e-05, 6.6877e-05, 2.2217e-05, 2.2065e-05, 2.9155e-05, + 2.2153e-05, 2.1587e-05], device='cuda:0') +2023-03-20 17:50:20,330 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-20 17:50:25,591 INFO [train.py:901] (0/2) Epoch 2, batch 1050, loss[loss=0.2839, simple_loss=0.3204, pruned_loss=0.1237, over 7232.00 frames. ], tot_loss[loss=0.3058, simple_loss=0.3387, pruned_loss=0.1365, over 1433939.91 frames. ], batch size: 45, lr: 4.41e-02, grad_scale: 8.0 +2023-03-20 17:50:27,083 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.96 vs. limit=2.0 +2023-03-20 17:50:40,161 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=3902.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 17:50:40,540 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.085e+02 3.872e+02 4.717e+02 6.455e+02 1.344e+03, threshold=9.435e+02, percent-clipped=6.0 +2023-03-20 17:50:43,059 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-20 17:50:47,641 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-20 17:50:51,706 INFO [train.py:901] (0/2) Epoch 2, batch 1100, loss[loss=0.2988, simple_loss=0.3375, pruned_loss=0.13, over 7274.00 frames. ], tot_loss[loss=0.3037, simple_loss=0.3369, pruned_loss=0.1353, over 1433083.80 frames. ], batch size: 70, lr: 4.40e-02, grad_scale: 8.0 +2023-03-20 17:50:52,400 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=3926.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:50:54,979 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2080, 2.4075, 3.1920, 2.7242, 3.5620, 3.1985, 1.9113, 3.1756], + device='cuda:0'), covar=tensor([0.0125, 0.0610, 0.0072, 0.0180, 0.0065, 0.0143, 0.1426, 0.0220], + device='cuda:0'), in_proj_covar=tensor([0.0018, 0.0021, 0.0019, 0.0018, 0.0019, 0.0020, 0.0036, 0.0020], + device='cuda:0'), out_proj_covar=tensor([1.8340e-05, 2.3618e-05, 2.0206e-05, 1.9256e-05, 1.6321e-05, 1.9971e-05, + 4.7734e-05, 2.2263e-05], device='cuda:0') +2023-03-20 17:50:59,223 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.12 vs. limit=2.0 +2023-03-20 17:51:01,631 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=3944.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:51:17,282 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-20 17:51:17,790 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 17:51:18,396 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6848, 3.2224, 3.6107, 3.1598, 3.6306, 3.5422, 1.9915, 3.6430], + device='cuda:0'), covar=tensor([0.0092, 0.0243, 0.0066, 0.0193, 0.0059, 0.0113, 0.1217, 0.0125], + device='cuda:0'), in_proj_covar=tensor([0.0018, 0.0021, 0.0019, 0.0019, 0.0018, 0.0021, 0.0037, 0.0021], + device='cuda:0'), out_proj_covar=tensor([1.8319e-05, 2.3680e-05, 2.0100e-05, 2.0082e-05, 1.6362e-05, 2.0774e-05, + 4.8747e-05, 2.3229e-05], device='cuda:0') +2023-03-20 17:51:18,783 INFO [train.py:901] (0/2) Epoch 2, batch 1150, loss[loss=0.285, simple_loss=0.3322, pruned_loss=0.1189, over 7329.00 frames. ], tot_loss[loss=0.3038, simple_loss=0.3371, pruned_loss=0.1352, over 1436338.01 frames. ], batch size: 59, lr: 4.39e-02, grad_scale: 8.0 +2023-03-20 17:51:25,046 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=3987.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:51:27,461 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=3992.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:51:30,413 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-20 17:51:30,928 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-20 17:51:31,747 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-4000.pt +2023-03-20 17:51:36,691 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.009e+02 4.029e+02 4.723e+02 6.247e+02 1.092e+03, threshold=9.446e+02, percent-clipped=3.0 +2023-03-20 17:51:48,210 INFO [train.py:901] (0/2) Epoch 2, batch 1200, loss[loss=0.2852, simple_loss=0.3055, pruned_loss=0.1325, over 7055.00 frames. ], tot_loss[loss=0.3026, simple_loss=0.3361, pruned_loss=0.1346, over 1436546.91 frames. ], batch size: 35, lr: 4.38e-02, grad_scale: 8.0 +2023-03-20 17:51:51,848 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=4032.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:52:02,885 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7019, 3.8178, 3.7217, 3.6768, 3.7702, 3.5534, 3.8752, 3.1580], + device='cuda:0'), covar=tensor([0.0094, 0.0117, 0.0146, 0.0114, 0.0108, 0.0238, 0.0096, 0.0234], + device='cuda:0'), in_proj_covar=tensor([0.0013, 0.0015, 0.0016, 0.0016, 0.0015, 0.0017, 0.0018, 0.0018], + device='cuda:0'), out_proj_covar=tensor([1.3257e-05, 2.0746e-05, 1.8007e-05, 2.2201e-05, 1.8125e-05, 2.1346e-05, + 2.3299e-05, 2.4183e-05], device='cuda:0') +2023-03-20 17:52:08,967 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-20 17:52:15,182 INFO [train.py:901] (0/2) Epoch 2, batch 1250, loss[loss=0.3608, simple_loss=0.3749, pruned_loss=0.1734, over 7153.00 frames. ], tot_loss[loss=0.3021, simple_loss=0.3357, pruned_loss=0.1342, over 1436220.15 frames. ], batch size: 98, lr: 4.37e-02, grad_scale: 8.0 +2023-03-20 17:52:20,548 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.96 vs. limit=2.0 +2023-03-20 17:52:29,535 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.597e+02 4.003e+02 4.997e+02 6.171e+02 1.262e+03, threshold=9.993e+02, percent-clipped=4.0 +2023-03-20 17:52:32,126 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-20 17:52:36,231 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-20 17:52:37,323 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-20 17:52:40,933 INFO [train.py:901] (0/2) Epoch 2, batch 1300, loss[loss=0.2611, simple_loss=0.2872, pruned_loss=0.1175, over 7021.00 frames. ], tot_loss[loss=0.303, simple_loss=0.3368, pruned_loss=0.1346, over 1438480.54 frames. ], batch size: 35, lr: 4.36e-02, grad_scale: 8.0 +2023-03-20 17:52:50,479 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=4141.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:53:02,195 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-20 17:53:04,746 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-20 17:53:07,835 INFO [train.py:901] (0/2) Epoch 2, batch 1350, loss[loss=0.343, simple_loss=0.3721, pruned_loss=0.157, over 7203.00 frames. ], tot_loss[loss=0.3018, simple_loss=0.336, pruned_loss=0.1338, over 1440426.11 frames. ], batch size: 50, lr: 4.35e-02, grad_scale: 8.0 +2023-03-20 17:53:08,393 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-20 17:53:18,695 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-20 17:53:22,256 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=4202.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 17:53:22,310 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=4202.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:53:22,644 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.237e+02 3.810e+02 4.674e+02 6.452e+02 1.079e+03, threshold=9.348e+02, percent-clipped=2.0 +2023-03-20 17:53:34,474 INFO [train.py:901] (0/2) Epoch 2, batch 1400, loss[loss=0.3208, simple_loss=0.3552, pruned_loss=0.1431, over 7264.00 frames. ], tot_loss[loss=0.3006, simple_loss=0.3355, pruned_loss=0.1328, over 1442413.69 frames. ], batch size: 55, lr: 4.34e-02, grad_scale: 8.0 +2023-03-20 17:53:48,077 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=4250.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:53:50,848 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.24 vs. limit=5.0 +2023-03-20 17:53:51,555 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-20 17:53:59,355 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5177, 2.9493, 2.6689, 2.7085, 3.2114, 2.8359, 3.2769, 3.2422], + device='cuda:0'), covar=tensor([0.0825, 0.0162, 0.0791, 0.1059, 0.0316, 0.0366, 0.0228, 0.0263], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0026, 0.0030, 0.0031, 0.0032, 0.0027, 0.0025, 0.0032], + device='cuda:0'), out_proj_covar=tensor([2.9161e-05, 2.2838e-05, 2.9192e-05, 3.2340e-05, 3.3464e-05, 2.8577e-05, + 2.3672e-05, 3.5304e-05], device='cuda:0') +2023-03-20 17:54:00,775 INFO [train.py:901] (0/2) Epoch 2, batch 1450, loss[loss=0.3346, simple_loss=0.3595, pruned_loss=0.1548, over 7313.00 frames. ], tot_loss[loss=0.3021, simple_loss=0.3366, pruned_loss=0.1338, over 1443763.14 frames. ], batch size: 80, lr: 4.33e-02, grad_scale: 8.0 +2023-03-20 17:54:04,451 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=4282.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:54:07,599 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=4288.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:54:14,276 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-20 17:54:15,759 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.055e+02 3.966e+02 5.157e+02 6.398e+02 1.604e+03, threshold=1.031e+03, percent-clipped=6.0 +2023-03-20 17:54:16,379 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7406, 2.7834, 4.0198, 2.9216, 3.8531, 4.0560, 1.7321, 4.0594], + device='cuda:0'), covar=tensor([0.0056, 0.0372, 0.0038, 0.0266, 0.0038, 0.0050, 0.1287, 0.0078], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0022, 0.0021, 0.0021, 0.0019, 0.0021, 0.0042, 0.0021], + device='cuda:0'), out_proj_covar=tensor([1.9701e-05, 2.6442e-05, 2.1540e-05, 2.4125e-05, 1.7824e-05, 2.2267e-05, + 5.5648e-05, 2.4531e-05], device='cuda:0') +2023-03-20 17:54:20,856 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.97 vs. limit=2.0 +2023-03-20 17:54:27,658 INFO [train.py:901] (0/2) Epoch 2, batch 1500, loss[loss=0.2779, simple_loss=0.321, pruned_loss=0.1174, over 7283.00 frames. ], tot_loss[loss=0.2999, simple_loss=0.3353, pruned_loss=0.1323, over 1445156.30 frames. ], batch size: 66, lr: 4.32e-02, grad_scale: 8.0 +2023-03-20 17:54:31,379 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-20 17:54:31,482 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=4332.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:54:40,207 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=4349.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:54:46,205 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.00 vs. limit=2.0 +2023-03-20 17:54:53,523 INFO [train.py:901] (0/2) Epoch 2, batch 1550, loss[loss=0.2974, simple_loss=0.3402, pruned_loss=0.1273, over 7333.00 frames. ], tot_loss[loss=0.2982, simple_loss=0.3341, pruned_loss=0.1312, over 1444943.49 frames. ], batch size: 63, lr: 4.31e-02, grad_scale: 8.0 +2023-03-20 17:54:56,651 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-20 17:54:56,695 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=4380.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 17:55:04,108 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.00 vs. limit=2.0 +2023-03-20 17:55:09,256 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.966e+02 3.883e+02 4.668e+02 5.736e+02 1.136e+03, threshold=9.337e+02, percent-clipped=1.0 +2023-03-20 17:55:17,730 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-03-20 17:55:20,409 INFO [train.py:901] (0/2) Epoch 2, batch 1600, loss[loss=0.2921, simple_loss=0.3312, pruned_loss=0.1266, over 7361.00 frames. ], tot_loss[loss=0.2972, simple_loss=0.3335, pruned_loss=0.1304, over 1443557.09 frames. ], batch size: 73, lr: 4.30e-02, grad_scale: 8.0 +2023-03-20 17:55:27,464 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-20 17:55:28,446 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-20 17:55:31,495 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-20 17:55:33,262 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.00 vs. limit=2.0 +2023-03-20 17:55:34,337 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.05 vs. limit=2.0 +2023-03-20 17:55:41,806 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-20 17:55:46,534 INFO [train.py:901] (0/2) Epoch 2, batch 1650, loss[loss=0.2837, simple_loss=0.3267, pruned_loss=0.1203, over 7271.00 frames. ], tot_loss[loss=0.2957, simple_loss=0.3322, pruned_loss=0.1296, over 1443062.95 frames. ], batch size: 77, lr: 4.29e-02, grad_scale: 8.0 +2023-03-20 17:55:46,549 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-20 17:55:53,621 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.35 vs. limit=2.0 +2023-03-20 17:55:55,037 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3350, 3.2447, 2.4948, 3.1387, 3.1437, 3.3271, 3.1115, 1.9692], + device='cuda:0'), covar=tensor([0.0100, 0.0163, 0.0836, 0.0223, 0.0077, 0.0119, 0.1052, 0.0975], + device='cuda:0'), in_proj_covar=tensor([0.0026, 0.0032, 0.0051, 0.0034, 0.0026, 0.0027, 0.0060, 0.0055], + device='cuda:0'), out_proj_covar=tensor([1.7222e-05, 2.1808e-05, 3.9836e-05, 2.3091e-05, 1.6900e-05, 1.8666e-05, + 4.9153e-05, 3.9923e-05], device='cuda:0') +2023-03-20 17:55:55,369 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-20 17:55:58,506 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=4497.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:55:58,567 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1829, 2.6201, 2.1546, 2.8622, 1.8823, 2.2627, 2.7011, 2.0574], + device='cuda:0'), covar=tensor([0.0522, 0.0300, 0.0674, 0.0172, 0.1056, 0.0612, 0.0283, 0.0809], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0026, 0.0029, 0.0030, 0.0027, 0.0025, 0.0028, 0.0025], + device='cuda:0'), out_proj_covar=tensor([2.9625e-05, 2.6456e-05, 3.0048e-05, 3.1236e-05, 3.0429e-05, 2.9384e-05, + 2.8269e-05, 3.1029e-05], device='cuda:0') +2023-03-20 17:56:01,468 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.052e+02 3.748e+02 4.507e+02 5.653e+02 9.298e+02, threshold=9.013e+02, percent-clipped=1.0 +2023-03-20 17:56:12,154 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.61 vs. limit=5.0 +2023-03-20 17:56:12,509 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3342, 4.2301, 4.1197, 3.7066, 4.2030, 4.3600, 4.6153, 4.2031], + device='cuda:0'), covar=tensor([0.0176, 0.0097, 0.0174, 0.0235, 0.0139, 0.0090, 0.0135, 0.0160], + device='cuda:0'), in_proj_covar=tensor([0.0029, 0.0027, 0.0032, 0.0028, 0.0037, 0.0031, 0.0027, 0.0028], + device='cuda:0'), out_proj_covar=tensor([3.9221e-05, 3.2375e-05, 4.7816e-05, 4.1894e-05, 5.3213e-05, 3.9203e-05, + 3.7374e-05, 3.5941e-05], device='cuda:0') +2023-03-20 17:56:12,928 INFO [train.py:901] (0/2) Epoch 2, batch 1700, loss[loss=0.2642, simple_loss=0.2987, pruned_loss=0.1148, over 7216.00 frames. ], tot_loss[loss=0.2933, simple_loss=0.3305, pruned_loss=0.128, over 1443429.95 frames. ], batch size: 39, lr: 4.28e-02, grad_scale: 8.0 +2023-03-20 17:56:12,955 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 17:56:17,009 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-20 17:56:27,939 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-20 17:56:34,454 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.97 vs. limit=2.0 +2023-03-20 17:56:39,100 INFO [train.py:901] (0/2) Epoch 2, batch 1750, loss[loss=0.3106, simple_loss=0.3444, pruned_loss=0.1384, over 7267.00 frames. ], tot_loss[loss=0.2938, simple_loss=0.331, pruned_loss=0.1283, over 1444398.44 frames. ], batch size: 70, lr: 4.27e-02, grad_scale: 8.0 +2023-03-20 17:56:43,360 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=4582.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:56:51,378 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.97 vs. limit=2.0 +2023-03-20 17:56:54,016 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-20 17:56:54,506 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.543e+02 3.909e+02 5.009e+02 6.432e+02 1.209e+03, threshold=1.002e+03, percent-clipped=8.0 +2023-03-20 17:56:55,031 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-20 17:57:05,835 INFO [train.py:901] (0/2) Epoch 2, batch 1800, loss[loss=0.2988, simple_loss=0.3327, pruned_loss=0.1324, over 7307.00 frames. ], tot_loss[loss=0.294, simple_loss=0.3311, pruned_loss=0.1284, over 1443840.15 frames. ], batch size: 86, lr: 4.25e-02, grad_scale: 8.0 +2023-03-20 17:57:08,431 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=4630.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:57:15,741 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8547, 2.9039, 2.0593, 2.3680, 2.4194, 2.2427, 2.8379, 2.6504], + device='cuda:0'), covar=tensor([0.0231, 0.0100, 0.1064, 0.0572, 0.0499, 0.0476, 0.0244, 0.0319], + device='cuda:0'), in_proj_covar=tensor([0.0024, 0.0022, 0.0026, 0.0027, 0.0029, 0.0024, 0.0023, 0.0030], + device='cuda:0'), out_proj_covar=tensor([2.6566e-05, 2.0813e-05, 2.7784e-05, 3.1076e-05, 3.2444e-05, 2.5726e-05, + 2.3264e-05, 3.3370e-05], device='cuda:0') +2023-03-20 17:57:16,204 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=4644.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:57:16,655 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-20 17:57:31,217 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-20 17:57:32,699 INFO [train.py:901] (0/2) Epoch 2, batch 1850, loss[loss=0.2829, simple_loss=0.3265, pruned_loss=0.1196, over 7316.00 frames. ], tot_loss[loss=0.2921, simple_loss=0.3299, pruned_loss=0.1272, over 1443059.38 frames. ], batch size: 59, lr: 4.24e-02, grad_scale: 8.0 +2023-03-20 17:57:33,840 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2492, 3.5618, 3.5136, 3.4737, 3.1609, 3.5739, 3.8253, 3.9064], + device='cuda:0'), covar=tensor([0.0368, 0.0255, 0.0266, 0.0264, 0.0458, 0.0342, 0.0242, 0.0155], + device='cuda:0'), in_proj_covar=tensor([0.0059, 0.0062, 0.0058, 0.0076, 0.0064, 0.0054, 0.0049, 0.0053], + device='cuda:0'), out_proj_covar=tensor([7.4342e-05, 7.9063e-05, 7.3265e-05, 1.0661e-04, 8.5323e-05, 6.9993e-05, + 6.2584e-05, 6.7993e-05], device='cuda:0') +2023-03-20 17:57:40,927 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-20 17:57:47,094 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.532e+02 3.943e+02 4.713e+02 6.122e+02 1.050e+03, threshold=9.426e+02, percent-clipped=1.0 +2023-03-20 17:57:48,976 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.47 vs. limit=5.0 +2023-03-20 17:57:50,833 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5110, 1.4115, 2.4379, 2.1686, 1.6066, 2.2918, 2.4602, 1.7180], + device='cuda:0'), covar=tensor([0.1377, 0.1508, 0.0360, 0.0386, 0.1083, 0.0530, 0.0194, 0.0587], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0044, 0.0042, 0.0044, 0.0041, 0.0043, 0.0038, 0.0040], + device='cuda:0'), out_proj_covar=tensor([4.2007e-05, 4.7556e-05, 5.1089e-05, 4.4885e-05, 4.4748e-05, 4.4082e-05, + 3.4939e-05, 3.7635e-05], device='cuda:0') +2023-03-20 17:57:54,072 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=6.37 vs. limit=5.0 +2023-03-20 17:57:56,381 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-20 17:57:58,962 INFO [train.py:901] (0/2) Epoch 2, batch 1900, loss[loss=0.2764, simple_loss=0.3215, pruned_loss=0.1156, over 7347.00 frames. ], tot_loss[loss=0.2923, simple_loss=0.3301, pruned_loss=0.1272, over 1444280.71 frames. ], batch size: 73, lr: 4.23e-02, grad_scale: 8.0 +2023-03-20 17:58:23,203 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-20 17:58:25,291 INFO [train.py:901] (0/2) Epoch 2, batch 1950, loss[loss=0.2429, simple_loss=0.2885, pruned_loss=0.09864, over 7151.00 frames. ], tot_loss[loss=0.2908, simple_loss=0.329, pruned_loss=0.1263, over 1442969.61 frames. ], batch size: 41, lr: 4.22e-02, grad_scale: 8.0 +2023-03-20 17:58:34,340 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-20 17:58:36,487 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=4797.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:58:38,764 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-20 17:58:39,750 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.425e+02 3.763e+02 4.603e+02 6.493e+02 1.635e+03, threshold=9.206e+02, percent-clipped=4.0 +2023-03-20 17:58:39,791 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-20 17:58:50,412 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.14 vs. limit=2.0 +2023-03-20 17:58:51,646 INFO [train.py:901] (0/2) Epoch 2, batch 2000, loss[loss=0.2941, simple_loss=0.3283, pruned_loss=0.13, over 7240.00 frames. ], tot_loss[loss=0.2909, simple_loss=0.329, pruned_loss=0.1264, over 1442748.06 frames. ], batch size: 45, lr: 4.21e-02, grad_scale: 8.0 +2023-03-20 17:58:58,054 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-20 17:59:02,686 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=4845.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 17:59:09,324 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-20 17:59:17,522 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-20 17:59:18,003 INFO [train.py:901] (0/2) Epoch 2, batch 2050, loss[loss=0.3199, simple_loss=0.358, pruned_loss=0.1409, over 7258.00 frames. ], tot_loss[loss=0.2888, simple_loss=0.3276, pruned_loss=0.125, over 1442042.20 frames. ], batch size: 89, lr: 4.20e-02, grad_scale: 8.0 +2023-03-20 17:59:24,559 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.52 vs. limit=2.0 +2023-03-20 17:59:27,945 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9629, 1.2382, 2.1702, 1.7848, 1.6860, 1.8271, 2.0810, 1.5631], + device='cuda:0'), covar=tensor([0.0395, 0.1327, 0.0373, 0.0468, 0.0996, 0.0571, 0.0235, 0.0488], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0039, 0.0040, 0.0043, 0.0039, 0.0040, 0.0036, 0.0038], + device='cuda:0'), out_proj_covar=tensor([4.0099e-05, 4.2798e-05, 4.8908e-05, 4.4826e-05, 4.4245e-05, 4.3547e-05, + 3.3672e-05, 3.6634e-05], device='cuda:0') +2023-03-20 17:59:33,000 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.296e+02 3.547e+02 4.263e+02 5.510e+02 1.038e+03, threshold=8.526e+02, percent-clipped=1.0 +2023-03-20 17:59:34,872 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.00 vs. limit=2.0 +2023-03-20 17:59:44,893 INFO [train.py:901] (0/2) Epoch 2, batch 2100, loss[loss=0.271, simple_loss=0.3155, pruned_loss=0.1132, over 7357.00 frames. ], tot_loss[loss=0.2874, simple_loss=0.3264, pruned_loss=0.1241, over 1442833.42 frames. ], batch size: 63, lr: 4.19e-02, grad_scale: 8.0 +2023-03-20 17:59:45,284 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.00 vs. limit=2.0 +2023-03-20 17:59:45,679 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=6.12 vs. limit=5.0 +2023-03-20 17:59:46,491 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1108, 2.7046, 2.1793, 2.7200, 0.9631, 2.1979, 2.6052, 1.9794], + device='cuda:0'), covar=tensor([0.0768, 0.0213, 0.0659, 0.0270, 0.1853, 0.0928, 0.0397, 0.1066], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0027, 0.0031, 0.0030, 0.0028, 0.0026, 0.0029, 0.0026], + device='cuda:0'), out_proj_covar=tensor([3.3413e-05, 2.7851e-05, 3.3833e-05, 3.3583e-05, 3.2567e-05, 3.3404e-05, + 3.1828e-05, 3.4103e-05], device='cuda:0') +2023-03-20 17:59:51,547 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-20 17:59:54,115 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-20 17:59:54,717 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=4944.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:00:03,177 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.04 vs. limit=2.0 +2023-03-20 18:00:10,567 INFO [train.py:901] (0/2) Epoch 2, batch 2150, loss[loss=0.3029, simple_loss=0.3455, pruned_loss=0.1301, over 7274.00 frames. ], tot_loss[loss=0.288, simple_loss=0.327, pruned_loss=0.1245, over 1440203.39 frames. ], batch size: 77, lr: 4.18e-02, grad_scale: 8.0 +2023-03-20 18:00:12,841 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.77 vs. limit=5.0 +2023-03-20 18:00:19,865 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=4992.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:00:23,074 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3159, 2.9241, 3.1207, 2.8056, 3.2331, 3.1137, 1.5626, 3.2845], + device='cuda:0'), covar=tensor([0.0056, 0.0273, 0.0110, 0.0127, 0.0053, 0.0138, 0.1312, 0.0140], + device='cuda:0'), in_proj_covar=tensor([0.0023, 0.0026, 0.0027, 0.0025, 0.0024, 0.0027, 0.0049, 0.0027], + device='cuda:0'), out_proj_covar=tensor([2.3861e-05, 3.3030e-05, 2.9822e-05, 2.9967e-05, 2.2380e-05, 2.9197e-05, + 6.8932e-05, 3.1439e-05], device='cuda:0') +2023-03-20 18:00:25,465 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2021, 2.6182, 2.5598, 2.7391, 1.5508, 2.4119, 2.6256, 2.2830], + device='cuda:0'), covar=tensor([0.0416, 0.0142, 0.0274, 0.0188, 0.1121, 0.0685, 0.0193, 0.0413], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0026, 0.0028, 0.0029, 0.0026, 0.0025, 0.0028, 0.0023], + device='cuda:0'), out_proj_covar=tensor([3.0248e-05, 2.7039e-05, 3.2411e-05, 3.2919e-05, 3.1116e-05, 3.1710e-05, + 3.0646e-05, 3.0567e-05], device='cuda:0') +2023-03-20 18:00:25,810 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.955e+02 4.119e+02 4.986e+02 6.315e+02 1.234e+03, threshold=9.973e+02, percent-clipped=7.0 +2023-03-20 18:00:37,646 INFO [train.py:901] (0/2) Epoch 2, batch 2200, loss[loss=0.238, simple_loss=0.2654, pruned_loss=0.1053, over 5978.00 frames. ], tot_loss[loss=0.2884, simple_loss=0.3272, pruned_loss=0.1248, over 1437868.29 frames. ], batch size: 26, lr: 4.17e-02, grad_scale: 8.0 +2023-03-20 18:00:40,194 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-20 18:01:03,874 INFO [train.py:901] (0/2) Epoch 2, batch 2250, loss[loss=0.3254, simple_loss=0.3507, pruned_loss=0.15, over 7320.00 frames. ], tot_loss[loss=0.2888, simple_loss=0.3277, pruned_loss=0.125, over 1439110.64 frames. ], batch size: 59, lr: 4.16e-02, grad_scale: 8.0 +2023-03-20 18:01:14,658 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-20 18:01:14,673 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-20 18:01:18,732 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.254e+02 3.930e+02 5.031e+02 6.019e+02 2.322e+03, threshold=1.006e+03, percent-clipped=4.0 +2023-03-20 18:01:18,917 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1366, 2.6350, 2.5583, 3.1995, 2.8719, 3.3037, 2.7074, 2.9207], + device='cuda:0'), covar=tensor([0.1541, 0.0652, 0.2438, 0.0189, 0.0080, 0.0129, 0.0053, 0.0111], + device='cuda:0'), in_proj_covar=tensor([0.0117, 0.0082, 0.0125, 0.0050, 0.0054, 0.0063, 0.0048, 0.0052], + device='cuda:0'), out_proj_covar=tensor([8.8034e-05, 6.0545e-05, 1.0194e-04, 3.7849e-05, 3.8228e-05, 4.4967e-05, + 3.3959e-05, 3.5647e-05], device='cuda:0') +2023-03-20 18:01:28,429 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-20 18:01:29,960 INFO [train.py:901] (0/2) Epoch 2, batch 2300, loss[loss=0.2733, simple_loss=0.3223, pruned_loss=0.1122, over 7275.00 frames. ], tot_loss[loss=0.2882, simple_loss=0.3273, pruned_loss=0.1246, over 1439490.12 frames. ], batch size: 70, lr: 4.15e-02, grad_scale: 8.0 +2023-03-20 18:01:30,863 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.98 vs. limit=2.0 +2023-03-20 18:01:50,496 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4858, 3.6440, 3.6368, 3.5756, 3.5787, 3.4275, 3.7301, 3.1758], + device='cuda:0'), covar=tensor([0.0104, 0.0149, 0.0137, 0.0104, 0.0134, 0.0161, 0.0156, 0.0171], + device='cuda:0'), in_proj_covar=tensor([0.0016, 0.0017, 0.0017, 0.0017, 0.0016, 0.0018, 0.0022, 0.0019], + device='cuda:0'), out_proj_covar=tensor([2.1567e-05, 3.0194e-05, 2.5437e-05, 2.8761e-05, 2.6404e-05, 2.6549e-05, + 3.4830e-05, 3.0050e-05], device='cuda:0') +2023-03-20 18:01:55,976 INFO [train.py:901] (0/2) Epoch 2, batch 2350, loss[loss=0.2427, simple_loss=0.2743, pruned_loss=0.1055, over 7021.00 frames. ], tot_loss[loss=0.2873, simple_loss=0.3266, pruned_loss=0.124, over 1439538.50 frames. ], batch size: 35, lr: 4.14e-02, grad_scale: 8.0 +2023-03-20 18:02:11,290 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.216e+02 3.739e+02 4.530e+02 5.733e+02 1.414e+03, threshold=9.059e+02, percent-clipped=1.0 +2023-03-20 18:02:15,353 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-20 18:02:22,461 INFO [train.py:901] (0/2) Epoch 2, batch 2400, loss[loss=0.2583, simple_loss=0.3044, pruned_loss=0.1061, over 7218.00 frames. ], tot_loss[loss=0.2873, simple_loss=0.3266, pruned_loss=0.124, over 1441256.10 frames. ], batch size: 50, lr: 4.13e-02, grad_scale: 8.0 +2023-03-20 18:02:22,478 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-20 18:02:34,443 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-20 18:02:37,587 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-20 18:02:45,136 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.99 vs. limit=2.0 +2023-03-20 18:02:49,398 INFO [train.py:901] (0/2) Epoch 2, batch 2450, loss[loss=0.3299, simple_loss=0.3654, pruned_loss=0.1472, over 6713.00 frames. ], tot_loss[loss=0.2868, simple_loss=0.3262, pruned_loss=0.1237, over 1440213.29 frames. ], batch size: 107, lr: 4.12e-02, grad_scale: 8.0 +2023-03-20 18:02:59,203 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.26 vs. limit=5.0 +2023-03-20 18:03:03,829 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.303e+02 3.507e+02 4.658e+02 5.816e+02 1.903e+03, threshold=9.317e+02, percent-clipped=8.0 +2023-03-20 18:03:05,425 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-20 18:03:15,084 INFO [train.py:901] (0/2) Epoch 2, batch 2500, loss[loss=0.37, simple_loss=0.3889, pruned_loss=0.1756, over 6560.00 frames. ], tot_loss[loss=0.2854, simple_loss=0.3245, pruned_loss=0.1232, over 1438273.24 frames. ], batch size: 106, lr: 4.11e-02, grad_scale: 8.0 +2023-03-20 18:03:31,843 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-20 18:03:41,957 INFO [train.py:901] (0/2) Epoch 2, batch 2550, loss[loss=0.3637, simple_loss=0.3898, pruned_loss=0.1689, over 6673.00 frames. ], tot_loss[loss=0.2856, simple_loss=0.3245, pruned_loss=0.1234, over 1440282.00 frames. ], batch size: 107, lr: 4.10e-02, grad_scale: 8.0 +2023-03-20 18:03:56,655 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.829e+02 4.460e+02 5.219e+02 6.582e+02 1.378e+03, threshold=1.044e+03, percent-clipped=5.0 +2023-03-20 18:04:07,524 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.98 vs. limit=5.0 +2023-03-20 18:04:08,623 INFO [train.py:901] (0/2) Epoch 2, batch 2600, loss[loss=0.2911, simple_loss=0.3381, pruned_loss=0.122, over 7330.00 frames. ], tot_loss[loss=0.2846, simple_loss=0.3242, pruned_loss=0.1225, over 1441282.90 frames. ], batch size: 61, lr: 4.09e-02, grad_scale: 16.0 +2023-03-20 18:04:30,769 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6903, 3.1677, 2.2282, 3.6303, 1.3286, 2.7134, 2.6567, 1.9492], + device='cuda:0'), covar=tensor([0.0238, 0.0106, 0.0648, 0.0085, 0.1543, 0.0511, 0.0272, 0.0758], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0026, 0.0028, 0.0026, 0.0025, 0.0025, 0.0027, 0.0023], + device='cuda:0'), out_proj_covar=tensor([3.0963e-05, 2.9468e-05, 3.3982e-05, 3.2710e-05, 3.1924e-05, 3.3023e-05, + 3.2755e-05, 3.4354e-05], device='cuda:0') +2023-03-20 18:04:34,280 INFO [train.py:901] (0/2) Epoch 2, batch 2650, loss[loss=0.2083, simple_loss=0.2387, pruned_loss=0.08891, over 5875.00 frames. ], tot_loss[loss=0.284, simple_loss=0.3235, pruned_loss=0.1223, over 1439635.42 frames. ], batch size: 25, lr: 4.08e-02, grad_scale: 16.0 +2023-03-20 18:04:48,315 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.928e+02 3.829e+02 4.633e+02 5.948e+02 1.485e+03, threshold=9.266e+02, percent-clipped=1.0 +2023-03-20 18:04:59,254 INFO [train.py:901] (0/2) Epoch 2, batch 2700, loss[loss=0.314, simple_loss=0.3482, pruned_loss=0.1399, over 7246.00 frames. ], tot_loss[loss=0.2837, simple_loss=0.3234, pruned_loss=0.122, over 1439423.26 frames. ], batch size: 55, lr: 4.07e-02, grad_scale: 16.0 +2023-03-20 18:05:00,129 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.95 vs. limit=2.0 +2023-03-20 18:05:24,194 INFO [train.py:901] (0/2) Epoch 2, batch 2750, loss[loss=0.2604, simple_loss=0.3062, pruned_loss=0.1073, over 7151.00 frames. ], tot_loss[loss=0.2829, simple_loss=0.3226, pruned_loss=0.1216, over 1438093.92 frames. ], batch size: 41, lr: 4.06e-02, grad_scale: 16.0 +2023-03-20 18:05:38,367 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.897e+02 3.942e+02 4.778e+02 5.507e+02 1.235e+03, threshold=9.556e+02, percent-clipped=3.0 +2023-03-20 18:05:49,299 INFO [train.py:901] (0/2) Epoch 2, batch 2800, loss[loss=0.2735, simple_loss=0.3178, pruned_loss=0.1147, over 7273.00 frames. ], tot_loss[loss=0.2822, simple_loss=0.3226, pruned_loss=0.1209, over 1440534.02 frames. ], batch size: 52, lr: 4.05e-02, grad_scale: 16.0 +2023-03-20 18:05:52,326 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7801, 4.0189, 4.0902, 4.1305, 3.9771, 3.7019, 4.2277, 4.1793], + device='cuda:0'), covar=tensor([0.0525, 0.0250, 0.0296, 0.0500, 0.0467, 0.0407, 0.0328, 0.0295], + device='cuda:0'), in_proj_covar=tensor([0.0065, 0.0049, 0.0060, 0.0054, 0.0053, 0.0058, 0.0053, 0.0052], + device='cuda:0'), out_proj_covar=tensor([8.5463e-05, 5.9254e-05, 7.7331e-05, 7.5504e-05, 7.1658e-05, 7.5428e-05, + 6.8327e-05, 6.9688e-05], device='cuda:0') +2023-03-20 18:06:01,978 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-2.pt +2023-03-20 18:06:19,830 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-20 18:06:19,887 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-20 18:06:19,913 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-20 18:06:19,931 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-20 18:06:20,109 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-20 18:06:20,129 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-20 18:06:20,140 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-20 18:06:20,579 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-20 18:06:20,655 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-20 18:06:20,670 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-20 18:06:20,688 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-20 18:06:20,857 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-20 18:06:20,863 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-20 18:06:20,871 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-20 18:06:21,012 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-20 18:06:21,014 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-20 18:06:21,074 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-20 18:06:23,287 INFO [train.py:901] (0/2) Epoch 3, batch 0, loss[loss=0.2637, simple_loss=0.3004, pruned_loss=0.1135, over 7200.00 frames. ], tot_loss[loss=0.2637, simple_loss=0.3004, pruned_loss=0.1135, over 7200.00 frames. ], batch size: 39, lr: 3.96e-02, grad_scale: 16.0 +2023-03-20 18:06:23,288 INFO [train.py:926] (0/2) Computing validation loss +2023-03-20 18:06:33,402 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6609, 2.6608, 2.6327, 1.9911, 1.3565, 1.8338, 2.8718, 1.9472], + device='cuda:0'), covar=tensor([0.0820, 0.0422, 0.0453, 0.0847, 0.2144, 0.0891, 0.0785, 0.1145], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0033, 0.0031, 0.0037, 0.0037, 0.0037, 0.0039, 0.0031], + device='cuda:0'), out_proj_covar=tensor([4.5547e-05, 3.9348e-05, 3.6175e-05, 4.1247e-05, 4.3620e-05, 4.3305e-05, + 4.5908e-05, 3.7488e-05], device='cuda:0') +2023-03-20 18:06:48,491 INFO [train.py:935] (0/2) Epoch 3, validation: loss=0.2204, simple_loss=0.3002, pruned_loss=0.07031, over 1622729.00 frames. +2023-03-20 18:06:48,492 INFO [train.py:936] (0/2) Maximum memory allocated so far is 11706MB +2023-03-20 18:06:49,143 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=5650.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 18:06:55,569 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-20 18:06:57,493 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.48 vs. limit=5.0 +2023-03-20 18:07:03,542 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6365, 3.3301, 3.1010, 3.6401, 2.9134, 2.2592, 3.4747, 3.0775], + device='cuda:0'), covar=tensor([0.0234, 0.0089, 0.0293, 0.0119, 0.0258, 0.1305, 0.0304, 0.0610], + device='cuda:0'), in_proj_covar=tensor([0.0061, 0.0051, 0.0076, 0.0067, 0.0084, 0.0175, 0.0062, 0.0136], + device='cuda:0'), out_proj_covar=tensor([3.9468e-05, 3.2239e-05, 4.7373e-05, 4.3606e-05, 5.2705e-05, 1.1837e-04, + 4.1161e-05, 9.8390e-05], device='cuda:0') +2023-03-20 18:07:07,897 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-20 18:07:15,079 INFO [train.py:901] (0/2) Epoch 3, batch 50, loss[loss=0.2624, simple_loss=0.3139, pruned_loss=0.1055, over 7331.00 frames. ], tot_loss[loss=0.2754, simple_loss=0.317, pruned_loss=0.1169, over 323950.40 frames. ], batch size: 61, lr: 3.95e-02, grad_scale: 16.0 +2023-03-20 18:07:16,075 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-20 18:07:17,060 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.426e+02 3.989e+02 5.085e+02 6.289e+02 1.206e+03, threshold=1.017e+03, percent-clipped=2.0 +2023-03-20 18:07:18,596 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-20 18:07:19,751 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2773, 3.1201, 2.3756, 2.9835, 2.5887, 3.1880, 3.2289, 3.1442], + device='cuda:0'), covar=tensor([0.0113, 0.0465, 0.3415, 0.0095, 0.4246, 0.0176, 0.0591, 0.0115], + device='cuda:0'), in_proj_covar=tensor([0.0085, 0.0124, 0.0247, 0.0086, 0.0218, 0.0098, 0.0152, 0.0086], + device='cuda:0'), out_proj_covar=tensor([6.1852e-05, 1.0211e-04, 1.7937e-04, 6.4899e-05, 1.6796e-04, 7.0939e-05, + 1.1896e-04, 6.3953e-05], device='cuda:0') +2023-03-20 18:07:21,273 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=5711.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 18:07:22,126 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-20 18:07:34,918 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3550, 4.1757, 4.2149, 4.4801, 4.6974, 4.6620, 3.9569, 3.9972], + device='cuda:0'), covar=tensor([0.0616, 0.1348, 0.1929, 0.1295, 0.0434, 0.1082, 0.0731, 0.0821], + device='cuda:0'), in_proj_covar=tensor([0.0078, 0.0134, 0.0147, 0.0112, 0.0086, 0.0119, 0.0077, 0.0095], + device='cuda:0'), out_proj_covar=tensor([8.5203e-05, 1.5309e-04, 1.6929e-04, 1.3669e-04, 9.4911e-05, 1.3660e-04, + 8.4501e-05, 1.0248e-04], device='cuda:0') +2023-03-20 18:07:38,396 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-20 18:07:38,408 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-20 18:07:40,414 INFO [train.py:901] (0/2) Epoch 3, batch 100, loss[loss=0.3131, simple_loss=0.3459, pruned_loss=0.1402, over 7310.00 frames. ], tot_loss[loss=0.2814, simple_loss=0.3228, pruned_loss=0.12, over 571037.47 frames. ], batch size: 49, lr: 3.95e-02, grad_scale: 16.0 +2023-03-20 18:07:42,210 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.12 vs. limit=2.0 +2023-03-20 18:07:43,591 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=5755.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:08:07,088 INFO [train.py:901] (0/2) Epoch 3, batch 150, loss[loss=0.2568, simple_loss=0.3004, pruned_loss=0.1066, over 7338.00 frames. ], tot_loss[loss=0.2801, simple_loss=0.3221, pruned_loss=0.119, over 765230.53 frames. ], batch size: 44, lr: 3.94e-02, grad_scale: 16.0 +2023-03-20 18:08:09,474 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.257e+02 3.860e+02 4.707e+02 5.554e+02 1.675e+03, threshold=9.414e+02, percent-clipped=1.0 +2023-03-20 18:08:14,724 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3923, 3.6917, 3.2485, 3.6244, 3.0015, 3.3419, 3.5610, 3.0817], + device='cuda:0'), covar=tensor([0.0131, 0.0134, 0.0166, 0.0175, 0.0184, 0.0169, 0.0165, 0.0187], + device='cuda:0'), in_proj_covar=tensor([0.0016, 0.0019, 0.0018, 0.0018, 0.0017, 0.0019, 0.0022, 0.0020], + device='cuda:0'), out_proj_covar=tensor([2.5972e-05, 3.5917e-05, 3.0266e-05, 3.4129e-05, 3.0433e-05, 3.2148e-05, + 3.9965e-05, 3.6141e-05], device='cuda:0') +2023-03-20 18:08:16,305 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=5816.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:08:33,342 INFO [train.py:901] (0/2) Epoch 3, batch 200, loss[loss=0.297, simple_loss=0.3399, pruned_loss=0.1271, over 7342.00 frames. ], tot_loss[loss=0.2827, simple_loss=0.3244, pruned_loss=0.1205, over 914953.49 frames. ], batch size: 61, lr: 3.93e-02, grad_scale: 16.0 +2023-03-20 18:08:37,918 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-20 18:08:38,014 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=5858.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:08:43,606 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-20 18:08:50,077 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-20 18:08:59,373 INFO [train.py:901] (0/2) Epoch 3, batch 250, loss[loss=0.2744, simple_loss=0.3181, pruned_loss=0.1154, over 7274.00 frames. ], tot_loss[loss=0.2796, simple_loss=0.3214, pruned_loss=0.1189, over 1029787.48 frames. ], batch size: 47, lr: 3.92e-02, grad_scale: 16.0 +2023-03-20 18:09:01,496 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.164e+02 3.767e+02 5.039e+02 6.362e+02 1.347e+03, threshold=1.008e+03, percent-clipped=5.0 +2023-03-20 18:09:03,651 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-20 18:09:09,906 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=5919.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:09:23,807 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.83 vs. limit=5.0 +2023-03-20 18:09:25,094 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4920, 4.5068, 4.3358, 4.5550, 4.8052, 4.8276, 4.3859, 4.0176], + device='cuda:0'), covar=tensor([0.0475, 0.1197, 0.1840, 0.1983, 0.0522, 0.1033, 0.0586, 0.0850], + device='cuda:0'), in_proj_covar=tensor([0.0078, 0.0135, 0.0147, 0.0117, 0.0086, 0.0124, 0.0076, 0.0095], + device='cuda:0'), out_proj_covar=tensor([8.5081e-05, 1.5432e-04, 1.7013e-04, 1.4055e-04, 9.4644e-05, 1.4407e-04, + 8.3911e-05, 1.0241e-04], device='cuda:0') +2023-03-20 18:09:25,532 INFO [train.py:901] (0/2) Epoch 3, batch 300, loss[loss=0.257, simple_loss=0.309, pruned_loss=0.1025, over 7315.00 frames. ], tot_loss[loss=0.2792, simple_loss=0.3204, pruned_loss=0.119, over 1121032.57 frames. ], batch size: 59, lr: 3.91e-02, grad_scale: 16.0 +2023-03-20 18:09:25,554 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-20 18:09:32,575 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5364, 1.3826, 2.1973, 1.7906, 1.3109, 1.4045, 1.7369, 1.4163], + device='cuda:0'), covar=tensor([0.0541, 0.0852, 0.0383, 0.0634, 0.1266, 0.1459, 0.0218, 0.0627], + device='cuda:0'), in_proj_covar=tensor([0.0033, 0.0035, 0.0034, 0.0036, 0.0033, 0.0034, 0.0034, 0.0036], + device='cuda:0'), out_proj_covar=tensor([3.8787e-05, 4.3664e-05, 4.8725e-05, 3.8526e-05, 4.2817e-05, 4.0723e-05, + 3.1936e-05, 3.8922e-05], device='cuda:0') +2023-03-20 18:09:32,715 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.86 vs. limit=2.0 +2023-03-20 18:09:36,463 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-20 18:09:47,156 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=5990.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:09:51,478 INFO [train.py:901] (0/2) Epoch 3, batch 350, loss[loss=0.2685, simple_loss=0.3208, pruned_loss=0.1081, over 7368.00 frames. ], tot_loss[loss=0.2771, simple_loss=0.3188, pruned_loss=0.1178, over 1192054.88 frames. ], batch size: 51, lr: 3.90e-02, grad_scale: 16.0 +2023-03-20 18:09:53,925 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.502e+02 3.944e+02 4.876e+02 5.780e+02 1.019e+03, threshold=9.752e+02, percent-clipped=1.0 +2023-03-20 18:09:55,509 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6006.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 18:10:11,579 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-20 18:10:14,861 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6043.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:10:17,765 INFO [train.py:901] (0/2) Epoch 3, batch 400, loss[loss=0.2754, simple_loss=0.3194, pruned_loss=0.1157, over 7279.00 frames. ], tot_loss[loss=0.2765, simple_loss=0.3185, pruned_loss=0.1172, over 1248874.95 frames. ], batch size: 57, lr: 3.89e-02, grad_scale: 16.0 +2023-03-20 18:10:19,579 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6051.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:10:28,253 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6091, 2.7649, 2.3239, 2.2924, 2.4412, 2.2805, 2.2570, 2.8937], + device='cuda:0'), covar=tensor([0.0478, 0.0149, 0.0756, 0.0718, 0.0467, 0.0744, 0.0623, 0.0294], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0024, 0.0026, 0.0027, 0.0028, 0.0028, 0.0029, 0.0030], + device='cuda:0'), out_proj_covar=tensor([3.6319e-05, 2.8352e-05, 3.5164e-05, 3.6203e-05, 3.7100e-05, 3.7171e-05, + 3.5283e-05, 4.0282e-05], device='cuda:0') +2023-03-20 18:10:43,820 INFO [train.py:901] (0/2) Epoch 3, batch 450, loss[loss=0.3112, simple_loss=0.3528, pruned_loss=0.1348, over 7118.00 frames. ], tot_loss[loss=0.2778, simple_loss=0.3198, pruned_loss=0.1179, over 1292710.72 frames. ], batch size: 98, lr: 3.88e-02, grad_scale: 16.0 +2023-03-20 18:10:45,831 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.319e+02 3.819e+02 4.941e+02 5.954e+02 1.043e+03, threshold=9.881e+02, percent-clipped=2.0 +2023-03-20 18:10:46,517 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6104.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:10:50,576 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6111.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:10:52,035 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-20 18:10:52,589 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-20 18:11:03,554 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6135.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:11:08,024 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.95 vs. limit=2.0 +2023-03-20 18:11:10,622 INFO [train.py:901] (0/2) Epoch 3, batch 500, loss[loss=0.2687, simple_loss=0.3107, pruned_loss=0.1134, over 7287.00 frames. ], tot_loss[loss=0.2771, simple_loss=0.3193, pruned_loss=0.1174, over 1324795.88 frames. ], batch size: 66, lr: 3.87e-02, grad_scale: 16.0 +2023-03-20 18:11:10,803 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5513, 3.3993, 2.7318, 3.2300, 2.6746, 2.1855, 3.4489, 2.9526], + device='cuda:0'), covar=tensor([0.0142, 0.0079, 0.0329, 0.0216, 0.0429, 0.1447, 0.0222, 0.0992], + device='cuda:0'), in_proj_covar=tensor([0.0064, 0.0058, 0.0089, 0.0077, 0.0101, 0.0192, 0.0071, 0.0154], + device='cuda:0'), out_proj_covar=tensor([4.3753e-05, 3.7568e-05, 5.5729e-05, 5.1199e-05, 6.6168e-05, 1.3292e-04, + 4.9941e-05, 1.1173e-04], device='cuda:0') +2023-03-20 18:11:21,433 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6170.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:11:22,589 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.02 vs. limit=2.0 +2023-03-20 18:11:25,864 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-20 18:11:27,350 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-20 18:11:27,892 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-20 18:11:30,444 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-20 18:11:35,152 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6196.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:11:35,535 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-20 18:11:36,535 INFO [train.py:901] (0/2) Epoch 3, batch 550, loss[loss=0.2358, simple_loss=0.2836, pruned_loss=0.09402, over 7139.00 frames. ], tot_loss[loss=0.2771, simple_loss=0.3193, pruned_loss=0.1175, over 1350534.24 frames. ], batch size: 41, lr: 3.86e-02, grad_scale: 16.0 +2023-03-20 18:11:38,853 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.922e+02 3.999e+02 4.896e+02 5.931e+02 1.258e+03, threshold=9.791e+02, percent-clipped=1.0 +2023-03-20 18:11:39,029 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6203.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:11:43,062 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6211.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:11:44,539 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6214.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:11:47,107 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-20 18:11:53,928 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6231.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:11:55,899 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-20 18:11:57,308 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.02 vs. limit=2.0 +2023-03-20 18:11:59,471 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-20 18:12:03,024 INFO [train.py:901] (0/2) Epoch 3, batch 600, loss[loss=0.269, simple_loss=0.3194, pruned_loss=0.1093, over 7256.00 frames. ], tot_loss[loss=0.2767, simple_loss=0.3192, pruned_loss=0.1171, over 1369324.02 frames. ], batch size: 47, lr: 3.85e-02, grad_scale: 16.0 +2023-03-20 18:12:06,207 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-20 18:12:10,976 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6264.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:12:11,953 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6266.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:12:15,029 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6272.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:12:20,737 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.76 vs. limit=5.0 +2023-03-20 18:12:22,510 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-20 18:12:27,380 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.94 vs. limit=2.0 +2023-03-20 18:12:29,110 INFO [train.py:901] (0/2) Epoch 3, batch 650, loss[loss=0.2623, simple_loss=0.3046, pruned_loss=0.11, over 7336.00 frames. ], tot_loss[loss=0.277, simple_loss=0.3192, pruned_loss=0.1174, over 1385267.84 frames. ], batch size: 54, lr: 3.84e-02, grad_scale: 16.0 +2023-03-20 18:12:31,154 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.321e+02 3.541e+02 4.453e+02 5.741e+02 1.629e+03, threshold=8.906e+02, percent-clipped=5.0 +2023-03-20 18:12:31,692 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-20 18:12:32,775 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6306.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 18:12:33,464 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.05 vs. limit=2.0 +2023-03-20 18:12:44,082 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6327.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:12:48,024 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-20 18:12:49,292 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.04 vs. limit=2.0 +2023-03-20 18:12:53,589 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6346.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:12:55,025 INFO [train.py:901] (0/2) Epoch 3, batch 700, loss[loss=0.2905, simple_loss=0.3302, pruned_loss=0.1254, over 7333.00 frames. ], tot_loss[loss=0.2779, simple_loss=0.3197, pruned_loss=0.118, over 1398978.41 frames. ], batch size: 49, lr: 3.83e-02, grad_scale: 16.0 +2023-03-20 18:12:56,550 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-20 18:12:57,606 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=6354.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 18:13:20,815 INFO [train.py:901] (0/2) Epoch 3, batch 750, loss[loss=0.3129, simple_loss=0.3498, pruned_loss=0.138, over 7299.00 frames. ], tot_loss[loss=0.2776, simple_loss=0.3192, pruned_loss=0.118, over 1409688.79 frames. ], batch size: 68, lr: 3.82e-02, grad_scale: 16.0 +2023-03-20 18:13:21,514 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-20 18:13:21,565 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6399.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:13:22,346 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-20 18:13:22,444 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3291, 4.1901, 3.1848, 3.6370, 3.4982, 3.7437, 3.7217, 3.4155], + device='cuda:0'), covar=tensor([0.0129, 0.0084, 0.0171, 0.0178, 0.0159, 0.0114, 0.0158, 0.0150], + device='cuda:0'), in_proj_covar=tensor([0.0018, 0.0020, 0.0020, 0.0019, 0.0018, 0.0019, 0.0023, 0.0021], + device='cuda:0'), out_proj_covar=tensor([3.2410e-05, 4.1555e-05, 3.9469e-05, 3.8501e-05, 3.5835e-05, 3.4702e-05, + 4.3640e-05, 4.1294e-05], device='cuda:0') +2023-03-20 18:13:23,906 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.231e+02 3.739e+02 4.769e+02 6.356e+02 1.531e+03, threshold=9.539e+02, percent-clipped=9.0 +2023-03-20 18:13:28,116 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6411.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:13:36,772 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-20 18:13:40,971 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6857, 2.5089, 2.2380, 2.1295, 1.4771, 2.4688, 2.5623, 2.1214], + device='cuda:0'), covar=tensor([0.0715, 0.0623, 0.0603, 0.0731, 0.1161, 0.0510, 0.0777, 0.0669], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0035, 0.0035, 0.0040, 0.0035, 0.0040, 0.0041, 0.0035], + device='cuda:0'), out_proj_covar=tensor([5.4280e-05, 4.4688e-05, 4.5046e-05, 4.8203e-05, 4.5034e-05, 5.1085e-05, + 5.3930e-05, 4.6148e-05], device='cuda:0') +2023-03-20 18:13:41,361 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-20 18:13:48,033 INFO [train.py:901] (0/2) Epoch 3, batch 800, loss[loss=0.2865, simple_loss=0.3204, pruned_loss=0.1263, over 7205.00 frames. ], tot_loss[loss=0.2766, simple_loss=0.3185, pruned_loss=0.1174, over 1415705.06 frames. ], batch size: 50, lr: 3.81e-02, grad_scale: 16.0 +2023-03-20 18:13:48,058 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-20 18:13:49,642 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-20 18:13:53,257 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=6459.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:13:54,846 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6462.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:13:57,277 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6467.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:13:59,734 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-20 18:14:05,741 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.15 vs. limit=5.0 +2023-03-20 18:14:10,183 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6491.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:14:14,097 INFO [train.py:901] (0/2) Epoch 3, batch 850, loss[loss=0.2878, simple_loss=0.3236, pruned_loss=0.126, over 7316.00 frames. ], tot_loss[loss=0.2768, simple_loss=0.3185, pruned_loss=0.1176, over 1418374.36 frames. ], batch size: 49, lr: 3.80e-02, grad_scale: 16.0 +2023-03-20 18:14:16,099 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.432e+02 3.714e+02 4.604e+02 5.774e+02 1.158e+03, threshold=9.207e+02, percent-clipped=4.0 +2023-03-20 18:14:18,177 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-20 18:14:18,660 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-20 18:14:21,876 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6514.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:14:23,666 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.14 vs. limit=2.0 +2023-03-20 18:14:23,821 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-20 18:14:26,500 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6523.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 18:14:27,336 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-20 18:14:27,931 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6526.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:14:29,020 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6528.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:14:40,311 INFO [train.py:901] (0/2) Epoch 3, batch 900, loss[loss=0.2633, simple_loss=0.3055, pruned_loss=0.1105, over 7232.00 frames. ], tot_loss[loss=0.2761, simple_loss=0.3182, pruned_loss=0.117, over 1423712.42 frames. ], batch size: 45, lr: 3.79e-02, grad_scale: 16.0 +2023-03-20 18:14:45,524 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6559.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:14:47,054 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=6562.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:14:47,586 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4798, 4.5172, 4.4006, 4.6984, 4.9450, 4.8195, 4.2199, 4.1423], + device='cuda:0'), covar=tensor([0.0539, 0.1136, 0.1496, 0.1552, 0.0408, 0.1088, 0.0718, 0.0786], + device='cuda:0'), in_proj_covar=tensor([0.0081, 0.0133, 0.0145, 0.0119, 0.0091, 0.0132, 0.0080, 0.0099], + device='cuda:0'), out_proj_covar=tensor([8.7053e-05, 1.4842e-04, 1.6784e-04, 1.4426e-04, 9.8969e-05, 1.5484e-04, + 8.6150e-05, 1.0319e-04], device='cuda:0') +2023-03-20 18:14:49,689 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6567.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:15:06,114 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-20 18:15:06,586 INFO [train.py:901] (0/2) Epoch 3, batch 950, loss[loss=0.2636, simple_loss=0.3151, pruned_loss=0.106, over 7327.00 frames. ], tot_loss[loss=0.275, simple_loss=0.3175, pruned_loss=0.1162, over 1425955.12 frames. ], batch size: 59, lr: 3.78e-02, grad_scale: 16.0 +2023-03-20 18:15:08,980 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.940e+02 3.725e+02 4.764e+02 5.478e+02 1.056e+03, threshold=9.528e+02, percent-clipped=3.0 +2023-03-20 18:15:19,393 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6622.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:15:23,058 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6629.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:15:30,035 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-20 18:15:31,650 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6646.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:15:33,031 INFO [train.py:901] (0/2) Epoch 3, batch 1000, loss[loss=0.2946, simple_loss=0.3297, pruned_loss=0.1298, over 7298.00 frames. ], tot_loss[loss=0.2747, simple_loss=0.3176, pruned_loss=0.1159, over 1430732.85 frames. ], batch size: 49, lr: 3.78e-02, grad_scale: 16.0 +2023-03-20 18:15:39,923 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.89 vs. limit=2.0 +2023-03-20 18:15:49,714 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-20 18:15:54,395 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6690.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:15:56,338 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=6694.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:15:58,865 INFO [train.py:901] (0/2) Epoch 3, batch 1050, loss[loss=0.254, simple_loss=0.313, pruned_loss=0.09747, over 7213.00 frames. ], tot_loss[loss=0.2757, simple_loss=0.3185, pruned_loss=0.1164, over 1433320.00 frames. ], batch size: 50, lr: 3.77e-02, grad_scale: 16.0 +2023-03-20 18:15:58,974 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6699.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:16:01,470 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.188e+02 3.614e+02 4.562e+02 6.271e+02 1.467e+03, threshold=9.125e+02, percent-clipped=4.0 +2023-03-20 18:16:11,095 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-20 18:16:15,165 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-20 18:16:24,533 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=6747.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:16:25,489 INFO [train.py:901] (0/2) Epoch 3, batch 1100, loss[loss=0.2846, simple_loss=0.3171, pruned_loss=0.126, over 7264.00 frames. ], tot_loss[loss=0.2748, simple_loss=0.3179, pruned_loss=0.1158, over 1435971.25 frames. ], batch size: 52, lr: 3.76e-02, grad_scale: 16.0 +2023-03-20 18:16:26,188 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9116, 2.9148, 2.1183, 3.0637, 2.5715, 2.8293, 2.3344, 1.8968], + device='cuda:0'), covar=tensor([0.0102, 0.0196, 0.0925, 0.0166, 0.0065, 0.0248, 0.1098, 0.0988], + device='cuda:0'), in_proj_covar=tensor([0.0056, 0.0063, 0.0121, 0.0062, 0.0055, 0.0057, 0.0135, 0.0120], + device='cuda:0'), out_proj_covar=tensor([5.1389e-05, 5.9056e-05, 1.0902e-04, 5.7791e-05, 5.0900e-05, 5.1461e-05, + 1.2395e-04, 1.0617e-04], device='cuda:0') +2023-03-20 18:16:38,139 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7355, 4.8515, 4.7260, 5.0459, 5.0992, 5.1095, 4.7351, 4.3998], + device='cuda:0'), covar=tensor([0.0447, 0.0926, 0.1497, 0.1081, 0.0419, 0.0861, 0.0391, 0.0673], + device='cuda:0'), in_proj_covar=tensor([0.0081, 0.0139, 0.0146, 0.0120, 0.0093, 0.0131, 0.0078, 0.0097], + device='cuda:0'), out_proj_covar=tensor([8.7555e-05, 1.5655e-04, 1.6726e-04, 1.4370e-04, 1.0028e-04, 1.5482e-04, + 8.3782e-05, 1.0290e-04], device='cuda:0') +2023-03-20 18:16:42,401 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8319, 2.1057, 2.3390, 2.4640, 2.9206, 3.2053, 2.3246, 2.7022], + device='cuda:0'), covar=tensor([0.1326, 0.0720, 0.1898, 0.0240, 0.0057, 0.0058, 0.0083, 0.0138], + device='cuda:0'), in_proj_covar=tensor([0.0163, 0.0115, 0.0157, 0.0072, 0.0061, 0.0068, 0.0063, 0.0064], + device='cuda:0'), out_proj_covar=tensor([1.2874e-04, 9.4579e-05, 1.2960e-04, 6.4380e-05, 4.9157e-05, 5.3784e-05, + 5.1342e-05, 4.9745e-05], device='cuda:0') +2023-03-20 18:16:43,753 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-20 18:16:44,267 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 18:16:47,401 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6791.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:16:51,365 INFO [train.py:901] (0/2) Epoch 3, batch 1150, loss[loss=0.2453, simple_loss=0.3036, pruned_loss=0.09354, over 7346.00 frames. ], tot_loss[loss=0.2734, simple_loss=0.3167, pruned_loss=0.115, over 1438735.15 frames. ], batch size: 61, lr: 3.75e-02, grad_scale: 16.0 +2023-03-20 18:16:53,754 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.995e+02 3.595e+02 4.369e+02 5.397e+02 1.197e+03, threshold=8.738e+02, percent-clipped=1.0 +2023-03-20 18:16:56,489 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8405, 3.0849, 3.0218, 3.1645, 2.8123, 2.8682, 2.8021, 2.7159], + device='cuda:0'), covar=tensor([0.0136, 0.0152, 0.0156, 0.0146, 0.0211, 0.0171, 0.0216, 0.0241], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0022, 0.0020, 0.0020, 0.0020, 0.0022, 0.0025, 0.0023], + device='cuda:0'), out_proj_covar=tensor([3.6912e-05, 4.8020e-05, 4.0327e-05, 4.1070e-05, 4.3405e-05, 4.2489e-05, + 5.0253e-05, 4.7514e-05], device='cuda:0') +2023-03-20 18:16:57,422 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-20 18:16:57,902 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-20 18:16:58,159 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.84 vs. limit=2.0 +2023-03-20 18:17:01,549 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6818.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 18:17:03,971 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6823.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:17:05,522 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6826.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:17:12,659 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=6839.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:17:17,597 INFO [train.py:901] (0/2) Epoch 3, batch 1200, loss[loss=0.2869, simple_loss=0.3281, pruned_loss=0.1229, over 7242.00 frames. ], tot_loss[loss=0.273, simple_loss=0.3159, pruned_loss=0.1151, over 1439117.53 frames. ], batch size: 55, lr: 3.74e-02, grad_scale: 16.0 +2023-03-20 18:17:22,745 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6859.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:17:27,388 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6867.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:17:30,863 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=6874.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:17:31,345 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-20 18:17:33,001 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6878.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 18:17:43,504 INFO [train.py:901] (0/2) Epoch 3, batch 1250, loss[loss=0.2834, simple_loss=0.33, pruned_loss=0.1184, over 7284.00 frames. ], tot_loss[loss=0.2734, simple_loss=0.3165, pruned_loss=0.1151, over 1440397.90 frames. ], batch size: 66, lr: 3.73e-02, grad_scale: 16.0 +2023-03-20 18:17:45,503 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.075e+02 4.156e+02 4.876e+02 5.701e+02 1.197e+03, threshold=9.752e+02, percent-clipped=5.0 +2023-03-20 18:17:46,388 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.95 vs. limit=2.0 +2023-03-20 18:17:47,644 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=6907.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:17:48,770 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=6909.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:17:51,818 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=6915.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:17:55,496 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-20 18:17:56,080 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=6922.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:17:59,549 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-20 18:18:01,063 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-20 18:18:04,757 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6939.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 18:18:07,798 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5179, 2.7167, 2.1909, 3.0140, 2.5787, 2.8503, 2.2229, 1.9152], + device='cuda:0'), covar=tensor([0.0039, 0.0153, 0.0667, 0.0159, 0.0100, 0.0078, 0.0692, 0.0670], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0064, 0.0120, 0.0066, 0.0060, 0.0056, 0.0133, 0.0123], + device='cuda:0'), out_proj_covar=tensor([4.8835e-05, 6.1466e-05, 1.0993e-04, 6.1279e-05, 5.5565e-05, 5.1272e-05, + 1.2428e-04, 1.1058e-04], device='cuda:0') +2023-03-20 18:18:09,653 INFO [train.py:901] (0/2) Epoch 3, batch 1300, loss[loss=0.2717, simple_loss=0.317, pruned_loss=0.1132, over 7273.00 frames. ], tot_loss[loss=0.2727, simple_loss=0.3163, pruned_loss=0.1145, over 1442012.56 frames. ], batch size: 77, lr: 3.72e-02, grad_scale: 16.0 +2023-03-20 18:18:20,965 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=6970.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:18:21,068 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=6970.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:18:24,483 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-20 18:18:26,566 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-20 18:18:28,632 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=6985.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:18:29,613 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-20 18:18:35,688 INFO [train.py:901] (0/2) Epoch 3, batch 1350, loss[loss=0.2887, simple_loss=0.3198, pruned_loss=0.1288, over 7306.00 frames. ], tot_loss[loss=0.2722, simple_loss=0.3164, pruned_loss=0.114, over 1443852.25 frames. ], batch size: 86, lr: 3.71e-02, grad_scale: 16.0 +2023-03-20 18:18:37,958 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.522e+02 3.667e+02 4.621e+02 5.823e+02 1.455e+03, threshold=9.242e+02, percent-clipped=1.0 +2023-03-20 18:18:40,704 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-20 18:18:45,550 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=8.09 vs. limit=5.0 +2023-03-20 18:19:02,189 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=7048.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 18:19:02,503 INFO [train.py:901] (0/2) Epoch 3, batch 1400, loss[loss=0.4017, simple_loss=0.4031, pruned_loss=0.2001, over 6802.00 frames. ], tot_loss[loss=0.2734, simple_loss=0.3168, pruned_loss=0.115, over 1442401.30 frames. ], batch size: 107, lr: 3.70e-02, grad_scale: 16.0 +2023-03-20 18:19:07,644 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5012, 4.5275, 4.4851, 4.7618, 4.9962, 4.7687, 4.4874, 4.1594], + device='cuda:0'), covar=tensor([0.0615, 0.1250, 0.1761, 0.1185, 0.0403, 0.1275, 0.0563, 0.0915], + device='cuda:0'), in_proj_covar=tensor([0.0084, 0.0144, 0.0151, 0.0126, 0.0099, 0.0136, 0.0084, 0.0103], + device='cuda:0'), out_proj_covar=tensor([8.9834e-05, 1.6033e-04, 1.7408e-04, 1.5155e-04, 1.0505e-04, 1.6105e-04, + 9.0431e-05, 1.0814e-04], device='cuda:0') +2023-03-20 18:19:13,215 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-20 18:19:24,873 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.08 vs. limit=2.0 +2023-03-20 18:19:28,592 INFO [train.py:901] (0/2) Epoch 3, batch 1450, loss[loss=0.2388, simple_loss=0.2928, pruned_loss=0.09245, over 7344.00 frames. ], tot_loss[loss=0.2718, simple_loss=0.3159, pruned_loss=0.1139, over 1443000.12 frames. ], batch size: 63, lr: 3.70e-02, grad_scale: 16.0 +2023-03-20 18:19:30,653 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.276e+02 3.678e+02 4.650e+02 5.818e+02 1.874e+03, threshold=9.299e+02, percent-clipped=3.0 +2023-03-20 18:19:33,890 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=7109.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 18:19:38,429 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=7118.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 18:19:38,808 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-20 18:19:40,899 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=7123.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:19:45,501 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9721, 2.6598, 2.1602, 2.9353, 1.6035, 2.8500, 2.4655, 2.8148], + device='cuda:0'), covar=tensor([0.0097, 0.0483, 0.3191, 0.0086, 0.6855, 0.0090, 0.0756, 0.0121], + device='cuda:0'), in_proj_covar=tensor([0.0094, 0.0133, 0.0258, 0.0091, 0.0234, 0.0093, 0.0159, 0.0097], + device='cuda:0'), out_proj_covar=tensor([7.2664e-05, 1.0971e-04, 1.9034e-04, 7.4114e-05, 1.8282e-04, 7.3112e-05, + 1.2841e-04, 7.3765e-05], device='cuda:0') +2023-03-20 18:19:54,142 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-20 18:19:54,635 INFO [train.py:901] (0/2) Epoch 3, batch 1500, loss[loss=0.2526, simple_loss=0.3056, pruned_loss=0.09982, over 7290.00 frames. ], tot_loss[loss=0.272, simple_loss=0.316, pruned_loss=0.1141, over 1442753.83 frames. ], batch size: 66, lr: 3.69e-02, grad_scale: 16.0 +2023-03-20 18:19:55,216 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5958, 4.5633, 4.7455, 4.7469, 4.4910, 4.2626, 4.8053, 4.6772], + device='cuda:0'), covar=tensor([0.0272, 0.0241, 0.0228, 0.0487, 0.0363, 0.0291, 0.0238, 0.0329], + device='cuda:0'), in_proj_covar=tensor([0.0072, 0.0058, 0.0073, 0.0065, 0.0056, 0.0066, 0.0060, 0.0057], + device='cuda:0'), out_proj_covar=tensor([1.0098e-04, 7.7199e-05, 1.0356e-04, 9.6996e-05, 7.9234e-05, 9.4593e-05, + 8.4288e-05, 8.0807e-05], device='cuda:0') +2023-03-20 18:20:01,838 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=7163.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:20:03,300 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=7166.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:20:06,374 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=7171.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:20:10,343 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9833, 4.8152, 4.8683, 5.2202, 5.3295, 5.3608, 4.9109, 4.4984], + device='cuda:0'), covar=tensor([0.0522, 0.1477, 0.1571, 0.1043, 0.0348, 0.0832, 0.0524, 0.0947], + device='cuda:0'), in_proj_covar=tensor([0.0084, 0.0148, 0.0154, 0.0128, 0.0102, 0.0139, 0.0088, 0.0108], + device='cuda:0'), out_proj_covar=tensor([9.0678e-05, 1.6574e-04, 1.7802e-04, 1.5385e-04, 1.0741e-04, 1.6296e-04, + 9.3604e-05, 1.1399e-04], device='cuda:0') +2023-03-20 18:20:17,889 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-20 18:20:20,405 INFO [train.py:901] (0/2) Epoch 3, batch 1550, loss[loss=0.2193, simple_loss=0.2619, pruned_loss=0.08832, over 7017.00 frames. ], tot_loss[loss=0.2726, simple_loss=0.3163, pruned_loss=0.1144, over 1442286.70 frames. ], batch size: 35, lr: 3.68e-02, grad_scale: 16.0 +2023-03-20 18:20:22,741 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.058e+02 3.898e+02 4.822e+02 6.502e+02 1.544e+03, threshold=9.644e+02, percent-clipped=7.0 +2023-03-20 18:20:34,282 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6116, 2.8763, 1.9176, 2.4160, 2.2947, 1.9487, 1.8858, 2.6573], + device='cuda:0'), covar=tensor([0.0689, 0.0150, 0.0818, 0.0708, 0.0670, 0.0892, 0.0857, 0.0451], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0022, 0.0025, 0.0025, 0.0025, 0.0026, 0.0028, 0.0024], + device='cuda:0'), out_proj_covar=tensor([3.6644e-05, 3.0348e-05, 3.8530e-05, 3.7922e-05, 3.7736e-05, 3.8662e-05, + 4.0775e-05, 3.5467e-05], device='cuda:0') +2023-03-20 18:20:34,312 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=7224.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 18:20:39,254 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=7234.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 18:20:46,676 INFO [train.py:901] (0/2) Epoch 3, batch 1600, loss[loss=0.2902, simple_loss=0.3329, pruned_loss=0.1238, over 7274.00 frames. ], tot_loss[loss=0.2723, simple_loss=0.3162, pruned_loss=0.1142, over 1442372.64 frames. ], batch size: 89, lr: 3.67e-02, grad_scale: 16.0 +2023-03-20 18:20:50,254 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-20 18:20:50,770 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-20 18:20:52,062 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.07 vs. limit=2.0 +2023-03-20 18:20:53,814 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-20 18:20:55,479 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=7265.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:21:00,613 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=7275.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:21:04,605 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-20 18:21:05,441 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-20 18:21:05,721 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=7285.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:21:08,696 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-20 18:21:13,315 INFO [train.py:901] (0/2) Epoch 3, batch 1650, loss[loss=0.2783, simple_loss=0.3124, pruned_loss=0.1222, over 7338.00 frames. ], tot_loss[loss=0.2727, simple_loss=0.316, pruned_loss=0.1147, over 1442908.08 frames. ], batch size: 61, lr: 3.66e-02, grad_scale: 16.0 +2023-03-20 18:21:15,324 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.873e+02 4.152e+02 5.169e+02 6.417e+02 1.445e+03, threshold=1.034e+03, percent-clipped=2.0 +2023-03-20 18:21:17,401 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-20 18:21:20,474 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4956, 4.5502, 4.5687, 4.7796, 4.9494, 4.9272, 4.1102, 4.3158], + device='cuda:0'), covar=tensor([0.0529, 0.0925, 0.1311, 0.0925, 0.0394, 0.0825, 0.0581, 0.0709], + device='cuda:0'), in_proj_covar=tensor([0.0086, 0.0147, 0.0153, 0.0124, 0.0103, 0.0138, 0.0083, 0.0104], + device='cuda:0'), out_proj_covar=tensor([9.2461e-05, 1.6468e-04, 1.7400e-04, 1.5011e-04, 1.0921e-04, 1.6219e-04, + 8.9282e-05, 1.0896e-04], device='cuda:0') +2023-03-20 18:21:30,691 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=7333.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:21:32,336 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=7336.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:21:35,790 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 18:21:38,776 INFO [train.py:901] (0/2) Epoch 3, batch 1700, loss[loss=0.2637, simple_loss=0.3078, pruned_loss=0.1098, over 7283.00 frames. ], tot_loss[loss=0.2715, simple_loss=0.3149, pruned_loss=0.114, over 1443301.31 frames. ], batch size: 66, lr: 3.65e-02, grad_scale: 16.0 +2023-03-20 18:21:39,796 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-20 18:21:51,203 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-20 18:22:05,303 INFO [train.py:901] (0/2) Epoch 3, batch 1750, loss[loss=0.2663, simple_loss=0.3067, pruned_loss=0.1129, over 7266.00 frames. ], tot_loss[loss=0.2706, simple_loss=0.3142, pruned_loss=0.1135, over 1442587.69 frames. ], batch size: 47, lr: 3.64e-02, grad_scale: 16.0 +2023-03-20 18:22:06,980 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.02 vs. limit=2.0 +2023-03-20 18:22:07,663 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.208e+02 3.884e+02 4.725e+02 5.852e+02 1.126e+03, threshold=9.450e+02, percent-clipped=2.0 +2023-03-20 18:22:08,225 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=7404.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 18:22:09,789 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=7407.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:22:16,241 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-20 18:22:17,245 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-20 18:22:31,633 INFO [train.py:901] (0/2) Epoch 3, batch 1800, loss[loss=0.2128, simple_loss=0.2695, pruned_loss=0.07799, over 7218.00 frames. ], tot_loss[loss=0.2683, simple_loss=0.3124, pruned_loss=0.1121, over 1442540.94 frames. ], batch size: 45, lr: 3.64e-02, grad_scale: 32.0 +2023-03-20 18:22:38,306 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1107, 3.7006, 3.4508, 3.5963, 3.3222, 3.2009, 3.4148, 3.0835], + device='cuda:0'), covar=tensor([0.0132, 0.0119, 0.0126, 0.0117, 0.0175, 0.0173, 0.0152, 0.0199], + device='cuda:0'), in_proj_covar=tensor([0.0022, 0.0022, 0.0021, 0.0021, 0.0020, 0.0022, 0.0025, 0.0023], + device='cuda:0'), out_proj_covar=tensor([4.6693e-05, 5.1827e-05, 4.6760e-05, 4.6930e-05, 4.6852e-05, 4.5877e-05, + 5.6227e-05, 5.0100e-05], device='cuda:0') +2023-03-20 18:22:40,207 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-20 18:22:41,054 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.04 vs. limit=2.0 +2023-03-20 18:22:41,385 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=7468.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:22:52,879 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-20 18:22:53,674 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-03-20 18:22:54,958 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1314, 3.5862, 3.1873, 2.9435, 3.4099, 2.8643, 1.2693, 3.6328], + device='cuda:0'), covar=tensor([0.0055, 0.0052, 0.0114, 0.0111, 0.0048, 0.0283, 0.1475, 0.0089], + device='cuda:0'), in_proj_covar=tensor([0.0032, 0.0033, 0.0041, 0.0034, 0.0036, 0.0047, 0.0069, 0.0038], + device='cuda:0'), out_proj_covar=tensor([3.7259e-05, 4.4999e-05, 5.1283e-05, 4.5647e-05, 4.0095e-05, 6.2537e-05, + 1.0100e-04, 4.7933e-05], device='cuda:0') +2023-03-20 18:22:57,429 INFO [train.py:901] (0/2) Epoch 3, batch 1850, loss[loss=0.2699, simple_loss=0.3172, pruned_loss=0.1113, over 7315.00 frames. ], tot_loss[loss=0.2682, simple_loss=0.3123, pruned_loss=0.112, over 1442432.08 frames. ], batch size: 83, lr: 3.63e-02, grad_scale: 32.0 +2023-03-20 18:22:59,427 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.848e+02 3.617e+02 4.496e+02 5.652e+02 1.259e+03, threshold=8.993e+02, percent-clipped=2.0 +2023-03-20 18:23:02,448 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-20 18:23:07,608 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=7519.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 18:23:15,734 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=7534.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 18:23:19,697 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-20 18:23:22,243 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.2913, 5.0925, 5.1838, 5.4084, 5.6213, 5.6229, 4.8863, 4.8459], + device='cuda:0'), covar=tensor([0.0394, 0.1167, 0.1629, 0.1084, 0.0298, 0.0671, 0.0539, 0.0668], + device='cuda:0'), in_proj_covar=tensor([0.0085, 0.0147, 0.0150, 0.0120, 0.0100, 0.0139, 0.0083, 0.0100], + device='cuda:0'), out_proj_covar=tensor([9.2514e-05, 1.6587e-04, 1.7259e-04, 1.4614e-04, 1.0674e-04, 1.6345e-04, + 9.0201e-05, 1.0491e-04], device='cuda:0') +2023-03-20 18:23:23,226 INFO [train.py:901] (0/2) Epoch 3, batch 1900, loss[loss=0.2595, simple_loss=0.3024, pruned_loss=0.1083, over 7201.00 frames. ], tot_loss[loss=0.2693, simple_loss=0.3133, pruned_loss=0.1127, over 1443015.39 frames. ], batch size: 39, lr: 3.62e-02, grad_scale: 32.0 +2023-03-20 18:23:31,912 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=7565.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:23:40,264 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=7582.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 18:23:44,711 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-20 18:23:46,399 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.8989, 0.9019, 0.9184, 0.5860, 1.7698, 0.6026, 0.8070, 0.7302], + device='cuda:0'), covar=tensor([0.0415, 0.0219, 0.0796, 0.0733, 0.0141, 0.0670, 0.0451, 0.0411], + device='cuda:0'), in_proj_covar=tensor([0.0024, 0.0024, 0.0022, 0.0023, 0.0022, 0.0025, 0.0024, 0.0023], + device='cuda:0'), out_proj_covar=tensor([2.6448e-05, 2.3512e-05, 2.5274e-05, 2.7218e-05, 2.1547e-05, 2.8235e-05, + 3.0650e-05, 2.6927e-05], device='cuda:0') +2023-03-20 18:23:48,793 INFO [train.py:901] (0/2) Epoch 3, batch 1950, loss[loss=0.2272, simple_loss=0.282, pruned_loss=0.08616, over 7354.00 frames. ], tot_loss[loss=0.2705, simple_loss=0.3144, pruned_loss=0.1133, over 1442581.61 frames. ], batch size: 44, lr: 3.61e-02, grad_scale: 32.0 +2023-03-20 18:23:51,174 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.297e+02 4.106e+02 4.807e+02 6.855e+02 1.279e+03, threshold=9.614e+02, percent-clipped=5.0 +2023-03-20 18:23:56,340 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-20 18:23:56,376 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=7613.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:24:02,039 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-20 18:24:02,555 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-20 18:24:02,648 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=7624.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:24:06,096 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=7631.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:24:09,402 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.63 vs. limit=5.0 +2023-03-20 18:24:15,188 INFO [train.py:901] (0/2) Epoch 3, batch 2000, loss[loss=0.2536, simple_loss=0.2969, pruned_loss=0.1051, over 7262.00 frames. ], tot_loss[loss=0.2697, simple_loss=0.3136, pruned_loss=0.1129, over 1441490.15 frames. ], batch size: 64, lr: 3.60e-02, grad_scale: 32.0 +2023-03-20 18:24:15,553 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.13 vs. limit=2.0 +2023-03-20 18:24:20,424 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-20 18:24:23,624 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7681, 2.1179, 1.7766, 1.6770, 1.7624, 1.5136, 2.1310, 2.0420], + device='cuda:0'), covar=tensor([0.0837, 0.0478, 0.0870, 0.0588, 0.0687, 0.0783, 0.0406, 0.0453], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0034, 0.0038, 0.0042, 0.0033, 0.0038, 0.0041, 0.0036], + device='cuda:0'), out_proj_covar=tensor([6.3710e-05, 4.9290e-05, 5.6803e-05, 5.9199e-05, 4.9619e-05, 5.5855e-05, + 6.3234e-05, 5.4026e-05], device='cuda:0') +2023-03-20 18:24:31,537 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-20 18:24:34,174 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=7685.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:24:39,194 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-20 18:24:41,213 INFO [train.py:901] (0/2) Epoch 3, batch 2050, loss[loss=0.2686, simple_loss=0.3202, pruned_loss=0.1085, over 7213.00 frames. ], tot_loss[loss=0.2686, simple_loss=0.3131, pruned_loss=0.112, over 1440481.83 frames. ], batch size: 93, lr: 3.59e-02, grad_scale: 32.0 +2023-03-20 18:24:43,217 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.099e+02 3.470e+02 4.322e+02 5.464e+02 1.086e+03, threshold=8.645e+02, percent-clipped=2.0 +2023-03-20 18:24:43,818 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=7704.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 18:25:04,841 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=7743.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:25:07,331 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0605, 4.2260, 4.0927, 4.4021, 4.6344, 4.5140, 3.9030, 3.9273], + device='cuda:0'), covar=tensor([0.0698, 0.1298, 0.1720, 0.1520, 0.0576, 0.1299, 0.0886, 0.0969], + device='cuda:0'), in_proj_covar=tensor([0.0086, 0.0148, 0.0148, 0.0126, 0.0100, 0.0142, 0.0085, 0.0104], + device='cuda:0'), out_proj_covar=tensor([9.5035e-05, 1.6748e-04, 1.6826e-04, 1.5141e-04, 1.0677e-04, 1.6786e-04, + 9.3975e-05, 1.0859e-04], device='cuda:0') +2023-03-20 18:25:07,724 INFO [train.py:901] (0/2) Epoch 3, batch 2100, loss[loss=0.2876, simple_loss=0.3244, pruned_loss=0.1254, over 7315.00 frames. ], tot_loss[loss=0.2672, simple_loss=0.3118, pruned_loss=0.1113, over 1441409.30 frames. ], batch size: 80, lr: 3.59e-02, grad_scale: 16.0 +2023-03-20 18:25:09,335 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=7752.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 18:25:14,209 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-20 18:25:14,818 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=7763.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:25:17,239 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-20 18:25:33,571 INFO [train.py:901] (0/2) Epoch 3, batch 2150, loss[loss=0.287, simple_loss=0.3283, pruned_loss=0.1229, over 7351.00 frames. ], tot_loss[loss=0.268, simple_loss=0.3128, pruned_loss=0.1116, over 1444149.27 frames. ], batch size: 63, lr: 3.58e-02, grad_scale: 16.0 +2023-03-20 18:25:35,548 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=7802.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:25:36,406 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.942e+02 3.904e+02 5.108e+02 6.529e+02 1.416e+03, threshold=1.022e+03, percent-clipped=8.0 +2023-03-20 18:25:36,595 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=7804.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:25:42,574 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2145, 4.3407, 4.2742, 4.7362, 4.8276, 4.7719, 4.3730, 4.1064], + device='cuda:0'), covar=tensor([0.0595, 0.1231, 0.1552, 0.1117, 0.0437, 0.0952, 0.0578, 0.0831], + device='cuda:0'), in_proj_covar=tensor([0.0084, 0.0142, 0.0143, 0.0119, 0.0098, 0.0137, 0.0082, 0.0099], + device='cuda:0'), out_proj_covar=tensor([9.2664e-05, 1.6192e-04, 1.6344e-04, 1.4350e-04, 1.0423e-04, 1.6192e-04, + 9.0959e-05, 1.0412e-04], device='cuda:0') +2023-03-20 18:25:44,180 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=7819.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 18:25:59,744 INFO [train.py:901] (0/2) Epoch 3, batch 2200, loss[loss=0.2462, simple_loss=0.299, pruned_loss=0.09664, over 7278.00 frames. ], tot_loss[loss=0.2689, simple_loss=0.3135, pruned_loss=0.1121, over 1444408.74 frames. ], batch size: 70, lr: 3.57e-02, grad_scale: 16.0 +2023-03-20 18:26:02,268 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-20 18:26:06,986 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=7863.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:26:08,898 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=7867.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:26:25,706 INFO [train.py:901] (0/2) Epoch 3, batch 2250, loss[loss=0.2651, simple_loss=0.3102, pruned_loss=0.11, over 7307.00 frames. ], tot_loss[loss=0.2684, simple_loss=0.3129, pruned_loss=0.1119, over 1442580.26 frames. ], batch size: 49, lr: 3.56e-02, grad_scale: 16.0 +2023-03-20 18:26:28,202 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.204e+02 3.580e+02 4.617e+02 6.250e+02 1.322e+03, threshold=9.234e+02, percent-clipped=2.0 +2023-03-20 18:26:34,530 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1732, 2.2876, 2.7404, 3.2264, 2.7019, 2.7830, 3.2403, 2.9121], + device='cuda:0'), covar=tensor([0.1356, 0.0814, 0.1583, 0.0228, 0.0068, 0.0188, 0.0103, 0.0107], + device='cuda:0'), in_proj_covar=tensor([0.0186, 0.0130, 0.0174, 0.0083, 0.0067, 0.0078, 0.0072, 0.0072], + device='cuda:0'), out_proj_covar=tensor([1.5112e-04, 1.1305e-04, 1.4600e-04, 7.7749e-05, 5.6543e-05, 6.3882e-05, + 6.3092e-05, 5.9429e-05], device='cuda:0') +2023-03-20 18:26:36,935 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-20 18:26:37,411 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-20 18:26:42,560 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=7931.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:26:49,637 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0439, 2.3248, 2.4407, 3.1048, 2.5020, 2.7369, 2.9186, 2.5032], + device='cuda:0'), covar=tensor([0.1287, 0.0606, 0.1670, 0.0223, 0.0051, 0.0078, 0.0063, 0.0091], + device='cuda:0'), in_proj_covar=tensor([0.0187, 0.0129, 0.0175, 0.0084, 0.0067, 0.0080, 0.0071, 0.0072], + device='cuda:0'), out_proj_covar=tensor([1.5241e-04, 1.1252e-04, 1.4652e-04, 7.8350e-05, 5.6378e-05, 6.5505e-05, + 6.2479e-05, 5.9958e-05], device='cuda:0') +2023-03-20 18:26:49,996 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-20 18:26:51,462 INFO [train.py:901] (0/2) Epoch 3, batch 2300, loss[loss=0.278, simple_loss=0.3083, pruned_loss=0.1239, over 7191.00 frames. ], tot_loss[loss=0.2675, simple_loss=0.3118, pruned_loss=0.1116, over 1439338.13 frames. ], batch size: 39, lr: 3.55e-02, grad_scale: 16.0 +2023-03-20 18:27:07,550 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=7979.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:27:08,063 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=7980.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:27:17,727 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7232, 3.7075, 3.7356, 4.1000, 4.2879, 4.2028, 3.6849, 3.4330], + device='cuda:0'), covar=tensor([0.0754, 0.1535, 0.1911, 0.1258, 0.0507, 0.1133, 0.0776, 0.1206], + device='cuda:0'), in_proj_covar=tensor([0.0082, 0.0144, 0.0145, 0.0115, 0.0099, 0.0134, 0.0081, 0.0099], + device='cuda:0'), out_proj_covar=tensor([9.0054e-05, 1.6356e-04, 1.6351e-04, 1.4067e-04, 1.0495e-04, 1.5925e-04, + 8.9209e-05, 1.0299e-04], device='cuda:0') +2023-03-20 18:27:18,153 INFO [train.py:901] (0/2) Epoch 3, batch 2350, loss[loss=0.3076, simple_loss=0.3468, pruned_loss=0.1342, over 6717.00 frames. ], tot_loss[loss=0.2676, simple_loss=0.3122, pruned_loss=0.1115, over 1440374.23 frames. ], batch size: 106, lr: 3.55e-02, grad_scale: 16.0 +2023-03-20 18:27:19,007 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-8000.pt +2023-03-20 18:27:24,524 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.608e+02 3.747e+02 4.691e+02 6.358e+02 1.596e+03, threshold=9.383e+02, percent-clipped=3.0 +2023-03-20 18:27:38,065 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-20 18:27:46,306 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-20 18:27:47,817 INFO [train.py:901] (0/2) Epoch 3, batch 2400, loss[loss=0.2458, simple_loss=0.2826, pruned_loss=0.1045, over 6933.00 frames. ], tot_loss[loss=0.2674, simple_loss=0.3123, pruned_loss=0.1112, over 1441291.48 frames. ], batch size: 35, lr: 3.54e-02, grad_scale: 16.0 +2023-03-20 18:27:48,604 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.97 vs. limit=2.0 +2023-03-20 18:27:54,950 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=8063.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:27:56,932 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-20 18:27:59,531 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-20 18:28:13,981 INFO [train.py:901] (0/2) Epoch 3, batch 2450, loss[loss=0.2721, simple_loss=0.3179, pruned_loss=0.1131, over 7353.00 frames. ], tot_loss[loss=0.2667, simple_loss=0.3117, pruned_loss=0.1109, over 1443132.31 frames. ], batch size: 73, lr: 3.53e-02, grad_scale: 16.0 +2023-03-20 18:28:14,072 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=8099.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:28:16,492 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.331e+02 4.135e+02 4.933e+02 6.435e+02 1.136e+03, threshold=9.866e+02, percent-clipped=3.0 +2023-03-20 18:28:20,178 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=8111.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:28:26,232 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-20 18:28:40,172 INFO [train.py:901] (0/2) Epoch 3, batch 2500, loss[loss=0.2, simple_loss=0.256, pruned_loss=0.07198, over 7165.00 frames. ], tot_loss[loss=0.2662, simple_loss=0.3116, pruned_loss=0.1104, over 1442091.28 frames. ], batch size: 41, lr: 3.52e-02, grad_scale: 16.0 +2023-03-20 18:28:44,797 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=8158.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:28:53,964 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-20 18:28:55,264 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.43 vs. limit=2.0 +2023-03-20 18:29:06,266 INFO [train.py:901] (0/2) Epoch 3, batch 2550, loss[loss=0.2724, simple_loss=0.322, pruned_loss=0.1114, over 7244.00 frames. ], tot_loss[loss=0.2666, simple_loss=0.3119, pruned_loss=0.1107, over 1442302.97 frames. ], batch size: 47, lr: 3.51e-02, grad_scale: 16.0 +2023-03-20 18:29:09,137 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.629e+02 3.850e+02 5.119e+02 6.158e+02 1.209e+03, threshold=1.024e+03, percent-clipped=1.0 +2023-03-20 18:29:32,417 INFO [train.py:901] (0/2) Epoch 3, batch 2600, loss[loss=0.2284, simple_loss=0.277, pruned_loss=0.08991, over 7143.00 frames. ], tot_loss[loss=0.2657, simple_loss=0.3113, pruned_loss=0.1101, over 1444398.20 frames. ], batch size: 41, lr: 3.51e-02, grad_scale: 16.0 +2023-03-20 18:29:40,669 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=8265.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:29:48,076 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=8280.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:29:57,296 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.01 vs. limit=2.0 +2023-03-20 18:29:58,015 INFO [train.py:901] (0/2) Epoch 3, batch 2650, loss[loss=0.2579, simple_loss=0.3093, pruned_loss=0.1032, over 7351.00 frames. ], tot_loss[loss=0.2659, simple_loss=0.3114, pruned_loss=0.1101, over 1445380.48 frames. ], batch size: 63, lr: 3.50e-02, grad_scale: 16.0 +2023-03-20 18:30:00,552 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.150e+02 3.672e+02 4.487e+02 5.762e+02 1.165e+03, threshold=8.975e+02, percent-clipped=3.0 +2023-03-20 18:30:11,790 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=8326.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:30:12,695 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=8328.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:30:23,023 INFO [train.py:901] (0/2) Epoch 3, batch 2700, loss[loss=0.2301, simple_loss=0.2904, pruned_loss=0.08492, over 7287.00 frames. ], tot_loss[loss=0.2657, simple_loss=0.3111, pruned_loss=0.1101, over 1443693.72 frames. ], batch size: 68, lr: 3.49e-02, grad_scale: 16.0 +2023-03-20 18:30:31,781 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.02 vs. limit=2.0 +2023-03-20 18:30:32,749 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.05 vs. limit=2.0 +2023-03-20 18:30:48,012 INFO [train.py:901] (0/2) Epoch 3, batch 2750, loss[loss=0.31, simple_loss=0.3422, pruned_loss=0.1389, over 7271.00 frames. ], tot_loss[loss=0.265, simple_loss=0.3106, pruned_loss=0.1097, over 1442397.83 frames. ], batch size: 57, lr: 3.48e-02, grad_scale: 16.0 +2023-03-20 18:30:48,104 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=8399.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:30:50,821 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.974e+02 3.728e+02 5.076e+02 6.308e+02 1.197e+03, threshold=1.015e+03, percent-clipped=5.0 +2023-03-20 18:30:56,844 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3158, 4.3603, 4.2690, 4.5900, 4.7265, 4.7068, 3.9558, 4.2068], + device='cuda:0'), covar=tensor([0.0683, 0.1599, 0.1734, 0.1397, 0.0438, 0.1015, 0.0797, 0.0950], + device='cuda:0'), in_proj_covar=tensor([0.0090, 0.0156, 0.0160, 0.0129, 0.0107, 0.0147, 0.0091, 0.0114], + device='cuda:0'), out_proj_covar=tensor([9.7695e-05, 1.7780e-04, 1.8074e-04, 1.5397e-04, 1.1171e-04, 1.7420e-04, + 9.6477e-05, 1.1598e-04], device='cuda:0') +2023-03-20 18:31:12,356 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=8447.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:31:13,267 INFO [train.py:901] (0/2) Epoch 3, batch 2800, loss[loss=0.283, simple_loss=0.324, pruned_loss=0.121, over 7281.00 frames. ], tot_loss[loss=0.266, simple_loss=0.3112, pruned_loss=0.1104, over 1440992.47 frames. ], batch size: 64, lr: 3.48e-02, grad_scale: 16.0 +2023-03-20 18:31:16,424 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1047, 2.7393, 2.1839, 2.8213, 1.5960, 2.8540, 3.1675, 2.9935], + device='cuda:0'), covar=tensor([0.0135, 0.0538, 0.2602, 0.0098, 0.6017, 0.0084, 0.0409, 0.0076], + device='cuda:0'), in_proj_covar=tensor([0.0097, 0.0150, 0.0253, 0.0094, 0.0241, 0.0096, 0.0161, 0.0100], + device='cuda:0'), out_proj_covar=tensor([7.8339e-05, 1.2478e-04, 1.9236e-04, 8.0318e-05, 1.9231e-04, 7.7814e-05, + 1.3246e-04, 8.2701e-05], device='cuda:0') +2023-03-20 18:31:16,614 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.14 vs. limit=2.0 +2023-03-20 18:31:17,826 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=8458.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:31:25,918 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-3.pt +2023-03-20 18:31:44,451 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-20 18:31:44,506 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-20 18:31:44,532 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-20 18:31:44,551 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-20 18:31:44,728 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-20 18:31:44,744 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-20 18:31:44,755 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-20 18:31:44,858 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-20 18:31:44,933 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-20 18:31:44,949 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-20 18:31:44,970 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-20 18:31:45,534 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-20 18:31:45,539 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-20 18:31:45,547 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-20 18:31:45,687 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-20 18:31:45,690 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-20 18:31:45,749 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-20 18:31:48,089 INFO [train.py:901] (0/2) Epoch 4, batch 0, loss[loss=0.2557, simple_loss=0.3088, pruned_loss=0.1013, over 7278.00 frames. ], tot_loss[loss=0.2557, simple_loss=0.3088, pruned_loss=0.1013, over 7278.00 frames. ], batch size: 77, lr: 3.37e-02, grad_scale: 16.0 +2023-03-20 18:31:48,091 INFO [train.py:926] (0/2) Computing validation loss +2023-03-20 18:31:58,052 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8830, 1.6173, 1.7557, 1.4476, 1.6083, 1.7130, 2.2232, 1.8587], + device='cuda:0'), covar=tensor([0.1431, 0.1196, 0.1579, 0.1217, 0.1859, 0.1604, 0.0731, 0.1117], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0036, 0.0037, 0.0041, 0.0036, 0.0041, 0.0038, 0.0039], + device='cuda:0'), out_proj_covar=tensor([6.8607e-05, 5.6603e-05, 5.8609e-05, 6.2529e-05, 5.9236e-05, 6.4035e-05, + 6.5089e-05, 6.3142e-05], device='cuda:0') +2023-03-20 18:32:13,093 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3380, 3.1700, 2.9746, 3.1370, 2.3723, 2.2862, 3.5836, 2.9211], + device='cuda:0'), covar=tensor([0.0081, 0.0102, 0.0293, 0.0031, 0.0267, 0.0741, 0.0127, 0.0449], + device='cuda:0'), in_proj_covar=tensor([0.0090, 0.0083, 0.0129, 0.0088, 0.0153, 0.0222, 0.0097, 0.0202], + device='cuda:0'), out_proj_covar=tensor([7.2142e-05, 6.4976e-05, 9.3871e-05, 6.1678e-05, 1.1609e-04, 1.6592e-04, + 7.8601e-05, 1.5781e-04], device='cuda:0') +2023-03-20 18:32:13,388 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5896, 4.5543, 4.7037, 4.9309, 5.2024, 5.1369, 4.7108, 4.4918], + device='cuda:0'), covar=tensor([0.0565, 0.0953, 0.2092, 0.1724, 0.0474, 0.1046, 0.0684, 0.0751], + device='cuda:0'), in_proj_covar=tensor([0.0091, 0.0156, 0.0163, 0.0125, 0.0107, 0.0152, 0.0091, 0.0113], + device='cuda:0'), out_proj_covar=tensor([9.7805e-05, 1.7717e-04, 1.8289e-04, 1.5127e-04, 1.1127e-04, 1.7997e-04, + 9.5953e-05, 1.1631e-04], device='cuda:0') +2023-03-20 18:32:14,287 INFO [train.py:935] (0/2) Epoch 4, validation: loss=0.2099, simple_loss=0.2916, pruned_loss=0.06411, over 1622729.00 frames. +2023-03-20 18:32:14,288 INFO [train.py:936] (0/2) Maximum memory allocated so far is 11856MB +2023-03-20 18:32:20,176 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.42 vs. limit=5.0 +2023-03-20 18:32:21,411 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-20 18:32:29,773 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.888e+02 3.865e+02 4.722e+02 6.124e+02 1.429e+03, threshold=9.445e+02, percent-clipped=4.0 +2023-03-20 18:32:30,892 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=8506.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:32:32,843 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-20 18:32:33,662 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.00 vs. limit=2.0 +2023-03-20 18:32:39,374 INFO [train.py:901] (0/2) Epoch 4, batch 50, loss[loss=0.2296, simple_loss=0.2767, pruned_loss=0.09129, over 7157.00 frames. ], tot_loss[loss=0.2575, simple_loss=0.3046, pruned_loss=0.1052, over 323107.14 frames. ], batch size: 39, lr: 3.36e-02, grad_scale: 16.0 +2023-03-20 18:32:41,010 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-20 18:32:43,492 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-20 18:32:44,075 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2724, 4.3176, 4.2111, 4.1851, 3.9363, 4.1678, 4.2848, 4.5071], + device='cuda:0'), covar=tensor([0.0235, 0.0251, 0.0274, 0.0325, 0.0483, 0.0299, 0.0475, 0.0446], + device='cuda:0'), in_proj_covar=tensor([0.0070, 0.0074, 0.0069, 0.0088, 0.0081, 0.0059, 0.0059, 0.0060], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 18:32:46,035 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-20 18:32:52,246 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1305, 3.5245, 3.9904, 3.5563, 3.5411, 3.6167, 3.9901, 3.8001], + device='cuda:0'), covar=tensor([0.0085, 0.0204, 0.0113, 0.0230, 0.0242, 0.0176, 0.0132, 0.0138], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0037, 0.0040, 0.0035, 0.0044, 0.0044, 0.0034, 0.0035], + device='cuda:0'), out_proj_covar=tensor([6.8613e-05, 7.6211e-05, 8.0169e-05, 7.1678e-05, 9.3654e-05, 9.1555e-05, + 7.8075e-05, 6.8254e-05], device='cuda:0') +2023-03-20 18:33:03,332 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-20 18:33:03,834 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-20 18:33:05,781 INFO [train.py:901] (0/2) Epoch 4, batch 100, loss[loss=0.2704, simple_loss=0.3171, pruned_loss=0.1119, over 7275.00 frames. ], tot_loss[loss=0.2575, simple_loss=0.3056, pruned_loss=0.1047, over 573318.78 frames. ], batch size: 77, lr: 3.36e-02, grad_scale: 16.0 +2023-03-20 18:33:06,553 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.98 vs. limit=2.0 +2023-03-20 18:33:21,692 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.466e+02 3.550e+02 4.535e+02 5.650e+02 1.214e+03, threshold=9.071e+02, percent-clipped=4.0 +2023-03-20 18:33:28,569 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.6722, 0.8627, 0.7467, 0.5640, 1.3652, 0.9451, 0.5714, 0.6754], + device='cuda:0'), covar=tensor([0.0628, 0.0237, 0.0903, 0.0469, 0.0283, 0.0393, 0.0544, 0.0586], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0020, 0.0017, 0.0018, 0.0020, 0.0021, 0.0019, 0.0019], + device='cuda:0'), out_proj_covar=tensor([2.2693e-05, 1.9885e-05, 2.1928e-05, 2.0991e-05, 2.0002e-05, 2.4728e-05, + 2.4557e-05, 2.4037e-05], device='cuda:0') +2023-03-20 18:33:30,535 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.9596, 0.8357, 0.6590, 0.4424, 1.6641, 0.9517, 0.6348, 0.8188], + device='cuda:0'), covar=tensor([0.0386, 0.0263, 0.0743, 0.0936, 0.0162, 0.0450, 0.0960, 0.0403], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0020, 0.0018, 0.0018, 0.0020, 0.0021, 0.0019, 0.0019], + device='cuda:0'), out_proj_covar=tensor([2.2894e-05, 2.0006e-05, 2.2296e-05, 2.1346e-05, 2.0181e-05, 2.5037e-05, + 2.4847e-05, 2.4415e-05], device='cuda:0') +2023-03-20 18:33:30,971 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=8621.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:33:31,850 INFO [train.py:901] (0/2) Epoch 4, batch 150, loss[loss=0.266, simple_loss=0.3204, pruned_loss=0.1058, over 7275.00 frames. ], tot_loss[loss=0.2596, simple_loss=0.3074, pruned_loss=0.1059, over 767521.48 frames. ], batch size: 77, lr: 3.35e-02, grad_scale: 16.0 +2023-03-20 18:33:57,549 INFO [train.py:901] (0/2) Epoch 4, batch 200, loss[loss=0.2601, simple_loss=0.3112, pruned_loss=0.1045, over 7331.00 frames. ], tot_loss[loss=0.2605, simple_loss=0.3077, pruned_loss=0.1066, over 917294.36 frames. ], batch size: 75, lr: 3.34e-02, grad_scale: 16.0 +2023-03-20 18:34:04,254 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-20 18:34:09,173 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-20 18:34:13,663 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.118e+02 3.602e+02 4.396e+02 5.497e+02 1.117e+03, threshold=8.791e+02, percent-clipped=2.0 +2023-03-20 18:34:14,699 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-20 18:34:23,234 INFO [train.py:901] (0/2) Epoch 4, batch 250, loss[loss=0.2737, simple_loss=0.3214, pruned_loss=0.113, over 7224.00 frames. ], tot_loss[loss=0.2583, simple_loss=0.3061, pruned_loss=0.1052, over 1033650.78 frames. ], batch size: 93, lr: 3.33e-02, grad_scale: 16.0 +2023-03-20 18:34:27,298 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-20 18:34:42,562 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=8760.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:34:48,097 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-20 18:34:49,071 INFO [train.py:901] (0/2) Epoch 4, batch 300, loss[loss=0.2673, simple_loss=0.3158, pruned_loss=0.1094, over 7347.00 frames. ], tot_loss[loss=0.2575, simple_loss=0.3056, pruned_loss=0.1047, over 1126007.77 frames. ], batch size: 73, lr: 3.33e-02, grad_scale: 16.0 +2023-03-20 18:34:58,070 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-20 18:35:03,807 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.03 vs. limit=2.0 +2023-03-20 18:35:05,462 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.454e+02 3.664e+02 4.389e+02 5.371e+02 1.280e+03, threshold=8.778e+02, percent-clipped=5.0 +2023-03-20 18:35:14,076 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=8821.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:35:14,945 INFO [train.py:901] (0/2) Epoch 4, batch 350, loss[loss=0.2427, simple_loss=0.2922, pruned_loss=0.09665, over 7280.00 frames. ], tot_loss[loss=0.2573, simple_loss=0.3052, pruned_loss=0.1047, over 1198212.16 frames. ], batch size: 66, lr: 3.32e-02, grad_scale: 16.0 +2023-03-20 18:35:32,147 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-20 18:35:36,421 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7200, 3.2689, 2.4823, 2.7787, 1.9243, 2.5716, 2.0823, 2.5630], + device='cuda:0'), covar=tensor([0.0329, 0.0098, 0.0827, 0.0379, 0.0680, 0.0558, 0.0724, 0.0354], + device='cuda:0'), in_proj_covar=tensor([0.0024, 0.0022, 0.0028, 0.0026, 0.0026, 0.0026, 0.0030, 0.0026], + device='cuda:0'), out_proj_covar=tensor([4.1773e-05, 3.3956e-05, 4.9423e-05, 4.6090e-05, 4.5217e-05, 4.6324e-05, + 5.1189e-05, 4.4998e-05], device='cuda:0') +2023-03-20 18:35:41,199 INFO [train.py:901] (0/2) Epoch 4, batch 400, loss[loss=0.3049, simple_loss=0.3383, pruned_loss=0.1357, over 7317.00 frames. ], tot_loss[loss=0.2589, simple_loss=0.3064, pruned_loss=0.1057, over 1253820.36 frames. ], batch size: 49, lr: 3.31e-02, grad_scale: 16.0 +2023-03-20 18:35:56,945 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.279e+02 3.727e+02 4.768e+02 5.740e+02 1.240e+03, threshold=9.535e+02, percent-clipped=4.0 +2023-03-20 18:36:06,307 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=8921.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:36:07,226 INFO [train.py:901] (0/2) Epoch 4, batch 450, loss[loss=0.269, simple_loss=0.3181, pruned_loss=0.11, over 7307.00 frames. ], tot_loss[loss=0.2573, simple_loss=0.3053, pruned_loss=0.1047, over 1298133.71 frames. ], batch size: 80, lr: 3.31e-02, grad_scale: 16.0 +2023-03-20 18:36:09,378 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=8927.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:36:13,571 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.04 vs. limit=2.0 +2023-03-20 18:36:14,829 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-20 18:36:15,360 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-20 18:36:31,032 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=8969.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:36:32,989 INFO [train.py:901] (0/2) Epoch 4, batch 500, loss[loss=0.2293, simple_loss=0.2736, pruned_loss=0.09246, over 7139.00 frames. ], tot_loss[loss=0.2571, simple_loss=0.3047, pruned_loss=0.1048, over 1327736.38 frames. ], batch size: 41, lr: 3.30e-02, grad_scale: 16.0 +2023-03-20 18:36:37,250 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.92 vs. limit=2.0 +2023-03-20 18:36:39,667 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=8986.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:36:40,712 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=8988.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:36:45,484 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.10 vs. limit=2.0 +2023-03-20 18:36:46,840 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-20 18:36:48,590 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-20 18:36:49,162 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-20 18:36:49,619 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.254e+02 3.528e+02 4.568e+02 6.123e+02 9.889e+02, threshold=9.136e+02, percent-clipped=2.0 +2023-03-20 18:36:51,174 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-20 18:36:53,595 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.86 vs. limit=2.0 +2023-03-20 18:36:54,398 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=9013.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:36:55,286 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-20 18:36:59,303 INFO [train.py:901] (0/2) Epoch 4, batch 550, loss[loss=0.3142, simple_loss=0.3538, pruned_loss=0.1373, over 7235.00 frames. ], tot_loss[loss=0.258, simple_loss=0.3058, pruned_loss=0.1051, over 1354596.53 frames. ], batch size: 93, lr: 3.29e-02, grad_scale: 16.0 +2023-03-20 18:37:05,982 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-20 18:37:11,949 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=9047.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:37:14,858 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-20 18:37:18,516 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-20 18:37:18,797 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.09 vs. limit=2.0 +2023-03-20 18:37:25,075 INFO [train.py:901] (0/2) Epoch 4, batch 600, loss[loss=0.2652, simple_loss=0.3022, pruned_loss=0.1141, over 7258.00 frames. ], tot_loss[loss=0.2573, simple_loss=0.305, pruned_loss=0.1048, over 1371019.97 frames. ], batch size: 64, lr: 3.29e-02, grad_scale: 16.0 +2023-03-20 18:37:25,724 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=9074.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:37:26,115 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-20 18:37:27,385 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9562, 2.5990, 2.2386, 3.1362, 1.8729, 3.1893, 2.7896, 3.3278], + device='cuda:0'), covar=tensor([0.0064, 0.0408, 0.2647, 0.0150, 0.6615, 0.0126, 0.0543, 0.0145], + device='cuda:0'), in_proj_covar=tensor([0.0098, 0.0151, 0.0260, 0.0100, 0.0251, 0.0098, 0.0171, 0.0103], + device='cuda:0'), out_proj_covar=tensor([8.1791e-05, 1.2596e-04, 1.9891e-04, 8.4809e-05, 2.0197e-04, 8.2916e-05, + 1.3971e-04, 8.5179e-05], device='cuda:0') +2023-03-20 18:37:40,893 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8424, 3.6819, 3.6198, 3.0788, 3.6636, 2.7442, 1.4280, 3.9341], + device='cuda:0'), covar=tensor([0.0035, 0.0374, 0.0089, 0.0099, 0.0041, 0.0425, 0.1399, 0.0122], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0036, 0.0046, 0.0039, 0.0039, 0.0056, 0.0077, 0.0041], + device='cuda:0'), out_proj_covar=tensor([4.7571e-05, 5.7038e-05, 6.2977e-05, 5.5432e-05, 4.8728e-05, 8.0426e-05, + 1.1328e-04, 5.4806e-05], device='cuda:0') +2023-03-20 18:37:41,225 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.856e+02 3.802e+02 4.609e+02 5.829e+02 9.078e+02, threshold=9.219e+02, percent-clipped=0.0 +2023-03-20 18:37:42,771 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-20 18:37:47,418 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=9116.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:37:51,481 INFO [train.py:901] (0/2) Epoch 4, batch 650, loss[loss=0.2319, simple_loss=0.2778, pruned_loss=0.09301, over 7136.00 frames. ], tot_loss[loss=0.2566, simple_loss=0.304, pruned_loss=0.1046, over 1384212.81 frames. ], batch size: 41, lr: 3.28e-02, grad_scale: 16.0 +2023-03-20 18:37:52,519 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-20 18:37:56,219 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4853, 1.8153, 2.0692, 2.0260, 1.5695, 2.5441, 1.9498, 2.3946], + device='cuda:0'), covar=tensor([0.0119, 0.0673, 0.2626, 0.0163, 0.5302, 0.0111, 0.0800, 0.0085], + device='cuda:0'), in_proj_covar=tensor([0.0100, 0.0159, 0.0273, 0.0105, 0.0262, 0.0101, 0.0181, 0.0105], + device='cuda:0'), out_proj_covar=tensor([8.3734e-05, 1.3213e-04, 2.0869e-04, 8.9259e-05, 2.1085e-04, 8.6085e-05, + 1.4765e-04, 8.7737e-05], device='cuda:0') +2023-03-20 18:38:08,924 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-20 18:38:14,412 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.93 vs. limit=2.0 +2023-03-20 18:38:17,014 INFO [train.py:901] (0/2) Epoch 4, batch 700, loss[loss=0.2653, simple_loss=0.3174, pruned_loss=0.1066, over 7286.00 frames. ], tot_loss[loss=0.257, simple_loss=0.3048, pruned_loss=0.1046, over 1397392.33 frames. ], batch size: 77, lr: 3.27e-02, grad_scale: 16.0 +2023-03-20 18:38:17,533 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-20 18:38:18,163 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4321, 3.6226, 3.2172, 3.5263, 3.4919, 3.4668, 3.3150, 2.8512], + device='cuda:0'), covar=tensor([0.0091, 0.0121, 0.0120, 0.0117, 0.0098, 0.0086, 0.0111, 0.0203], + device='cuda:0'), in_proj_covar=tensor([0.0023, 0.0025, 0.0024, 0.0022, 0.0021, 0.0024, 0.0027, 0.0027], + device='cuda:0'), out_proj_covar=tensor([5.7205e-05, 6.9382e-05, 6.6417e-05, 5.6237e-05, 5.4179e-05, 5.9667e-05, + 7.2437e-05, 6.8464e-05], device='cuda:0') +2023-03-20 18:38:25,593 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6323, 2.2217, 3.2626, 3.5841, 3.3965, 3.2423, 2.8183, 3.1476], + device='cuda:0'), covar=tensor([0.1061, 0.0693, 0.1254, 0.0180, 0.0064, 0.0032, 0.0089, 0.0049], + device='cuda:0'), in_proj_covar=tensor([0.0206, 0.0148, 0.0197, 0.0101, 0.0077, 0.0079, 0.0077, 0.0082], + device='cuda:0'), out_proj_covar=tensor([1.7328e-04, 1.3298e-04, 1.6718e-04, 9.7193e-05, 6.9708e-05, 6.9178e-05, + 6.9651e-05, 7.4192e-05], device='cuda:0') +2023-03-20 18:38:33,366 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.850e+02 3.888e+02 4.537e+02 6.098e+02 1.462e+03, threshold=9.073e+02, percent-clipped=2.0 +2023-03-20 18:38:38,771 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.00 vs. limit=2.0 +2023-03-20 18:38:41,429 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-20 18:38:42,373 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-20 18:38:42,901 INFO [train.py:901] (0/2) Epoch 4, batch 750, loss[loss=0.2141, simple_loss=0.2715, pruned_loss=0.0783, over 7345.00 frames. ], tot_loss[loss=0.2572, simple_loss=0.3047, pruned_loss=0.1049, over 1406641.64 frames. ], batch size: 44, lr: 3.26e-02, grad_scale: 16.0 +2023-03-20 18:38:55,636 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-20 18:38:58,313 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5030, 4.7298, 4.5911, 4.9173, 5.0191, 5.0857, 4.6559, 4.5505], + device='cuda:0'), covar=tensor([0.0545, 0.1012, 0.1490, 0.1027, 0.0496, 0.0801, 0.0430, 0.0632], + device='cuda:0'), in_proj_covar=tensor([0.0096, 0.0158, 0.0157, 0.0124, 0.0108, 0.0152, 0.0092, 0.0115], + device='cuda:0'), out_proj_covar=tensor([1.0363e-04, 1.7764e-04, 1.7709e-04, 1.4837e-04, 1.1381e-04, 1.7756e-04, + 9.4331e-05, 1.1984e-04], device='cuda:0') +2023-03-20 18:39:00,790 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-20 18:39:06,832 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-20 18:39:08,375 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-20 18:39:08,865 INFO [train.py:901] (0/2) Epoch 4, batch 800, loss[loss=0.2397, simple_loss=0.296, pruned_loss=0.09172, over 7238.00 frames. ], tot_loss[loss=0.2553, simple_loss=0.3033, pruned_loss=0.1037, over 1413394.31 frames. ], batch size: 93, lr: 3.26e-02, grad_scale: 16.0 +2023-03-20 18:39:11,007 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3980, 3.4031, 3.2558, 3.5034, 3.5777, 3.4347, 3.2847, 3.1079], + device='cuda:0'), covar=tensor([0.0070, 0.0142, 0.0119, 0.0090, 0.0104, 0.0086, 0.0122, 0.0142], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0023, 0.0023, 0.0020, 0.0018, 0.0022, 0.0025, 0.0024], + device='cuda:0'), out_proj_covar=tensor([5.0988e-05, 6.1957e-05, 6.2685e-05, 5.1375e-05, 4.9656e-05, 5.6627e-05, + 6.6686e-05, 6.2749e-05], device='cuda:0') +2023-03-20 18:39:14,001 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=9283.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:39:19,093 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-20 18:39:25,235 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.196e+02 3.605e+02 4.493e+02 5.434e+02 9.732e+02, threshold=8.986e+02, percent-clipped=1.0 +2023-03-20 18:39:34,825 INFO [train.py:901] (0/2) Epoch 4, batch 850, loss[loss=0.2506, simple_loss=0.3118, pruned_loss=0.0947, over 7247.00 frames. ], tot_loss[loss=0.2557, simple_loss=0.3043, pruned_loss=0.1036, over 1422120.86 frames. ], batch size: 89, lr: 3.25e-02, grad_scale: 16.0 +2023-03-20 18:39:37,371 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-20 18:39:37,380 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-20 18:39:43,536 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-20 18:39:45,174 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=9342.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:39:47,169 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-20 18:39:54,262 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.6301, 0.7112, 1.0921, 0.8732, 0.5905, 0.6001, 0.9140, 0.7223], + device='cuda:0'), covar=tensor([0.0768, 0.2112, 0.0409, 0.0404, 0.2108, 0.2072, 0.0480, 0.0948], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0032, 0.0026, 0.0026, 0.0026, 0.0029, 0.0030, 0.0030], + device='cuda:0'), out_proj_covar=tensor([3.7016e-05, 5.1161e-05, 3.6091e-05, 3.6705e-05, 3.9401e-05, 4.3596e-05, + 3.7609e-05, 4.1468e-05], device='cuda:0') +2023-03-20 18:39:57,730 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1193, 4.0944, 3.9659, 4.3627, 4.5629, 4.5074, 3.9889, 3.9980], + device='cuda:0'), covar=tensor([0.0769, 0.1404, 0.2225, 0.1186, 0.0410, 0.1155, 0.0627, 0.0879], + device='cuda:0'), in_proj_covar=tensor([0.0094, 0.0156, 0.0160, 0.0125, 0.0106, 0.0157, 0.0090, 0.0114], + device='cuda:0'), out_proj_covar=tensor([1.0147e-04, 1.7449e-04, 1.7755e-04, 1.4898e-04, 1.1187e-04, 1.8264e-04, + 9.4266e-05, 1.1920e-04], device='cuda:0') +2023-03-20 18:39:58,745 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=9369.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:40:00,733 INFO [train.py:901] (0/2) Epoch 4, batch 900, loss[loss=0.258, simple_loss=0.3124, pruned_loss=0.1018, over 7324.00 frames. ], tot_loss[loss=0.2545, simple_loss=0.3036, pruned_loss=0.1027, over 1425099.84 frames. ], batch size: 61, lr: 3.24e-02, grad_scale: 16.0 +2023-03-20 18:40:07,590 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.9567, 1.6495, 0.9015, 1.6984, 0.8003, 1.8052, 1.3403, 1.0290], + device='cuda:0'), covar=tensor([0.0975, 0.0344, 0.0546, 0.0497, 0.0728, 0.0323, 0.0292, 0.0556], + device='cuda:0'), in_proj_covar=tensor([0.0022, 0.0023, 0.0023, 0.0022, 0.0022, 0.0023, 0.0021, 0.0023], + device='cuda:0'), out_proj_covar=tensor([4.0502e-05, 3.8030e-05, 3.7896e-05, 3.6956e-05, 3.9696e-05, 4.2093e-05, + 3.4538e-05, 4.2567e-05], device='cuda:0') +2023-03-20 18:40:17,374 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.070e+02 3.493e+02 4.561e+02 5.943e+02 1.247e+03, threshold=9.122e+02, percent-clipped=7.0 +2023-03-20 18:40:22,599 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=9414.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:40:23,575 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=9416.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:40:24,490 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-20 18:40:27,099 INFO [train.py:901] (0/2) Epoch 4, batch 950, loss[loss=0.2706, simple_loss=0.3271, pruned_loss=0.1071, over 7261.00 frames. ], tot_loss[loss=0.2548, simple_loss=0.304, pruned_loss=0.1029, over 1431087.38 frames. ], batch size: 64, lr: 3.24e-02, grad_scale: 16.0 +2023-03-20 18:40:32,907 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5945, 3.8438, 3.8681, 3.7480, 3.4510, 3.6659, 3.9491, 4.1304], + device='cuda:0'), covar=tensor([0.0317, 0.0209, 0.0242, 0.0257, 0.0536, 0.0361, 0.0363, 0.0189], + device='cuda:0'), in_proj_covar=tensor([0.0067, 0.0070, 0.0063, 0.0084, 0.0072, 0.0058, 0.0057, 0.0056], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0002, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 18:40:34,939 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.8507, 0.9514, 1.6309, 0.9585, 0.8070, 1.0987, 1.0184, 0.9192], + device='cuda:0'), covar=tensor([0.0841, 0.1065, 0.0148, 0.0624, 0.1339, 0.0845, 0.0637, 0.1076], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0031, 0.0025, 0.0027, 0.0026, 0.0030, 0.0030, 0.0029], + device='cuda:0'), out_proj_covar=tensor([3.8444e-05, 4.9851e-05, 3.4914e-05, 3.8322e-05, 3.9912e-05, 4.4197e-05, + 3.8308e-05, 4.0278e-05], device='cuda:0') +2023-03-20 18:40:45,158 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1294, 3.4417, 3.0944, 3.4104, 3.3432, 2.9944, 3.3699, 3.2420], + device='cuda:0'), covar=tensor([0.0107, 0.0147, 0.0140, 0.0096, 0.0100, 0.0142, 0.0109, 0.0131], + device='cuda:0'), in_proj_covar=tensor([0.0022, 0.0024, 0.0023, 0.0021, 0.0020, 0.0023, 0.0026, 0.0025], + device='cuda:0'), out_proj_covar=tensor([5.6151e-05, 6.6350e-05, 6.5396e-05, 5.4295e-05, 5.3915e-05, 5.9367e-05, + 7.1682e-05, 6.5994e-05], device='cuda:0') +2023-03-20 18:40:49,161 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=9464.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:40:49,635 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-20 18:40:53,611 INFO [train.py:901] (0/2) Epoch 4, batch 1000, loss[loss=0.2556, simple_loss=0.3011, pruned_loss=0.105, over 7361.00 frames. ], tot_loss[loss=0.2545, simple_loss=0.3036, pruned_loss=0.1027, over 1433437.90 frames. ], batch size: 73, lr: 3.23e-02, grad_scale: 16.0 +2023-03-20 18:40:54,754 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=9475.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:41:00,829 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.53 vs. limit=5.0 +2023-03-20 18:41:06,237 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8536, 2.9988, 2.5260, 2.6742, 3.0404, 3.1861, 1.4277, 2.7263], + device='cuda:0'), covar=tensor([0.0455, 0.0159, 0.0874, 0.1045, 0.0374, 0.0563, 0.1374, 0.0558], + device='cuda:0'), in_proj_covar=tensor([0.0023, 0.0022, 0.0025, 0.0026, 0.0023, 0.0025, 0.0029, 0.0025], + device='cuda:0'), out_proj_covar=tensor([4.4056e-05, 3.5563e-05, 4.7386e-05, 4.9374e-05, 4.1940e-05, 4.6307e-05, + 5.4777e-05, 4.5458e-05], device='cuda:0') +2023-03-20 18:41:09,172 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.060e+02 3.756e+02 4.692e+02 5.891e+02 2.084e+03, threshold=9.385e+02, percent-clipped=6.0 +2023-03-20 18:41:09,722 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-20 18:41:19,221 INFO [train.py:901] (0/2) Epoch 4, batch 1050, loss[loss=0.2426, simple_loss=0.2969, pruned_loss=0.09417, over 7254.00 frames. ], tot_loss[loss=0.2536, simple_loss=0.3029, pruned_loss=0.1021, over 1436084.25 frames. ], batch size: 89, lr: 3.22e-02, grad_scale: 16.0 +2023-03-20 18:41:24,483 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0033, 1.9705, 1.8026, 1.6482, 2.2061, 2.1642, 2.0126, 2.3395], + device='cuda:0'), covar=tensor([0.0451, 0.0479, 0.0951, 0.1047, 0.0441, 0.0470, 0.0833, 0.0659], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0035, 0.0034, 0.0036, 0.0033, 0.0038, 0.0035, 0.0035], + device='cuda:0'), out_proj_covar=tensor([6.9255e-05, 5.9801e-05, 5.9629e-05, 6.1325e-05, 5.8636e-05, 6.4714e-05, + 6.4846e-05, 6.2886e-05], device='cuda:0') +2023-03-20 18:41:25,508 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.7202, 1.0153, 0.6232, 1.1468, 1.3131, 0.6424, 0.7184, 1.0048], + device='cuda:0'), covar=tensor([0.0959, 0.0201, 0.0447, 0.0525, 0.0262, 0.0616, 0.0750, 0.0259], + device='cuda:0'), in_proj_covar=tensor([0.0021, 0.0020, 0.0019, 0.0020, 0.0020, 0.0019, 0.0020, 0.0019], + device='cuda:0'), out_proj_covar=tensor([2.4088e-05, 2.1451e-05, 2.4769e-05, 2.4306e-05, 2.2495e-05, 2.2792e-05, + 2.6245e-05, 2.5674e-05], device='cuda:0') +2023-03-20 18:41:28,769 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.03 vs. limit=2.0 +2023-03-20 18:41:32,360 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-20 18:41:36,925 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-20 18:41:44,817 INFO [train.py:901] (0/2) Epoch 4, batch 1100, loss[loss=0.2469, simple_loss=0.2971, pruned_loss=0.09835, over 7350.00 frames. ], tot_loss[loss=0.2531, simple_loss=0.3027, pruned_loss=0.1017, over 1438112.92 frames. ], batch size: 63, lr: 3.22e-02, grad_scale: 16.0 +2023-03-20 18:41:48,190 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.85 vs. limit=5.0 +2023-03-20 18:41:49,899 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=9583.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:41:50,475 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=9584.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:42:01,279 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.878e+02 3.483e+02 4.054e+02 5.375e+02 1.176e+03, threshold=8.109e+02, percent-clipped=1.0 +2023-03-20 18:42:05,548 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.05 vs. limit=2.0 +2023-03-20 18:42:05,764 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-20 18:42:05,779 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 18:42:10,814 INFO [train.py:901] (0/2) Epoch 4, batch 1150, loss[loss=0.2503, simple_loss=0.2994, pruned_loss=0.1006, over 7320.00 frames. ], tot_loss[loss=0.2514, simple_loss=0.3011, pruned_loss=0.1008, over 1437590.03 frames. ], batch size: 83, lr: 3.21e-02, grad_scale: 16.0 +2023-03-20 18:42:14,924 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=9631.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:42:19,015 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-20 18:42:19,498 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-20 18:42:20,834 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.05 vs. limit=2.0 +2023-03-20 18:42:21,089 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=9642.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:42:22,651 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=9645.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:42:29,018 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=9658.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:42:34,631 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=9669.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:42:36,525 INFO [train.py:901] (0/2) Epoch 4, batch 1200, loss[loss=0.2393, simple_loss=0.2936, pruned_loss=0.09252, over 7346.00 frames. ], tot_loss[loss=0.2535, simple_loss=0.3027, pruned_loss=0.1021, over 1438315.63 frames. ], batch size: 54, lr: 3.20e-02, grad_scale: 16.0 +2023-03-20 18:42:45,919 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=9690.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:42:50,470 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-20 18:42:51,550 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0798, 1.3784, 1.9692, 1.8993, 2.1156, 1.9853, 2.1956, 2.0058], + device='cuda:0'), covar=tensor([0.0533, 0.1253, 0.0518, 0.0500, 0.0644, 0.0598, 0.0360, 0.0591], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0035, 0.0033, 0.0034, 0.0033, 0.0037, 0.0033, 0.0035], + device='cuda:0'), out_proj_covar=tensor([6.9870e-05, 6.0299e-05, 5.7627e-05, 5.9254e-05, 5.9818e-05, 6.4260e-05, + 6.1163e-05, 6.2987e-05], device='cuda:0') +2023-03-20 18:42:52,841 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.497e+02 3.682e+02 4.480e+02 5.873e+02 1.157e+03, threshold=8.961e+02, percent-clipped=5.0 +2023-03-20 18:42:59,413 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=9717.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:43:01,168 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=9719.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:43:03,039 INFO [train.py:901] (0/2) Epoch 4, batch 1250, loss[loss=0.2474, simple_loss=0.3098, pruned_loss=0.09251, over 7306.00 frames. ], tot_loss[loss=0.2545, simple_loss=0.3037, pruned_loss=0.1027, over 1439301.73 frames. ], batch size: 61, lr: 3.20e-02, grad_scale: 16.0 +2023-03-20 18:43:05,684 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6885, 3.6571, 3.6389, 3.8199, 3.4189, 3.7742, 4.1303, 4.2207], + device='cuda:0'), covar=tensor([0.0303, 0.0291, 0.0324, 0.0241, 0.0504, 0.0276, 0.0371, 0.0195], + device='cuda:0'), in_proj_covar=tensor([0.0070, 0.0077, 0.0067, 0.0090, 0.0074, 0.0057, 0.0063, 0.0060], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0002, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 18:43:13,370 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=9743.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:43:14,277 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-20 18:43:18,199 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-20 18:43:19,192 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-20 18:43:19,792 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8136, 1.6405, 0.7583, 1.5139, 1.4932, 1.8671, 1.1940, 1.7926], + device='cuda:0'), covar=tensor([0.0666, 0.0471, 0.0689, 0.0635, 0.2082, 0.1714, 0.1148, 0.0346], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0022, 0.0021, 0.0021, 0.0022, 0.0023, 0.0022, 0.0021], + device='cuda:0'), out_proj_covar=tensor([3.7534e-05, 3.8691e-05, 3.5368e-05, 3.6332e-05, 4.0205e-05, 4.2332e-05, + 3.7211e-05, 4.0196e-05], device='cuda:0') +2023-03-20 18:43:26,787 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=9770.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:43:28,149 INFO [train.py:901] (0/2) Epoch 4, batch 1300, loss[loss=0.2615, simple_loss=0.3103, pruned_loss=0.1064, over 7338.00 frames. ], tot_loss[loss=0.2542, simple_loss=0.3036, pruned_loss=0.1024, over 1440600.25 frames. ], batch size: 54, lr: 3.19e-02, grad_scale: 32.0 +2023-03-20 18:43:42,660 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-20 18:43:44,625 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.204e+02 3.860e+02 4.926e+02 5.872e+02 1.226e+03, threshold=9.851e+02, percent-clipped=3.0 +2023-03-20 18:43:44,661 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-20 18:43:44,771 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=9804.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:43:49,283 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-20 18:43:54,838 INFO [train.py:901] (0/2) Epoch 4, batch 1350, loss[loss=0.258, simple_loss=0.3018, pruned_loss=0.1072, over 7361.00 frames. ], tot_loss[loss=0.2539, simple_loss=0.3033, pruned_loss=0.1022, over 1442255.38 frames. ], batch size: 51, lr: 3.18e-02, grad_scale: 32.0 +2023-03-20 18:43:59,383 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-20 18:44:11,104 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.8494, 0.8556, 0.3396, 0.8302, 0.7445, 0.8306, 0.8689, 0.7694], + device='cuda:0'), covar=tensor([0.0349, 0.0271, 0.0784, 0.0585, 0.0373, 0.0288, 0.0484, 0.0458], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0020, 0.0018, 0.0020, 0.0020, 0.0018, 0.0018, 0.0019], + device='cuda:0'), out_proj_covar=tensor([2.2184e-05, 2.0943e-05, 2.4479e-05, 2.5064e-05, 2.1810e-05, 2.1771e-05, + 2.4992e-05, 2.5627e-05], device='cuda:0') +2023-03-20 18:44:20,609 INFO [train.py:901] (0/2) Epoch 4, batch 1400, loss[loss=0.229, simple_loss=0.2843, pruned_loss=0.08689, over 7362.00 frames. ], tot_loss[loss=0.2528, simple_loss=0.3024, pruned_loss=0.1016, over 1441618.22 frames. ], batch size: 63, lr: 3.18e-02, grad_scale: 32.0 +2023-03-20 18:44:26,492 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.91 vs. limit=2.0 +2023-03-20 18:44:27,721 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6269, 3.1504, 2.3471, 2.6015, 2.7000, 2.4481, 1.8665, 2.5914], + device='cuda:0'), covar=tensor([0.0721, 0.0113, 0.0770, 0.1734, 0.0588, 0.0923, 0.0866, 0.0548], + device='cuda:0'), in_proj_covar=tensor([0.0024, 0.0022, 0.0024, 0.0026, 0.0024, 0.0026, 0.0029, 0.0025], + device='cuda:0'), out_proj_covar=tensor([4.6377e-05, 3.5842e-05, 4.7191e-05, 5.0513e-05, 4.4320e-05, 4.9009e-05, + 5.5666e-05, 4.7403e-05], device='cuda:0') +2023-03-20 18:44:31,200 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-20 18:44:36,036 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.02 vs. limit=2.0 +2023-03-20 18:44:37,217 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.715e+02 3.895e+02 4.499e+02 5.738e+02 1.302e+03, threshold=8.998e+02, percent-clipped=3.0 +2023-03-20 18:44:44,661 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.03 vs. limit=2.0 +2023-03-20 18:44:45,781 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.92 vs. limit=2.0 +2023-03-20 18:44:46,404 INFO [train.py:901] (0/2) Epoch 4, batch 1450, loss[loss=0.2739, simple_loss=0.3219, pruned_loss=0.113, over 7295.00 frames. ], tot_loss[loss=0.2522, simple_loss=0.3017, pruned_loss=0.1014, over 1439363.28 frames. ], batch size: 68, lr: 3.17e-02, grad_scale: 16.0 +2023-03-20 18:44:55,058 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=9940.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:44:56,995 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-20 18:45:12,170 INFO [train.py:901] (0/2) Epoch 4, batch 1500, loss[loss=0.257, simple_loss=0.3054, pruned_loss=0.1043, over 7342.00 frames. ], tot_loss[loss=0.251, simple_loss=0.3005, pruned_loss=0.1008, over 1439728.88 frames. ], batch size: 54, lr: 3.17e-02, grad_scale: 16.0 +2023-03-20 18:45:13,742 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-20 18:45:24,425 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3377, 1.9267, 2.0339, 2.8546, 2.8609, 2.8547, 2.8481, 2.6602], + device='cuda:0'), covar=tensor([0.1435, 0.0928, 0.2106, 0.0271, 0.0087, 0.0043, 0.0082, 0.0080], + device='cuda:0'), in_proj_covar=tensor([0.0230, 0.0156, 0.0211, 0.0114, 0.0083, 0.0087, 0.0084, 0.0088], + device='cuda:0'), out_proj_covar=tensor([1.9510e-04, 1.4168e-04, 1.8118e-04, 1.0939e-04, 7.6460e-05, 7.9474e-05, + 7.7586e-05, 8.1125e-05], device='cuda:0') +2023-03-20 18:45:29,014 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.227e+02 3.489e+02 4.498e+02 5.516e+02 1.678e+03, threshold=8.996e+02, percent-clipped=8.0 +2023-03-20 18:45:33,683 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=10014.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:45:37,625 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-20 18:45:38,114 INFO [train.py:901] (0/2) Epoch 4, batch 1550, loss[loss=0.2284, simple_loss=0.2756, pruned_loss=0.09056, over 7349.00 frames. ], tot_loss[loss=0.2515, simple_loss=0.3005, pruned_loss=0.1013, over 1437957.20 frames. ], batch size: 44, lr: 3.16e-02, grad_scale: 16.0 +2023-03-20 18:45:46,297 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=10039.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:45:58,526 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=10062.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 18:45:59,771 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-03-20 18:46:02,621 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=10070.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:46:04,054 INFO [train.py:901] (0/2) Epoch 4, batch 1600, loss[loss=0.2772, simple_loss=0.3263, pruned_loss=0.114, over 7265.00 frames. ], tot_loss[loss=0.2513, simple_loss=0.3002, pruned_loss=0.1012, over 1437823.87 frames. ], batch size: 89, lr: 3.15e-02, grad_scale: 16.0 +2023-03-20 18:46:11,145 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-20 18:46:12,143 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-20 18:46:14,794 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-20 18:46:17,906 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=10099.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:46:18,479 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=10100.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 18:46:19,962 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6603, 1.5340, 1.3683, 0.9608, 1.3039, 1.9985, 1.4280, 1.3869], + device='cuda:0'), covar=tensor([0.0294, 0.0536, 0.0287, 0.0259, 0.0571, 0.0272, 0.0522, 0.0184], + device='cuda:0'), in_proj_covar=tensor([0.0021, 0.0023, 0.0021, 0.0020, 0.0021, 0.0021, 0.0021, 0.0020], + device='cuda:0'), out_proj_covar=tensor([3.9132e-05, 4.0486e-05, 3.5887e-05, 3.4727e-05, 3.8924e-05, 4.0317e-05, + 3.6372e-05, 3.8719e-05], device='cuda:0') +2023-03-20 18:46:20,784 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.208e+02 3.857e+02 4.437e+02 5.410e+02 1.141e+03, threshold=8.873e+02, percent-clipped=2.0 +2023-03-20 18:46:24,239 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-20 18:46:27,350 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=10118.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:46:28,321 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-20 18:46:29,722 INFO [train.py:901] (0/2) Epoch 4, batch 1650, loss[loss=0.2621, simple_loss=0.3092, pruned_loss=0.1075, over 7313.00 frames. ], tot_loss[loss=0.2508, simple_loss=0.2998, pruned_loss=0.1009, over 1439603.94 frames. ], batch size: 83, lr: 3.15e-02, grad_scale: 16.0 +2023-03-20 18:46:29,884 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=10123.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 18:46:36,419 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-20 18:46:53,716 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 18:46:56,324 INFO [train.py:901] (0/2) Epoch 4, batch 1700, loss[loss=0.2529, simple_loss=0.3015, pruned_loss=0.1021, over 7276.00 frames. ], tot_loss[loss=0.25, simple_loss=0.2994, pruned_loss=0.1003, over 1438513.93 frames. ], batch size: 57, lr: 3.14e-02, grad_scale: 16.0 +2023-03-20 18:46:57,852 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-20 18:47:07,895 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-20 18:47:12,606 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.114e+02 3.691e+02 4.418e+02 5.552e+02 1.158e+03, threshold=8.837e+02, percent-clipped=2.0 +2023-03-20 18:47:22,346 INFO [train.py:901] (0/2) Epoch 4, batch 1750, loss[loss=0.2227, simple_loss=0.2765, pruned_loss=0.08442, over 7168.00 frames. ], tot_loss[loss=0.2495, simple_loss=0.2991, pruned_loss=0.0999, over 1439958.16 frames. ], batch size: 39, lr: 3.13e-02, grad_scale: 16.0 +2023-03-20 18:47:31,041 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=10240.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:47:33,999 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-20 18:47:35,022 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-20 18:47:48,134 INFO [train.py:901] (0/2) Epoch 4, batch 1800, loss[loss=0.2725, simple_loss=0.3196, pruned_loss=0.1127, over 7305.00 frames. ], tot_loss[loss=0.2494, simple_loss=0.2992, pruned_loss=0.09977, over 1441338.65 frames. ], batch size: 83, lr: 3.13e-02, grad_scale: 16.0 +2023-03-20 18:47:48,278 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.9962, 0.6692, 1.0174, 0.5300, 0.6878, 0.9514, 0.9931, 0.6488], + device='cuda:0'), covar=tensor([0.0143, 0.0261, 0.0518, 0.0698, 0.0201, 0.0393, 0.0315, 0.0617], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0021, 0.0018, 0.0021, 0.0020, 0.0019, 0.0019, 0.0021], + device='cuda:0'), out_proj_covar=tensor([2.2861e-05, 2.2417e-05, 2.4108e-05, 2.6110e-05, 2.2002e-05, 2.3370e-05, + 2.6619e-05, 2.8774e-05], device='cuda:0') +2023-03-20 18:47:55,818 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=10288.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:47:56,287 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-20 18:47:58,557 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.86 vs. limit=2.0 +2023-03-20 18:48:04,845 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.576e+02 3.726e+02 4.686e+02 5.832e+02 1.098e+03, threshold=9.372e+02, percent-clipped=3.0 +2023-03-20 18:48:09,468 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=10314.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:48:09,930 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-20 18:48:13,938 INFO [train.py:901] (0/2) Epoch 4, batch 1850, loss[loss=0.2564, simple_loss=0.3121, pruned_loss=0.1003, over 7300.00 frames. ], tot_loss[loss=0.2499, simple_loss=0.2996, pruned_loss=0.1001, over 1440482.90 frames. ], batch size: 86, lr: 3.12e-02, grad_scale: 16.0 +2023-03-20 18:48:19,543 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-20 18:48:29,796 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8120, 4.0692, 3.5171, 3.7062, 3.7800, 3.4175, 3.5600, 3.5081], + device='cuda:0'), covar=tensor([0.0085, 0.0101, 0.0121, 0.0088, 0.0082, 0.0124, 0.0141, 0.0130], + device='cuda:0'), in_proj_covar=tensor([0.0024, 0.0025, 0.0026, 0.0023, 0.0022, 0.0024, 0.0030, 0.0027], + device='cuda:0'), out_proj_covar=tensor([6.4633e-05, 7.3274e-05, 7.8804e-05, 6.1060e-05, 6.4502e-05, 6.5873e-05, + 8.8070e-05, 7.5958e-05], device='cuda:0') +2023-03-20 18:48:34,298 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=10362.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:48:37,337 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-20 18:48:39,768 INFO [train.py:901] (0/2) Epoch 4, batch 1900, loss[loss=0.277, simple_loss=0.3223, pruned_loss=0.1159, over 7279.00 frames. ], tot_loss[loss=0.2493, simple_loss=0.2992, pruned_loss=0.09972, over 1442984.50 frames. ], batch size: 66, lr: 3.12e-02, grad_scale: 16.0 +2023-03-20 18:48:51,616 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=10395.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 18:48:53,668 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=10399.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:48:53,841 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.88 vs. limit=2.0 +2023-03-20 18:48:56,774 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.239e+02 3.747e+02 4.648e+02 5.275e+02 9.127e+02, threshold=9.296e+02, percent-clipped=0.0 +2023-03-20 18:48:59,013 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=10409.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:49:02,388 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-20 18:49:03,462 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=10418.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 18:49:06,422 INFO [train.py:901] (0/2) Epoch 4, batch 1950, loss[loss=0.2432, simple_loss=0.3005, pruned_loss=0.09288, over 7270.00 frames. ], tot_loss[loss=0.248, simple_loss=0.2981, pruned_loss=0.09893, over 1443497.55 frames. ], batch size: 57, lr: 3.11e-02, grad_scale: 16.0 +2023-03-20 18:49:14,673 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-20 18:49:18,763 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-20 18:49:18,798 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=10447.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:49:19,744 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-20 18:49:22,360 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=10454.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:49:30,424 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=10470.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:49:31,810 INFO [train.py:901] (0/2) Epoch 4, batch 2000, loss[loss=0.264, simple_loss=0.3142, pruned_loss=0.1069, over 7365.00 frames. ], tot_loss[loss=0.2484, simple_loss=0.2985, pruned_loss=0.09912, over 1445895.45 frames. ], batch size: 73, lr: 3.10e-02, grad_scale: 16.0 +2023-03-20 18:49:36,384 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-20 18:49:48,103 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-20 18:49:48,600 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.129e+02 3.642e+02 4.667e+02 5.884e+02 9.942e+02, threshold=9.334e+02, percent-clipped=1.0 +2023-03-20 18:49:52,808 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.83 vs. limit=5.0 +2023-03-20 18:49:54,531 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=10515.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 18:49:56,887 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-20 18:49:58,401 INFO [train.py:901] (0/2) Epoch 4, batch 2050, loss[loss=0.2907, simple_loss=0.3362, pruned_loss=0.1226, over 7129.00 frames. ], tot_loss[loss=0.2493, simple_loss=0.2991, pruned_loss=0.09976, over 1442967.10 frames. ], batch size: 98, lr: 3.10e-02, grad_scale: 16.0 +2023-03-20 18:50:01,797 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.02 vs. limit=2.0 +2023-03-20 18:50:04,286 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.37 vs. limit=2.0 +2023-03-20 18:50:24,359 INFO [train.py:901] (0/2) Epoch 4, batch 2100, loss[loss=0.2496, simple_loss=0.3056, pruned_loss=0.09676, over 7251.00 frames. ], tot_loss[loss=0.2494, simple_loss=0.2996, pruned_loss=0.09955, over 1443498.64 frames. ], batch size: 55, lr: 3.09e-02, grad_scale: 16.0 +2023-03-20 18:50:30,388 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-20 18:50:33,335 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-20 18:50:41,274 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.704e+02 3.659e+02 4.368e+02 5.269e+02 1.016e+03, threshold=8.736e+02, percent-clipped=1.0 +2023-03-20 18:50:47,517 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0018, 3.1239, 2.2913, 3.1158, 2.5353, 2.9747, 2.2095, 1.8660], + device='cuda:0'), covar=tensor([0.0026, 0.0146, 0.0550, 0.0117, 0.0055, 0.0066, 0.0817, 0.0662], + device='cuda:0'), in_proj_covar=tensor([0.0079, 0.0097, 0.0190, 0.0088, 0.0088, 0.0095, 0.0200, 0.0192], + device='cuda:0'), out_proj_covar=tensor([8.7636e-05, 1.1267e-04, 1.9621e-04, 9.9180e-05, 9.8808e-05, 1.0855e-04, + 2.0924e-04, 1.9724e-04], device='cuda:0') +2023-03-20 18:50:50,278 INFO [train.py:901] (0/2) Epoch 4, batch 2150, loss[loss=0.2373, simple_loss=0.2963, pruned_loss=0.08918, over 7276.00 frames. ], tot_loss[loss=0.2489, simple_loss=0.2993, pruned_loss=0.09927, over 1443447.54 frames. ], batch size: 66, lr: 3.09e-02, grad_scale: 16.0 +2023-03-20 18:51:04,538 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1525, 4.3630, 4.0279, 4.1902, 3.8890, 4.3785, 4.6139, 4.7567], + device='cuda:0'), covar=tensor([0.0266, 0.0170, 0.0263, 0.0257, 0.0438, 0.0162, 0.0289, 0.0167], + device='cuda:0'), in_proj_covar=tensor([0.0075, 0.0076, 0.0071, 0.0090, 0.0078, 0.0058, 0.0064, 0.0064], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 18:51:16,245 INFO [train.py:901] (0/2) Epoch 4, batch 2200, loss[loss=0.2315, simple_loss=0.2829, pruned_loss=0.09003, over 7352.00 frames. ], tot_loss[loss=0.2473, simple_loss=0.2978, pruned_loss=0.09837, over 1440517.75 frames. ], batch size: 51, lr: 3.08e-02, grad_scale: 16.0 +2023-03-20 18:51:17,639 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.93 vs. limit=2.0 +2023-03-20 18:51:18,943 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2427, 3.7408, 3.6228, 4.2352, 3.2099, 2.8226, 4.3299, 3.4235], + device='cuda:0'), covar=tensor([0.0069, 0.0050, 0.0143, 0.0031, 0.0189, 0.0419, 0.0114, 0.0365], + device='cuda:0'), in_proj_covar=tensor([0.0124, 0.0109, 0.0169, 0.0108, 0.0203, 0.0253, 0.0129, 0.0247], + device='cuda:0'), out_proj_covar=tensor([1.0493e-04, 9.2138e-05, 1.3263e-04, 8.4240e-05, 1.6426e-04, 2.0092e-04, + 1.1307e-04, 1.9993e-04], device='cuda:0') +2023-03-20 18:51:19,783 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-20 18:51:28,083 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=10695.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:51:30,355 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.08 vs. limit=2.0 +2023-03-20 18:51:33,047 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.783e+02 3.779e+02 4.552e+02 5.953e+02 1.477e+03, threshold=9.104e+02, percent-clipped=10.0 +2023-03-20 18:51:39,705 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=10718.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 18:51:42,125 INFO [train.py:901] (0/2) Epoch 4, batch 2250, loss[loss=0.2619, simple_loss=0.3239, pruned_loss=0.09998, over 7240.00 frames. ], tot_loss[loss=0.2463, simple_loss=0.2974, pruned_loss=0.0976, over 1442496.01 frames. ], batch size: 93, lr: 3.07e-02, grad_scale: 16.0 +2023-03-20 18:51:52,470 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=10743.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:51:53,070 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=10744.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:51:53,956 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-20 18:51:53,967 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-20 18:52:04,181 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=10765.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:52:04,668 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=10766.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 18:52:08,289 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-20 18:52:08,785 INFO [train.py:901] (0/2) Epoch 4, batch 2300, loss[loss=0.2462, simple_loss=0.3015, pruned_loss=0.09549, over 7313.00 frames. ], tot_loss[loss=0.246, simple_loss=0.2973, pruned_loss=0.09738, over 1444347.28 frames. ], batch size: 83, lr: 3.07e-02, grad_scale: 16.0 +2023-03-20 18:52:09,481 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=10774.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:52:15,679 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2017, 1.5880, 2.0337, 1.6977, 2.1155, 1.6857, 2.0703, 1.9264], + device='cuda:0'), covar=tensor([0.0851, 0.0821, 0.0527, 0.0837, 0.1277, 0.0790, 0.0866, 0.1455], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0035, 0.0033, 0.0036, 0.0035, 0.0039, 0.0035, 0.0034], + device='cuda:0'), out_proj_covar=tensor([7.7227e-05, 6.4484e-05, 6.2578e-05, 6.6820e-05, 6.7185e-05, 7.1591e-05, + 6.9389e-05, 6.5003e-05], device='cuda:0') +2023-03-20 18:52:25,624 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.241e+02 3.504e+02 4.106e+02 5.041e+02 1.029e+03, threshold=8.211e+02, percent-clipped=2.0 +2023-03-20 18:52:25,813 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=10805.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:52:28,334 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=10810.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 18:52:34,790 INFO [train.py:901] (0/2) Epoch 4, batch 2350, loss[loss=0.2887, simple_loss=0.3209, pruned_loss=0.1282, over 7316.00 frames. ], tot_loss[loss=0.2456, simple_loss=0.2969, pruned_loss=0.09715, over 1444577.57 frames. ], batch size: 83, lr: 3.06e-02, grad_scale: 16.0 +2023-03-20 18:52:41,584 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=10835.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:52:52,872 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9492, 3.2914, 2.4221, 3.2236, 2.6391, 3.3306, 2.2966, 2.1275], + device='cuda:0'), covar=tensor([0.0048, 0.0212, 0.0535, 0.0087, 0.0086, 0.0070, 0.0712, 0.0604], + device='cuda:0'), in_proj_covar=tensor([0.0083, 0.0100, 0.0193, 0.0089, 0.0094, 0.0099, 0.0198, 0.0191], + device='cuda:0'), out_proj_covar=tensor([9.2710e-05, 1.1647e-04, 2.0111e-04, 1.0093e-04, 1.0453e-04, 1.1302e-04, + 2.0692e-04, 1.9532e-04], device='cuda:0') +2023-03-20 18:52:56,396 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-20 18:53:01,352 INFO [train.py:901] (0/2) Epoch 4, batch 2400, loss[loss=0.2188, simple_loss=0.2732, pruned_loss=0.08223, over 7193.00 frames. ], tot_loss[loss=0.2464, simple_loss=0.2976, pruned_loss=0.09759, over 1446276.27 frames. ], batch size: 39, lr: 3.06e-02, grad_scale: 16.0 +2023-03-20 18:53:02,392 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-20 18:53:05,336 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.85 vs. limit=2.0 +2023-03-20 18:53:07,602 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=10885.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 18:53:11,579 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9258, 4.2440, 4.3389, 4.2293, 4.2029, 3.6987, 4.3240, 4.2300], + device='cuda:0'), covar=tensor([0.0579, 0.0393, 0.0566, 0.0740, 0.0561, 0.0501, 0.0484, 0.0729], + device='cuda:0'), in_proj_covar=tensor([0.0096, 0.0080, 0.0094, 0.0078, 0.0073, 0.0089, 0.0078, 0.0071], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 18:53:13,011 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-20 18:53:16,114 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-20 18:53:17,636 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.227e+02 3.878e+02 4.809e+02 5.497e+02 2.353e+03, threshold=9.618e+02, percent-clipped=8.0 +2023-03-20 18:53:21,302 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2021, 1.2007, 1.4173, 1.0647, 1.1929, 1.7011, 1.2433, 0.9775], + device='cuda:0'), covar=tensor([0.0481, 0.0211, 0.0311, 0.0485, 0.1245, 0.0286, 0.0281, 0.0439], + device='cuda:0'), in_proj_covar=tensor([0.0022, 0.0022, 0.0021, 0.0020, 0.0023, 0.0020, 0.0021, 0.0021], + device='cuda:0'), out_proj_covar=tensor([4.2904e-05, 4.1873e-05, 3.6897e-05, 3.6082e-05, 4.4159e-05, 4.1338e-05, + 3.8178e-05, 4.3074e-05], device='cuda:0') +2023-03-20 18:53:22,717 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9827, 4.2659, 4.2920, 4.5446, 4.6534, 4.5747, 4.2007, 4.0608], + device='cuda:0'), covar=tensor([0.0896, 0.1289, 0.1551, 0.0981, 0.0485, 0.0936, 0.0435, 0.0777], + device='cuda:0'), in_proj_covar=tensor([0.0099, 0.0163, 0.0161, 0.0140, 0.0109, 0.0164, 0.0091, 0.0120], + device='cuda:0'), out_proj_covar=tensor([1.0869e-04, 1.7968e-04, 1.7437e-04, 1.6046e-04, 1.1749e-04, 1.8703e-04, + 9.7083e-05, 1.2553e-04], device='cuda:0') +2023-03-20 18:53:27,300 INFO [train.py:901] (0/2) Epoch 4, batch 2450, loss[loss=0.2424, simple_loss=0.2987, pruned_loss=0.09307, over 7246.00 frames. ], tot_loss[loss=0.2467, simple_loss=0.2979, pruned_loss=0.09772, over 1445614.95 frames. ], batch size: 55, lr: 3.05e-02, grad_scale: 16.0 +2023-03-20 18:53:39,729 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=10946.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 18:53:43,133 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-20 18:53:53,059 INFO [train.py:901] (0/2) Epoch 4, batch 2500, loss[loss=0.2388, simple_loss=0.2992, pruned_loss=0.08921, over 6743.00 frames. ], tot_loss[loss=0.2471, simple_loss=0.2979, pruned_loss=0.09811, over 1441724.10 frames. ], batch size: 107, lr: 3.04e-02, grad_scale: 16.0 +2023-03-20 18:54:07,984 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-20 18:54:09,457 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.005e+02 3.468e+02 4.610e+02 5.877e+02 1.140e+03, threshold=9.220e+02, percent-clipped=4.0 +2023-03-20 18:54:19,254 INFO [train.py:901] (0/2) Epoch 4, batch 2550, loss[loss=0.2609, simple_loss=0.3186, pruned_loss=0.1016, over 7148.00 frames. ], tot_loss[loss=0.2474, simple_loss=0.2982, pruned_loss=0.09824, over 1442464.02 frames. ], batch size: 98, lr: 3.04e-02, grad_scale: 16.0 +2023-03-20 18:54:32,839 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-03-20 18:54:41,105 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=11065.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:54:44,898 INFO [train.py:901] (0/2) Epoch 4, batch 2600, loss[loss=0.2141, simple_loss=0.2612, pruned_loss=0.08351, over 7220.00 frames. ], tot_loss[loss=0.2471, simple_loss=0.2978, pruned_loss=0.09822, over 1439997.80 frames. ], batch size: 39, lr: 3.03e-02, grad_scale: 16.0 +2023-03-20 18:54:55,811 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5768, 3.7325, 3.7352, 3.7059, 3.3776, 3.6298, 4.0978, 4.1676], + device='cuda:0'), covar=tensor([0.0367, 0.0213, 0.0247, 0.0245, 0.0493, 0.0399, 0.0295, 0.0174], + device='cuda:0'), in_proj_covar=tensor([0.0078, 0.0079, 0.0073, 0.0091, 0.0082, 0.0063, 0.0067, 0.0065], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 18:54:58,278 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=11100.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:55:00,599 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.035e+02 3.420e+02 4.169e+02 5.328e+02 1.067e+03, threshold=8.339e+02, percent-clipped=1.0 +2023-03-20 18:55:03,158 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=11110.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 18:55:04,596 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=11113.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:55:09,434 INFO [train.py:901] (0/2) Epoch 4, batch 2650, loss[loss=0.2376, simple_loss=0.2943, pruned_loss=0.09042, over 7289.00 frames. ], tot_loss[loss=0.2477, simple_loss=0.2982, pruned_loss=0.09857, over 1440707.04 frames. ], batch size: 57, lr: 3.03e-02, grad_scale: 16.0 +2023-03-20 18:55:11,991 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1357, 1.1138, 1.6114, 1.3756, 1.3543, 0.5862, 1.0309, 1.0320], + device='cuda:0'), covar=tensor([0.0455, 0.0848, 0.0178, 0.0441, 0.0381, 0.1096, 0.0282, 0.0522], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0033, 0.0025, 0.0028, 0.0025, 0.0026, 0.0030, 0.0029], + device='cuda:0'), out_proj_covar=tensor([3.9780e-05, 5.6404e-05, 3.4495e-05, 4.3433e-05, 3.9942e-05, 4.3317e-05, + 4.4048e-05, 4.3287e-05], device='cuda:0') +2023-03-20 18:55:12,915 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=11130.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:55:24,521 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.21 vs. limit=2.0 +2023-03-20 18:55:26,759 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=11158.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:55:28,132 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-03-20 18:55:31,319 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6113, 2.6441, 1.9723, 2.4958, 2.4181, 2.2855, 1.9944, 2.2962], + device='cuda:0'), covar=tensor([0.0656, 0.0266, 0.1074, 0.0823, 0.0483, 0.1819, 0.2226, 0.1049], + device='cuda:0'), in_proj_covar=tensor([0.0026, 0.0024, 0.0026, 0.0026, 0.0025, 0.0027, 0.0030, 0.0025], + device='cuda:0'), out_proj_covar=tensor([5.6119e-05, 4.6530e-05, 5.4774e-05, 5.4838e-05, 5.2725e-05, 5.4605e-05, + 6.3706e-05, 4.9971e-05], device='cuda:0') +2023-03-20 18:55:34,663 INFO [train.py:901] (0/2) Epoch 4, batch 2700, loss[loss=0.3116, simple_loss=0.3493, pruned_loss=0.137, over 6723.00 frames. ], tot_loss[loss=0.2464, simple_loss=0.2975, pruned_loss=0.09758, over 1441605.32 frames. ], batch size: 106, lr: 3.02e-02, grad_scale: 16.0 +2023-03-20 18:55:35,287 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=11174.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:55:50,738 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.463e+02 3.424e+02 4.477e+02 5.866e+02 1.423e+03, threshold=8.953e+02, percent-clipped=8.0 +2023-03-20 18:56:00,378 INFO [train.py:901] (0/2) Epoch 4, batch 2750, loss[loss=0.2735, simple_loss=0.3259, pruned_loss=0.1105, over 7309.00 frames. ], tot_loss[loss=0.246, simple_loss=0.2969, pruned_loss=0.09752, over 1441110.16 frames. ], batch size: 83, lr: 3.02e-02, grad_scale: 16.0 +2023-03-20 18:56:06,612 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=11235.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:56:09,607 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=11241.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 18:56:21,477 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.9517, 0.8097, 1.2152, 0.9547, 1.0127, 0.6467, 0.7252, 0.8211], + device='cuda:0'), covar=tensor([0.0369, 0.1336, 0.0199, 0.0268, 0.0614, 0.0861, 0.0263, 0.0218], + device='cuda:0'), in_proj_covar=tensor([0.0026, 0.0031, 0.0025, 0.0027, 0.0024, 0.0025, 0.0028, 0.0028], + device='cuda:0'), out_proj_covar=tensor([3.8130e-05, 5.4079e-05, 3.3527e-05, 4.1558e-05, 3.8184e-05, 4.1965e-05, + 4.0841e-05, 4.1331e-05], device='cuda:0') +2023-03-20 18:56:24,913 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2940, 3.4609, 3.1219, 2.9683, 3.3316, 3.3899, 3.1594, 3.1848], + device='cuda:0'), covar=tensor([0.0079, 0.0087, 0.0100, 0.0117, 0.0082, 0.0069, 0.0119, 0.0109], + device='cuda:0'), in_proj_covar=tensor([0.0023, 0.0024, 0.0026, 0.0023, 0.0022, 0.0024, 0.0029, 0.0028], + device='cuda:0'), out_proj_covar=tensor([6.4779e-05, 7.1169e-05, 8.1950e-05, 6.4324e-05, 6.6412e-05, 6.7249e-05, + 8.7824e-05, 8.2424e-05], device='cuda:0') +2023-03-20 18:56:25,287 INFO [train.py:901] (0/2) Epoch 4, batch 2800, loss[loss=0.2653, simple_loss=0.3121, pruned_loss=0.1093, over 7118.00 frames. ], tot_loss[loss=0.245, simple_loss=0.2964, pruned_loss=0.09686, over 1442940.50 frames. ], batch size: 98, lr: 3.01e-02, grad_scale: 16.0 +2023-03-20 18:56:26,906 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2129, 1.8662, 1.9865, 2.0287, 2.2827, 1.9531, 2.5456, 1.6727], + device='cuda:0'), covar=tensor([0.0294, 0.0714, 0.0628, 0.0602, 0.0769, 0.0610, 0.0302, 0.1158], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0036, 0.0032, 0.0035, 0.0035, 0.0037, 0.0032, 0.0033], + device='cuda:0'), out_proj_covar=tensor([7.7705e-05, 6.8536e-05, 6.3998e-05, 6.7342e-05, 6.9360e-05, 7.0219e-05, + 6.6245e-05, 6.4975e-05], device='cuda:0') +2023-03-20 18:56:29,779 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6878, 3.3655, 3.3621, 3.3876, 3.3391, 3.3758, 3.5458, 3.3866], + device='cuda:0'), covar=tensor([0.0099, 0.0191, 0.0166, 0.0212, 0.0229, 0.0162, 0.0193, 0.0162], + device='cuda:0'), in_proj_covar=tensor([0.0031, 0.0036, 0.0038, 0.0034, 0.0046, 0.0042, 0.0036, 0.0033], + device='cuda:0'), out_proj_covar=tensor([6.8725e-05, 8.4346e-05, 8.7416e-05, 8.0939e-05, 1.1079e-04, 1.0193e-04, + 8.9770e-05, 7.4386e-05], device='cuda:0') +2023-03-20 18:56:38,106 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-4.pt +2023-03-20 18:56:55,197 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-20 18:56:55,251 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-20 18:56:55,278 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-20 18:56:55,296 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-20 18:56:55,471 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-20 18:56:55,487 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-20 18:56:55,498 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-20 18:56:55,598 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-20 18:56:56,017 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-20 18:56:56,034 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-20 18:56:56,050 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-20 18:56:56,214 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-20 18:56:56,220 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-20 18:56:56,227 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-20 18:56:56,368 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-20 18:56:56,371 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-20 18:56:56,430 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-20 18:56:58,651 INFO [train.py:901] (0/2) Epoch 5, batch 0, loss[loss=0.2655, simple_loss=0.3146, pruned_loss=0.1082, over 7281.00 frames. ], tot_loss[loss=0.2655, simple_loss=0.3146, pruned_loss=0.1082, over 7281.00 frames. ], batch size: 77, lr: 2.90e-02, grad_scale: 16.0 +2023-03-20 18:56:58,652 INFO [train.py:926] (0/2) Computing validation loss +2023-03-20 18:57:14,106 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.9714, 0.9210, 1.1705, 1.1474, 1.0288, 0.5769, 0.8324, 0.9545], + device='cuda:0'), covar=tensor([0.0386, 0.0954, 0.0165, 0.0299, 0.0688, 0.1206, 0.0238, 0.0390], + device='cuda:0'), in_proj_covar=tensor([0.0026, 0.0031, 0.0024, 0.0027, 0.0024, 0.0026, 0.0028, 0.0028], + device='cuda:0'), out_proj_covar=tensor([3.7796e-05, 5.3574e-05, 3.3089e-05, 4.1911e-05, 3.8912e-05, 4.2765e-05, + 4.1202e-05, 4.1520e-05], device='cuda:0') +2023-03-20 18:57:20,089 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2301, 4.2240, 4.2287, 3.7462, 4.0841, 3.5837, 2.8042, 4.5993], + device='cuda:0'), covar=tensor([0.0011, 0.0076, 0.0039, 0.0130, 0.0018, 0.0303, 0.0750, 0.0033], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0042, 0.0054, 0.0044, 0.0043, 0.0066, 0.0087, 0.0047], + device='cuda:0'), out_proj_covar=tensor([5.0941e-05, 6.6504e-05, 7.8384e-05, 6.4523e-05, 5.3866e-05, 9.7078e-05, + 1.2793e-04, 6.5185e-05], device='cuda:0') +2023-03-20 18:57:21,762 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6104, 2.0948, 2.2571, 2.3320, 1.6746, 2.5480, 2.2640, 2.7054], + device='cuda:0'), covar=tensor([0.0034, 0.0661, 0.1861, 0.0028, 0.4786, 0.0055, 0.0737, 0.0041], + device='cuda:0'), in_proj_covar=tensor([0.0111, 0.0181, 0.0280, 0.0109, 0.0274, 0.0107, 0.0199, 0.0110], + device='cuda:0'), out_proj_covar=tensor([9.3705e-05, 1.5703e-04, 2.2274e-04, 9.5719e-05, 2.2956e-04, 9.6580e-05, + 1.6927e-04, 9.9041e-05], device='cuda:0') +2023-03-20 18:57:24,431 INFO [train.py:935] (0/2) Epoch 5, validation: loss=0.199, simple_loss=0.2825, pruned_loss=0.05778, over 1622729.00 frames. +2023-03-20 18:57:24,432 INFO [train.py:936] (0/2) Maximum memory allocated so far is 11856MB +2023-03-20 18:57:25,576 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1719, 1.8990, 1.9617, 1.9612, 2.2211, 1.9562, 2.4247, 1.6413], + device='cuda:0'), covar=tensor([0.0598, 0.0522, 0.0585, 0.0564, 0.0327, 0.0372, 0.0320, 0.0661], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0036, 0.0031, 0.0035, 0.0034, 0.0036, 0.0032, 0.0033], + device='cuda:0'), out_proj_covar=tensor([7.5958e-05, 6.9333e-05, 6.2876e-05, 6.7553e-05, 6.8749e-05, 6.9185e-05, + 6.5527e-05, 6.4656e-05], device='cuda:0') +2023-03-20 18:57:28,426 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.740e+02 3.320e+02 4.022e+02 5.081e+02 1.321e+03, threshold=8.045e+02, percent-clipped=4.0 +2023-03-20 18:57:30,945 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-20 18:57:41,996 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-20 18:57:48,510 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-20 18:57:49,494 INFO [train.py:901] (0/2) Epoch 5, batch 50, loss[loss=0.2154, simple_loss=0.2772, pruned_loss=0.07677, over 7284.00 frames. ], tot_loss[loss=0.2388, simple_loss=0.2911, pruned_loss=0.09327, over 324981.69 frames. ], batch size: 70, lr: 2.90e-02, grad_scale: 16.0 +2023-03-20 18:57:50,512 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-20 18:57:53,507 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-20 18:58:10,842 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-20 18:58:11,353 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-20 18:58:15,838 INFO [train.py:901] (0/2) Epoch 5, batch 100, loss[loss=0.2524, simple_loss=0.3009, pruned_loss=0.1019, over 7371.00 frames. ], tot_loss[loss=0.2422, simple_loss=0.2947, pruned_loss=0.09486, over 573172.70 frames. ], batch size: 51, lr: 2.89e-02, grad_scale: 16.0 +2023-03-20 18:58:17,827 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=11400.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:58:20,160 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.320e+02 3.482e+02 4.365e+02 5.725e+02 1.315e+03, threshold=8.730e+02, percent-clipped=7.0 +2023-03-20 18:58:32,668 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=11430.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:58:41,104 INFO [train.py:901] (0/2) Epoch 5, batch 150, loss[loss=0.2482, simple_loss=0.3077, pruned_loss=0.09432, over 7307.00 frames. ], tot_loss[loss=0.2424, simple_loss=0.2952, pruned_loss=0.09475, over 767920.77 frames. ], batch size: 49, lr: 2.89e-02, grad_scale: 16.0 +2023-03-20 18:58:41,666 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=11448.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:58:58,015 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=11478.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:59:06,803 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.41 vs. limit=5.0 +2023-03-20 18:59:07,527 INFO [train.py:901] (0/2) Epoch 5, batch 200, loss[loss=0.2781, simple_loss=0.3214, pruned_loss=0.1173, over 7278.00 frames. ], tot_loss[loss=0.2406, simple_loss=0.2933, pruned_loss=0.0939, over 916156.00 frames. ], batch size: 52, lr: 2.88e-02, grad_scale: 16.0 +2023-03-20 18:59:11,555 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.950e+02 3.488e+02 4.378e+02 5.184e+02 9.944e+02, threshold=8.756e+02, percent-clipped=2.0 +2023-03-20 18:59:12,084 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-20 18:59:16,513 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-20 18:59:23,545 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-20 18:59:24,131 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=11530.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:59:25,210 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=11532.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:59:30,294 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=11541.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 18:59:32,530 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.79 vs. limit=5.0 +2023-03-20 18:59:33,190 INFO [train.py:901] (0/2) Epoch 5, batch 250, loss[loss=0.2356, simple_loss=0.2958, pruned_loss=0.08772, over 7216.00 frames. ], tot_loss[loss=0.2394, simple_loss=0.2926, pruned_loss=0.09307, over 1031172.37 frames. ], batch size: 50, lr: 2.87e-02, grad_scale: 16.0 +2023-03-20 18:59:36,044 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.94 vs. limit=2.0 +2023-03-20 18:59:36,268 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-20 18:59:37,179 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.87 vs. limit=2.0 +2023-03-20 18:59:43,597 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.7698, 1.5035, 1.3338, 0.9764, 1.1251, 2.1083, 1.4449, 1.2007], + device='cuda:0'), covar=tensor([0.1153, 0.0644, 0.0394, 0.0292, 0.0741, 0.0150, 0.0271, 0.0439], + device='cuda:0'), in_proj_covar=tensor([0.0021, 0.0022, 0.0022, 0.0020, 0.0021, 0.0021, 0.0020, 0.0021], + device='cuda:0'), out_proj_covar=tensor([4.2059e-05, 4.3172e-05, 4.0272e-05, 3.5734e-05, 4.3315e-05, 4.2129e-05, + 3.8303e-05, 4.3977e-05], device='cuda:0') +2023-03-20 18:59:54,823 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=11589.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 18:59:56,854 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=11593.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 18:59:57,256 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-20 18:59:58,662 INFO [train.py:901] (0/2) Epoch 5, batch 300, loss[loss=0.2261, simple_loss=0.2862, pruned_loss=0.083, over 7316.00 frames. ], tot_loss[loss=0.2401, simple_loss=0.2933, pruned_loss=0.09341, over 1122702.41 frames. ], batch size: 83, lr: 2.87e-02, grad_scale: 16.0 +2023-03-20 19:00:03,002 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.958e+02 3.432e+02 4.391e+02 5.389e+02 9.062e+02, threshold=8.782e+02, percent-clipped=1.0 +2023-03-20 19:00:03,895 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.76 vs. limit=5.0 +2023-03-20 19:00:06,023 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-20 19:00:17,860 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.97 vs. limit=2.0 +2023-03-20 19:00:18,636 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=11635.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:00:25,219 INFO [train.py:901] (0/2) Epoch 5, batch 350, loss[loss=0.2566, simple_loss=0.3031, pruned_loss=0.105, over 7279.00 frames. ], tot_loss[loss=0.2402, simple_loss=0.2931, pruned_loss=0.09365, over 1193604.43 frames. ], batch size: 57, lr: 2.86e-02, grad_scale: 16.0 +2023-03-20 19:00:32,312 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.8444, 0.7500, 1.2355, 1.4568, 0.9603, 0.6581, 0.7962, 0.9040], + device='cuda:0'), covar=tensor([0.0933, 0.1907, 0.0406, 0.0344, 0.1114, 0.1240, 0.0416, 0.0611], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0032, 0.0026, 0.0027, 0.0028, 0.0026, 0.0029, 0.0030], + device='cuda:0'), out_proj_covar=tensor([4.0765e-05, 5.6155e-05, 3.5446e-05, 4.2445e-05, 4.4138e-05, 4.4816e-05, + 4.3962e-05, 4.4720e-05], device='cuda:0') +2023-03-20 19:00:38,019 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.14 vs. limit=2.0 +2023-03-20 19:00:41,250 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-20 19:00:49,972 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=11696.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:00:50,325 INFO [train.py:901] (0/2) Epoch 5, batch 400, loss[loss=0.2009, simple_loss=0.2599, pruned_loss=0.07092, over 7152.00 frames. ], tot_loss[loss=0.239, simple_loss=0.2921, pruned_loss=0.09295, over 1249265.12 frames. ], batch size: 39, lr: 2.86e-02, grad_scale: 16.0 +2023-03-20 19:00:51,469 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=11699.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:00:54,172 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.098e+02 3.501e+02 4.253e+02 5.454e+02 9.095e+02, threshold=8.507e+02, percent-clipped=1.0 +2023-03-20 19:01:12,360 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.7643, 1.7800, 1.4740, 1.2044, 1.2314, 1.8311, 1.5020, 1.1665], + device='cuda:0'), covar=tensor([0.1218, 0.0281, 0.0544, 0.0238, 0.0843, 0.0500, 0.0392, 0.0500], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0021, 0.0020, 0.0019, 0.0020, 0.0019, 0.0019, 0.0019], + device='cuda:0'), out_proj_covar=tensor([3.9559e-05, 4.0802e-05, 3.6617e-05, 3.4385e-05, 4.0375e-05, 3.9079e-05, + 3.7231e-05, 4.0583e-05], device='cuda:0') +2023-03-20 19:01:16,685 INFO [train.py:901] (0/2) Epoch 5, batch 450, loss[loss=0.2408, simple_loss=0.2977, pruned_loss=0.09192, over 7316.00 frames. ], tot_loss[loss=0.2397, simple_loss=0.2925, pruned_loss=0.09348, over 1294448.00 frames. ], batch size: 80, lr: 2.85e-02, grad_scale: 16.0 +2023-03-20 19:01:22,749 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-20 19:01:22,764 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-20 19:01:23,421 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=11760.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:01:41,722 INFO [train.py:901] (0/2) Epoch 5, batch 500, loss[loss=0.2394, simple_loss=0.2924, pruned_loss=0.09322, over 7319.00 frames. ], tot_loss[loss=0.2393, simple_loss=0.2926, pruned_loss=0.09297, over 1328969.17 frames. ], batch size: 83, lr: 2.85e-02, grad_scale: 16.0 +2023-03-20 19:01:46,705 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.845e+02 3.462e+02 4.263e+02 5.633e+02 1.413e+03, threshold=8.525e+02, percent-clipped=5.0 +2023-03-20 19:01:52,853 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=11816.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:01:54,733 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-20 19:01:55,470 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.54 vs. limit=2.0 +2023-03-20 19:01:56,241 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-20 19:01:56,762 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-20 19:01:59,272 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-20 19:01:59,902 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=11830.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:02:04,234 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-20 19:02:07,829 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=11846.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:02:08,201 INFO [train.py:901] (0/2) Epoch 5, batch 550, loss[loss=0.2725, simple_loss=0.3177, pruned_loss=0.1136, over 7267.00 frames. ], tot_loss[loss=0.2399, simple_loss=0.2927, pruned_loss=0.0935, over 1352621.78 frames. ], batch size: 57, lr: 2.84e-02, grad_scale: 16.0 +2023-03-20 19:02:14,274 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-20 19:02:14,377 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2223, 3.9855, 3.9752, 3.5685, 3.7167, 2.4702, 1.7710, 4.2985], + device='cuda:0'), covar=tensor([0.0015, 0.0028, 0.0056, 0.0063, 0.0023, 0.0388, 0.0915, 0.0037], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0042, 0.0056, 0.0043, 0.0044, 0.0065, 0.0085, 0.0044], + device='cuda:0'), out_proj_covar=tensor([5.2093e-05, 6.4014e-05, 7.9843e-05, 6.4118e-05, 5.4197e-05, 9.7196e-05, + 1.2560e-04, 6.0001e-05], device='cuda:0') +2023-03-20 19:02:22,371 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-20 19:02:23,524 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=11877.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:02:23,927 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=11878.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:02:24,023 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.8945, 0.8470, 1.2526, 1.3721, 0.9712, 0.9355, 0.8339, 0.8962], + device='cuda:0'), covar=tensor([0.0405, 0.1418, 0.0226, 0.0323, 0.0584, 0.0486, 0.0182, 0.0321], + device='cuda:0'), in_proj_covar=tensor([0.0026, 0.0030, 0.0025, 0.0026, 0.0026, 0.0025, 0.0028, 0.0028], + device='cuda:0'), out_proj_covar=tensor([4.0256e-05, 5.3661e-05, 3.5547e-05, 4.0848e-05, 4.2392e-05, 4.2390e-05, + 4.2800e-05, 4.2892e-05], device='cuda:0') +2023-03-20 19:02:25,851 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-20 19:02:29,611 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=11888.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:02:30,400 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.83 vs. limit=5.0 +2023-03-20 19:02:33,600 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-20 19:02:34,080 INFO [train.py:901] (0/2) Epoch 5, batch 600, loss[loss=0.2118, simple_loss=0.2708, pruned_loss=0.07644, over 7247.00 frames. ], tot_loss[loss=0.2396, simple_loss=0.2924, pruned_loss=0.09336, over 1371096.44 frames. ], batch size: 47, lr: 2.84e-02, grad_scale: 32.0 +2023-03-20 19:02:38,652 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.069e+02 3.397e+02 4.178e+02 5.537e+02 1.097e+03, threshold=8.356e+02, percent-clipped=4.0 +2023-03-20 19:02:39,810 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=11907.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:02:46,291 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=11920.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:02:49,175 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-20 19:02:57,601 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-20 19:02:59,596 INFO [train.py:901] (0/2) Epoch 5, batch 650, loss[loss=0.2406, simple_loss=0.295, pruned_loss=0.09308, over 7230.00 frames. ], tot_loss[loss=0.2395, simple_loss=0.2923, pruned_loss=0.09336, over 1386078.84 frames. ], batch size: 93, lr: 2.83e-02, grad_scale: 32.0 +2023-03-20 19:03:00,238 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=11948.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:03:12,515 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.58 vs. limit=5.0 +2023-03-20 19:03:16,866 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-20 19:03:17,479 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=11981.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:03:18,904 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=11984.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:03:22,869 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=11991.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:03:25,765 INFO [train.py:901] (0/2) Epoch 5, batch 700, loss[loss=0.2044, simple_loss=0.2617, pruned_loss=0.07354, over 7310.00 frames. ], tot_loss[loss=0.2396, simple_loss=0.2922, pruned_loss=0.09349, over 1397709.08 frames. ], batch size: 42, lr: 2.83e-02, grad_scale: 32.0 +2023-03-20 19:03:26,265 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-20 19:03:27,576 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-12000.pt +2023-03-20 19:03:33,386 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.580e+02 3.642e+02 4.392e+02 5.273e+02 9.555e+02, threshold=8.784e+02, percent-clipped=4.0 +2023-03-20 19:03:35,605 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=12009.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:03:52,605 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-20 19:03:53,080 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-20 19:03:53,707 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=12045.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:03:54,570 INFO [train.py:901] (0/2) Epoch 5, batch 750, loss[loss=0.2447, simple_loss=0.3051, pruned_loss=0.09217, over 7251.00 frames. ], tot_loss[loss=0.24, simple_loss=0.293, pruned_loss=0.09353, over 1408456.80 frames. ], batch size: 89, lr: 2.82e-02, grad_scale: 32.0 +2023-03-20 19:03:58,707 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=12055.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:04:00,720 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3946, 4.8698, 4.8361, 4.7869, 4.6052, 4.3723, 4.8424, 4.7256], + device='cuda:0'), covar=tensor([0.0489, 0.0318, 0.0528, 0.0483, 0.0343, 0.0299, 0.0364, 0.0473], + device='cuda:0'), in_proj_covar=tensor([0.0100, 0.0086, 0.0096, 0.0079, 0.0072, 0.0093, 0.0084, 0.0074], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0002, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 19:04:06,073 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.95 vs. limit=2.0 +2023-03-20 19:04:08,189 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-20 19:04:09,779 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=12076.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:04:13,234 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-20 19:04:18,857 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-20 19:04:20,818 INFO [train.py:901] (0/2) Epoch 5, batch 800, loss[loss=0.2256, simple_loss=0.2894, pruned_loss=0.08092, over 7299.00 frames. ], tot_loss[loss=0.2381, simple_loss=0.2915, pruned_loss=0.0924, over 1415803.42 frames. ], batch size: 68, lr: 2.82e-02, grad_scale: 32.0 +2023-03-20 19:04:20,838 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-20 19:04:24,790 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.858e+02 3.401e+02 4.258e+02 5.414e+02 1.071e+03, threshold=8.517e+02, percent-clipped=1.0 +2023-03-20 19:04:31,253 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-20 19:04:40,882 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=12137.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:04:43,488 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.02 vs. limit=2.0 +2023-03-20 19:04:45,701 INFO [train.py:901] (0/2) Epoch 5, batch 850, loss[loss=0.2417, simple_loss=0.296, pruned_loss=0.09372, over 7303.00 frames. ], tot_loss[loss=0.2371, simple_loss=0.2908, pruned_loss=0.09172, over 1421935.53 frames. ], batch size: 80, lr: 2.81e-02, grad_scale: 32.0 +2023-03-20 19:04:50,327 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-20 19:04:50,337 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-20 19:04:55,913 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-20 19:04:59,478 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=12172.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:04:59,931 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-20 19:05:07,429 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=12188.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:05:11,848 INFO [train.py:901] (0/2) Epoch 5, batch 900, loss[loss=0.2254, simple_loss=0.2862, pruned_loss=0.0823, over 7327.00 frames. ], tot_loss[loss=0.2367, simple_loss=0.2906, pruned_loss=0.09142, over 1426281.09 frames. ], batch size: 54, lr: 2.81e-02, grad_scale: 32.0 +2023-03-20 19:05:14,773 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=12202.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:05:16,175 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.142e+02 3.418e+02 4.341e+02 5.122e+02 8.381e+02, threshold=8.682e+02, percent-clipped=0.0 +2023-03-20 19:05:18,368 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0695, 2.1265, 2.3203, 3.2301, 3.1058, 3.2803, 2.9099, 2.8691], + device='cuda:0'), covar=tensor([0.0836, 0.0410, 0.1062, 0.0124, 0.0041, 0.0052, 0.0064, 0.0061], + device='cuda:0'), in_proj_covar=tensor([0.0246, 0.0170, 0.0219, 0.0132, 0.0089, 0.0089, 0.0086, 0.0092], + device='cuda:0'), out_proj_covar=tensor([2.1147e-04, 1.6085e-04, 1.9599e-04, 1.3042e-04, 8.5311e-05, 8.6081e-05, + 8.5518e-05, 8.9902e-05], device='cuda:0') +2023-03-20 19:05:19,759 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0391, 0.9691, 1.2913, 1.1452, 0.6288, 0.8756, 0.6162, 0.8746], + device='cuda:0'), covar=tensor([0.0793, 0.1307, 0.0490, 0.0326, 0.2060, 0.1493, 0.0287, 0.0769], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0030, 0.0027, 0.0029, 0.0029, 0.0028, 0.0029, 0.0030], + device='cuda:0'), out_proj_covar=tensor([4.3036e-05, 5.4671e-05, 3.8701e-05, 4.5623e-05, 4.6759e-05, 4.6769e-05, + 4.3510e-05, 4.8353e-05], device='cuda:0') +2023-03-20 19:05:31,676 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=12236.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:05:37,752 INFO [train.py:901] (0/2) Epoch 5, batch 950, loss[loss=0.2335, simple_loss=0.2918, pruned_loss=0.08753, over 7276.00 frames. ], tot_loss[loss=0.2363, simple_loss=0.2905, pruned_loss=0.0911, over 1430459.87 frames. ], batch size: 77, lr: 2.80e-02, grad_scale: 32.0 +2023-03-20 19:05:38,175 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-03-20 19:05:38,795 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-20 19:05:43,619 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6928, 1.6819, 1.6596, 1.8056, 1.9844, 1.8696, 2.0080, 1.8492], + device='cuda:0'), covar=tensor([0.1601, 0.0544, 0.1235, 0.0430, 0.0613, 0.0517, 0.0496, 0.0581], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0032, 0.0034, 0.0032, 0.0033, 0.0037, 0.0031, 0.0032], + device='cuda:0'), out_proj_covar=tensor([8.0354e-05, 6.4824e-05, 7.0434e-05, 6.5884e-05, 6.9354e-05, 7.5537e-05, + 6.7998e-05, 6.5840e-05], device='cuda:0') +2023-03-20 19:05:52,948 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=12276.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:06:00,450 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=12291.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:06:01,838 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-20 19:06:03,342 INFO [train.py:901] (0/2) Epoch 5, batch 1000, loss[loss=0.197, simple_loss=0.2436, pruned_loss=0.07519, over 6038.00 frames. ], tot_loss[loss=0.2361, simple_loss=0.2904, pruned_loss=0.0909, over 1434938.45 frames. ], batch size: 26, lr: 2.80e-02, grad_scale: 32.0 +2023-03-20 19:06:06,902 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=12304.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:06:07,289 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.908e+02 3.426e+02 4.315e+02 5.528e+02 1.209e+03, threshold=8.630e+02, percent-clipped=3.0 +2023-03-20 19:06:22,115 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2709, 3.8636, 3.8626, 3.7614, 3.8579, 3.8377, 4.1406, 3.6775], + device='cuda:0'), covar=tensor([0.0099, 0.0140, 0.0161, 0.0168, 0.0182, 0.0155, 0.0169, 0.0165], + device='cuda:0'), in_proj_covar=tensor([0.0034, 0.0039, 0.0043, 0.0036, 0.0051, 0.0047, 0.0039, 0.0039], + device='cuda:0'), out_proj_covar=tensor([7.5393e-05, 9.5927e-05, 1.0203e-04, 8.8716e-05, 1.2500e-04, 1.1731e-04, + 1.0205e-04, 9.0323e-05], device='cuda:0') +2023-03-20 19:06:22,523 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-20 19:06:25,594 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=12339.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:06:25,714 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4300, 2.4853, 3.0342, 3.6181, 3.2726, 3.5563, 3.5615, 3.3239], + device='cuda:0'), covar=tensor([0.1134, 0.0520, 0.1263, 0.0118, 0.0055, 0.0028, 0.0050, 0.0076], + device='cuda:0'), in_proj_covar=tensor([0.0253, 0.0174, 0.0225, 0.0134, 0.0092, 0.0091, 0.0087, 0.0092], + device='cuda:0'), out_proj_covar=tensor([2.1689e-04, 1.6484e-04, 2.0104e-04, 1.3170e-04, 8.8812e-05, 8.8477e-05, + 8.5324e-05, 9.1293e-05], device='cuda:0') +2023-03-20 19:06:26,125 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=12340.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:06:26,214 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.9362, 0.8329, 1.0057, 0.7675, 1.2330, 1.0867, 0.6001, 0.6124], + device='cuda:0'), covar=tensor([0.0189, 0.0264, 0.0245, 0.0191, 0.0111, 0.0248, 0.0293, 0.0335], + device='cuda:0'), in_proj_covar=tensor([0.0016, 0.0017, 0.0015, 0.0016, 0.0017, 0.0016, 0.0017, 0.0017], + device='cuda:0'), out_proj_covar=tensor([1.8929e-05, 1.9771e-05, 2.0705e-05, 2.0334e-05, 1.9435e-05, 2.1040e-05, + 2.2905e-05, 2.4119e-05], device='cuda:0') +2023-03-20 19:06:26,375 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.90 vs. limit=2.0 +2023-03-20 19:06:29,300 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.54 vs. limit=2.0 +2023-03-20 19:06:29,541 INFO [train.py:901] (0/2) Epoch 5, batch 1050, loss[loss=0.1992, simple_loss=0.2639, pruned_loss=0.0672, over 7352.00 frames. ], tot_loss[loss=0.236, simple_loss=0.2899, pruned_loss=0.09107, over 1437262.53 frames. ], batch size: 44, lr: 2.79e-02, grad_scale: 32.0 +2023-03-20 19:06:33,660 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=12355.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:06:44,517 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-20 19:06:47,991 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-20 19:06:54,360 INFO [train.py:901] (0/2) Epoch 5, batch 1100, loss[loss=0.2553, simple_loss=0.3115, pruned_loss=0.09951, over 7258.00 frames. ], tot_loss[loss=0.2364, simple_loss=0.2902, pruned_loss=0.09129, over 1438078.91 frames. ], batch size: 55, lr: 2.79e-02, grad_scale: 32.0 +2023-03-20 19:06:56,833 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9832, 4.0711, 4.1895, 4.4892, 4.6095, 4.5187, 3.8808, 4.0543], + device='cuda:0'), covar=tensor([0.0838, 0.1555, 0.1607, 0.1095, 0.0504, 0.1095, 0.0655, 0.0822], + device='cuda:0'), in_proj_covar=tensor([0.0108, 0.0181, 0.0172, 0.0154, 0.0126, 0.0183, 0.0101, 0.0125], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 19:06:57,782 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=12403.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:06:58,694 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.093e+02 3.365e+02 4.279e+02 5.255e+02 8.935e+02, threshold=8.557e+02, percent-clipped=2.0 +2023-03-20 19:07:09,451 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0810, 3.9018, 3.6931, 3.2631, 3.4139, 2.2765, 1.8437, 4.1396], + device='cuda:0'), covar=tensor([0.0014, 0.0058, 0.0079, 0.0064, 0.0055, 0.0461, 0.0827, 0.0030], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0044, 0.0059, 0.0045, 0.0047, 0.0068, 0.0085, 0.0047], + device='cuda:0'), out_proj_covar=tensor([5.2705e-05, 6.6999e-05, 8.4239e-05, 6.6384e-05, 5.9309e-05, 1.0128e-04, + 1.2766e-04, 6.3549e-05], device='cuda:0') +2023-03-20 19:07:13,498 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=12432.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:07:18,382 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-20 19:07:18,389 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 19:07:20,467 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=12446.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:07:20,807 INFO [train.py:901] (0/2) Epoch 5, batch 1150, loss[loss=0.3128, simple_loss=0.3413, pruned_loss=0.1421, over 6853.00 frames. ], tot_loss[loss=0.2368, simple_loss=0.2905, pruned_loss=0.09157, over 1439145.75 frames. ], batch size: 107, lr: 2.78e-02, grad_scale: 32.0 +2023-03-20 19:07:22,431 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4705, 2.3116, 1.9414, 2.6780, 1.3781, 2.3799, 1.3072, 3.1174], + device='cuda:0'), covar=tensor([0.0058, 0.0532, 0.2156, 0.0041, 0.5114, 0.0065, 0.1045, 0.0053], + device='cuda:0'), in_proj_covar=tensor([0.0113, 0.0186, 0.0286, 0.0106, 0.0282, 0.0106, 0.0209, 0.0111], + device='cuda:0'), out_proj_covar=tensor([9.7852e-05, 1.6440e-04, 2.2957e-04, 9.6353e-05, 2.4051e-04, 9.8317e-05, + 1.8084e-04, 1.0057e-04], device='cuda:0') +2023-03-20 19:07:30,237 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-20 19:07:30,711 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-20 19:07:33,383 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=12472.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:07:38,955 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9348, 2.2622, 1.6806, 1.6957, 1.9470, 2.0599, 2.1622, 1.8901], + device='cuda:0'), covar=tensor([0.1955, 0.0414, 0.0939, 0.0689, 0.0928, 0.0530, 0.0536, 0.0862], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0033, 0.0033, 0.0033, 0.0034, 0.0038, 0.0033, 0.0033], + device='cuda:0'), out_proj_covar=tensor([7.7568e-05, 6.7842e-05, 7.0518e-05, 6.8094e-05, 7.2784e-05, 7.7489e-05, + 7.1675e-05, 6.9091e-05], device='cuda:0') +2023-03-20 19:07:44,581 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.04 vs. limit=2.0 +2023-03-20 19:07:45,729 INFO [train.py:901] (0/2) Epoch 5, batch 1200, loss[loss=0.253, simple_loss=0.3009, pruned_loss=0.1025, over 7272.00 frames. ], tot_loss[loss=0.2371, simple_loss=0.2907, pruned_loss=0.09177, over 1439027.80 frames. ], batch size: 57, lr: 2.78e-02, grad_scale: 32.0 +2023-03-20 19:07:48,908 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=12502.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:07:50,273 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.922e+02 3.243e+02 4.028e+02 5.026e+02 9.782e+02, threshold=8.056e+02, percent-clipped=2.0 +2023-03-20 19:07:51,448 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=12507.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:07:58,506 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=12520.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:08:00,079 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.9572, 0.7586, 1.3106, 1.0000, 0.8788, 0.7624, 0.9647, 0.7852], + device='cuda:0'), covar=tensor([0.0893, 0.1885, 0.0204, 0.0284, 0.1086, 0.0677, 0.0302, 0.0984], + device='cuda:0'), in_proj_covar=tensor([0.0031, 0.0034, 0.0027, 0.0031, 0.0030, 0.0029, 0.0030, 0.0031], + device='cuda:0'), out_proj_covar=tensor([4.8400e-05, 5.9837e-05, 3.9281e-05, 4.8555e-05, 4.9545e-05, 4.8849e-05, + 4.5251e-05, 5.0287e-05], device='cuda:0') +2023-03-20 19:08:00,700 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=6.52 vs. limit=5.0 +2023-03-20 19:08:00,723 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.78 vs. limit=5.0 +2023-03-20 19:08:05,421 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-20 19:08:06,693 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.41 vs. limit=5.0 +2023-03-20 19:08:11,909 INFO [train.py:901] (0/2) Epoch 5, batch 1250, loss[loss=0.234, simple_loss=0.2981, pruned_loss=0.08493, over 7256.00 frames. ], tot_loss[loss=0.2371, simple_loss=0.2905, pruned_loss=0.0919, over 1438300.98 frames. ], batch size: 47, lr: 2.77e-02, grad_scale: 32.0 +2023-03-20 19:08:13,458 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=12550.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:08:26,597 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=12576.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:08:27,634 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4988, 3.7451, 3.1625, 3.5077, 3.5456, 3.7339, 3.4531, 3.0852], + device='cuda:0'), covar=tensor([0.0090, 0.0131, 0.0154, 0.0111, 0.0121, 0.0090, 0.0137, 0.0165], + device='cuda:0'), in_proj_covar=tensor([0.0026, 0.0029, 0.0030, 0.0026, 0.0027, 0.0029, 0.0035, 0.0033], + device='cuda:0'), out_proj_covar=tensor([7.5837e-05, 9.0487e-05, 1.0423e-04, 7.5680e-05, 8.6729e-05, 8.7380e-05, + 1.0995e-04, 1.0416e-04], device='cuda:0') +2023-03-20 19:08:28,025 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-20 19:08:32,643 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-20 19:08:33,661 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-20 19:08:37,212 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6891, 5.0854, 5.0918, 5.0311, 4.7405, 4.6216, 5.1485, 4.8404], + device='cuda:0'), covar=tensor([0.0380, 0.0375, 0.0586, 0.0399, 0.0451, 0.0284, 0.0379, 0.0608], + device='cuda:0'), in_proj_covar=tensor([0.0101, 0.0088, 0.0097, 0.0078, 0.0074, 0.0094, 0.0087, 0.0075], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0001, 0.0002, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 19:08:37,625 INFO [train.py:901] (0/2) Epoch 5, batch 1300, loss[loss=0.2421, simple_loss=0.3002, pruned_loss=0.09197, over 7230.00 frames. ], tot_loss[loss=0.2387, simple_loss=0.2919, pruned_loss=0.09276, over 1439420.29 frames. ], batch size: 45, lr: 2.77e-02, grad_scale: 32.0 +2023-03-20 19:08:42,124 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=12604.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:08:42,500 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.705e+02 3.330e+02 4.266e+02 5.499e+02 1.083e+03, threshold=8.532e+02, percent-clipped=4.0 +2023-03-20 19:08:52,196 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=12624.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:08:52,330 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9931, 3.6711, 3.3959, 3.9611, 2.8174, 2.5175, 3.9841, 3.2025], + device='cuda:0'), covar=tensor([0.0057, 0.0054, 0.0094, 0.0021, 0.0144, 0.0300, 0.0084, 0.0285], + device='cuda:0'), in_proj_covar=tensor([0.0153, 0.0137, 0.0189, 0.0122, 0.0227, 0.0265, 0.0160, 0.0265], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0002, 0.0001, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:0') +2023-03-20 19:08:54,323 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.4259, 0.7409, 0.4899, 0.4226, 0.7424, 0.7084, 0.2968, 0.5590], + device='cuda:0'), covar=tensor([0.0205, 0.0190, 0.0278, 0.0257, 0.0176, 0.0289, 0.0568, 0.0351], + device='cuda:0'), in_proj_covar=tensor([0.0016, 0.0017, 0.0015, 0.0015, 0.0017, 0.0018, 0.0017, 0.0018], + device='cuda:0'), out_proj_covar=tensor([1.9868e-05, 1.9630e-05, 2.0741e-05, 2.0222e-05, 2.0207e-05, 2.2820e-05, + 2.3036e-05, 2.6010e-05], device='cuda:0') +2023-03-20 19:08:58,123 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-20 19:09:00,154 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=12640.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:09:00,558 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-20 19:09:03,586 INFO [train.py:901] (0/2) Epoch 5, batch 1350, loss[loss=0.2003, simple_loss=0.245, pruned_loss=0.07776, over 7009.00 frames. ], tot_loss[loss=0.2377, simple_loss=0.2909, pruned_loss=0.09225, over 1438468.43 frames. ], batch size: 35, lr: 2.77e-02, grad_scale: 32.0 +2023-03-20 19:09:04,107 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-20 19:09:06,152 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=12652.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:09:13,526 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-20 19:09:24,677 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=12688.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:09:29,645 INFO [train.py:901] (0/2) Epoch 5, batch 1400, loss[loss=0.2288, simple_loss=0.2878, pruned_loss=0.08493, over 7257.00 frames. ], tot_loss[loss=0.2378, simple_loss=0.2912, pruned_loss=0.09225, over 1439211.57 frames. ], batch size: 55, lr: 2.76e-02, grad_scale: 32.0 +2023-03-20 19:09:33,684 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.299e+02 3.528e+02 4.338e+02 5.415e+02 8.083e+02, threshold=8.676e+02, percent-clipped=0.0 +2023-03-20 19:09:47,226 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-20 19:09:47,300 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=12732.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:09:54,832 INFO [train.py:901] (0/2) Epoch 5, batch 1450, loss[loss=0.2055, simple_loss=0.2642, pruned_loss=0.07337, over 7326.00 frames. ], tot_loss[loss=0.2368, simple_loss=0.2905, pruned_loss=0.09152, over 1440395.51 frames. ], batch size: 44, lr: 2.76e-02, grad_scale: 32.0 +2023-03-20 19:09:59,036 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4507, 3.4942, 3.2488, 3.3852, 3.6308, 3.5322, 3.3884, 3.1936], + device='cuda:0'), covar=tensor([0.0074, 0.0104, 0.0085, 0.0081, 0.0073, 0.0090, 0.0107, 0.0122], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0028, 0.0029, 0.0026, 0.0026, 0.0028, 0.0035, 0.0032], + device='cuda:0'), out_proj_covar=tensor([7.4347e-05, 8.7039e-05, 9.9686e-05, 7.6182e-05, 8.4834e-05, 8.5418e-05, + 1.0854e-04, 1.0134e-04], device='cuda:0') +2023-03-20 19:10:02,542 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6843, 5.1121, 5.1683, 5.0319, 4.7005, 4.5635, 5.2034, 4.8949], + device='cuda:0'), covar=tensor([0.0336, 0.0319, 0.0373, 0.0438, 0.0424, 0.0307, 0.0297, 0.0455], + device='cuda:0'), in_proj_covar=tensor([0.0099, 0.0088, 0.0093, 0.0076, 0.0071, 0.0094, 0.0086, 0.0073], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 19:10:02,583 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2368, 4.0634, 3.9580, 3.5561, 3.6719, 2.7303, 1.8412, 4.3471], + device='cuda:0'), covar=tensor([0.0014, 0.0119, 0.0067, 0.0074, 0.0046, 0.0412, 0.0866, 0.0035], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0046, 0.0060, 0.0049, 0.0050, 0.0073, 0.0089, 0.0050], + device='cuda:0'), out_proj_covar=tensor([5.4260e-05, 6.9665e-05, 8.6288e-05, 7.1318e-05, 6.5681e-05, 1.0929e-04, + 1.3314e-04, 6.9114e-05], device='cuda:0') +2023-03-20 19:10:03,102 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=12763.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:10:11,854 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-20 19:10:12,889 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=12780.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:10:21,347 INFO [train.py:901] (0/2) Epoch 5, batch 1500, loss[loss=0.2119, simple_loss=0.2641, pruned_loss=0.07987, over 7193.00 frames. ], tot_loss[loss=0.2358, simple_loss=0.2895, pruned_loss=0.09103, over 1440098.81 frames. ], batch size: 39, lr: 2.75e-02, grad_scale: 32.0 +2023-03-20 19:10:21,521 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9892, 3.6115, 3.1623, 3.8543, 2.8016, 2.4651, 3.8629, 3.1258], + device='cuda:0'), covar=tensor([0.0058, 0.0040, 0.0175, 0.0026, 0.0222, 0.0448, 0.0073, 0.0388], + device='cuda:0'), in_proj_covar=tensor([0.0148, 0.0131, 0.0187, 0.0119, 0.0226, 0.0260, 0.0158, 0.0261], + device='cuda:0'), out_proj_covar=tensor([1.2904e-04, 1.1708e-04, 1.5422e-04, 9.8292e-05, 1.9239e-04, 2.1666e-04, + 1.4178e-04, 2.2027e-04], device='cuda:0') +2023-03-20 19:10:23,351 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5559, 2.0128, 1.7443, 2.5690, 1.2406, 2.7309, 1.0353, 2.6354], + device='cuda:0'), covar=tensor([0.0044, 0.0575, 0.2057, 0.0058, 0.4539, 0.0052, 0.1033, 0.0067], + device='cuda:0'), in_proj_covar=tensor([0.0120, 0.0194, 0.0297, 0.0114, 0.0291, 0.0112, 0.0223, 0.0124], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0001, 0.0002, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 19:10:24,159 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=12802.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:10:25,594 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.949e+02 3.496e+02 4.325e+02 5.786e+02 9.440e+02, threshold=8.649e+02, percent-clipped=4.0 +2023-03-20 19:10:28,150 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-20 19:10:35,403 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=12824.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 19:10:46,766 INFO [train.py:901] (0/2) Epoch 5, batch 1550, loss[loss=0.2233, simple_loss=0.2839, pruned_loss=0.0813, over 7157.00 frames. ], tot_loss[loss=0.2363, simple_loss=0.2897, pruned_loss=0.09141, over 1441588.79 frames. ], batch size: 41, lr: 2.75e-02, grad_scale: 32.0 +2023-03-20 19:10:52,221 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-20 19:10:56,582 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0523, 2.0728, 2.2458, 2.9011, 2.9977, 2.9146, 2.3632, 3.1718], + device='cuda:0'), covar=tensor([0.0920, 0.0533, 0.1350, 0.0178, 0.0051, 0.0025, 0.0034, 0.0057], + device='cuda:0'), in_proj_covar=tensor([0.0243, 0.0173, 0.0223, 0.0133, 0.0090, 0.0088, 0.0087, 0.0090], + device='cuda:0'), out_proj_covar=tensor([2.1065e-04, 1.6479e-04, 2.0073e-04, 1.3082e-04, 8.6570e-05, 8.2999e-05, + 8.5433e-05, 9.0133e-05], device='cuda:0') +2023-03-20 19:11:08,760 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8219, 2.1921, 1.8009, 2.1672, 1.8956, 1.9580, 1.9727, 1.7277], + device='cuda:0'), covar=tensor([0.0342, 0.0416, 0.0622, 0.0240, 0.0427, 0.0434, 0.0491, 0.0669], + device='cuda:0'), in_proj_covar=tensor([0.0033, 0.0033, 0.0032, 0.0032, 0.0035, 0.0036, 0.0031, 0.0033], + device='cuda:0'), out_proj_covar=tensor([7.5751e-05, 6.7725e-05, 6.9430e-05, 6.6596e-05, 7.4618e-05, 7.6116e-05, + 7.0180e-05, 6.9307e-05], device='cuda:0') +2023-03-20 19:11:10,299 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6209, 1.9988, 1.8396, 2.4630, 1.4424, 2.8340, 1.0705, 2.4706], + device='cuda:0'), covar=tensor([0.0056, 0.0620, 0.1911, 0.0045, 0.4110, 0.0066, 0.0965, 0.0057], + device='cuda:0'), in_proj_covar=tensor([0.0121, 0.0200, 0.0293, 0.0115, 0.0293, 0.0115, 0.0226, 0.0126], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0001, 0.0003, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 19:11:13,171 INFO [train.py:901] (0/2) Epoch 5, batch 1600, loss[loss=0.243, simple_loss=0.2973, pruned_loss=0.09434, over 7239.00 frames. ], tot_loss[loss=0.2354, simple_loss=0.2891, pruned_loss=0.09083, over 1443824.81 frames. ], batch size: 89, lr: 2.74e-02, grad_scale: 32.0 +2023-03-20 19:11:17,086 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.145e+02 3.367e+02 4.098e+02 5.521e+02 1.235e+03, threshold=8.196e+02, percent-clipped=2.0 +2023-03-20 19:11:22,253 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-20 19:11:23,235 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-20 19:11:24,623 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.95 vs. limit=2.0 +2023-03-20 19:11:26,787 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-20 19:11:30,829 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5525, 3.7420, 3.9540, 4.0664, 4.0442, 4.1635, 4.3400, 3.9167], + device='cuda:0'), covar=tensor([0.0072, 0.0143, 0.0138, 0.0111, 0.0160, 0.0109, 0.0112, 0.0112], + device='cuda:0'), in_proj_covar=tensor([0.0033, 0.0037, 0.0041, 0.0034, 0.0049, 0.0045, 0.0039, 0.0038], + device='cuda:0'), out_proj_covar=tensor([7.7092e-05, 9.2745e-05, 9.7538e-05, 8.2965e-05, 1.2431e-04, 1.1278e-04, + 1.0121e-04, 8.9434e-05], device='cuda:0') +2023-03-20 19:11:36,605 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-20 19:11:38,695 INFO [train.py:901] (0/2) Epoch 5, batch 1650, loss[loss=0.2087, simple_loss=0.2546, pruned_loss=0.08138, over 6963.00 frames. ], tot_loss[loss=0.2353, simple_loss=0.2888, pruned_loss=0.09089, over 1439348.79 frames. ], batch size: 35, lr: 2.74e-02, grad_scale: 32.0 +2023-03-20 19:11:41,286 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-20 19:11:47,084 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8811, 2.0466, 1.9648, 2.6964, 1.2159, 2.5695, 1.0983, 2.7391], + device='cuda:0'), covar=tensor([0.0080, 0.0643, 0.1800, 0.0040, 0.5402, 0.0064, 0.1103, 0.0056], + device='cuda:0'), in_proj_covar=tensor([0.0120, 0.0199, 0.0290, 0.0115, 0.0293, 0.0115, 0.0221, 0.0126], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0001, 0.0003, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 19:11:49,451 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-20 19:11:49,535 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9458, 3.6132, 3.7453, 3.6641, 3.6307, 3.7804, 3.9395, 3.5019], + device='cuda:0'), covar=tensor([0.0170, 0.0173, 0.0147, 0.0160, 0.0243, 0.0132, 0.0194, 0.0178], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0039, 0.0042, 0.0035, 0.0052, 0.0047, 0.0040, 0.0040], + device='cuda:0'), out_proj_covar=tensor([8.0360e-05, 9.7273e-05, 9.8554e-05, 8.4908e-05, 1.3085e-04, 1.1753e-04, + 1.0632e-04, 9.3576e-05], device='cuda:0') +2023-03-20 19:11:56,590 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1204, 3.5984, 3.6687, 3.7217, 3.6476, 3.8770, 3.9099, 3.4405], + device='cuda:0'), covar=tensor([0.0135, 0.0158, 0.0162, 0.0165, 0.0256, 0.0127, 0.0219, 0.0197], + device='cuda:0'), in_proj_covar=tensor([0.0034, 0.0039, 0.0042, 0.0034, 0.0052, 0.0046, 0.0040, 0.0039], + device='cuda:0'), out_proj_covar=tensor([7.9637e-05, 9.6390e-05, 9.7857e-05, 8.4649e-05, 1.3036e-04, 1.1578e-04, + 1.0527e-04, 9.2334e-05], device='cuda:0') +2023-03-20 19:12:02,479 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9414, 4.2051, 3.9454, 4.1089, 3.8474, 4.3069, 4.5717, 4.5924], + device='cuda:0'), covar=tensor([0.0264, 0.0161, 0.0218, 0.0172, 0.0323, 0.0192, 0.0199, 0.0132], + device='cuda:0'), in_proj_covar=tensor([0.0084, 0.0089, 0.0082, 0.0095, 0.0091, 0.0068, 0.0069, 0.0071], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 19:12:04,412 INFO [train.py:901] (0/2) Epoch 5, batch 1700, loss[loss=0.2457, simple_loss=0.2941, pruned_loss=0.0987, over 7301.00 frames. ], tot_loss[loss=0.2336, simple_loss=0.2879, pruned_loss=0.08966, over 1442017.86 frames. ], batch size: 49, lr: 2.73e-02, grad_scale: 32.0 +2023-03-20 19:12:07,263 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 19:12:08,031 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.89 vs. limit=2.0 +2023-03-20 19:12:08,704 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.127e+02 3.688e+02 4.669e+02 5.560e+02 1.145e+03, threshold=9.338e+02, percent-clipped=6.0 +2023-03-20 19:12:10,765 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-20 19:12:20,792 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-20 19:12:31,117 INFO [train.py:901] (0/2) Epoch 5, batch 1750, loss[loss=0.2819, simple_loss=0.3291, pruned_loss=0.1174, over 6828.00 frames. ], tot_loss[loss=0.2346, simple_loss=0.289, pruned_loss=0.09013, over 1441470.40 frames. ], batch size: 106, lr: 2.73e-02, grad_scale: 32.0 +2023-03-20 19:12:32,236 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9020, 3.9328, 3.6771, 3.9499, 3.5158, 3.9306, 4.1188, 4.2725], + device='cuda:0'), covar=tensor([0.0232, 0.0176, 0.0285, 0.0183, 0.0468, 0.0232, 0.0262, 0.0182], + device='cuda:0'), in_proj_covar=tensor([0.0078, 0.0084, 0.0078, 0.0089, 0.0087, 0.0064, 0.0066, 0.0069], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 19:12:45,693 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-20 19:12:46,680 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-20 19:12:56,139 INFO [train.py:901] (0/2) Epoch 5, batch 1800, loss[loss=0.2291, simple_loss=0.2869, pruned_loss=0.08565, over 7319.00 frames. ], tot_loss[loss=0.2338, simple_loss=0.2883, pruned_loss=0.08961, over 1444074.91 frames. ], batch size: 80, lr: 2.72e-02, grad_scale: 32.0 +2023-03-20 19:12:58,731 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=13102.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:13:00,111 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.985e+02 3.231e+02 4.030e+02 4.960e+02 9.791e+02, threshold=8.060e+02, percent-clipped=2.0 +2023-03-20 19:13:07,148 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=13119.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 19:13:07,566 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-20 19:13:08,633 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.04 vs. limit=2.0 +2023-03-20 19:13:20,969 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.81 vs. limit=5.0 +2023-03-20 19:13:21,145 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-20 19:13:21,607 INFO [train.py:901] (0/2) Epoch 5, batch 1850, loss[loss=0.2443, simple_loss=0.2911, pruned_loss=0.09874, over 7316.00 frames. ], tot_loss[loss=0.2339, simple_loss=0.2889, pruned_loss=0.0894, over 1445397.90 frames. ], batch size: 49, lr: 2.72e-02, grad_scale: 32.0 +2023-03-20 19:13:22,740 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8780, 1.9148, 1.7609, 2.2448, 1.9315, 1.7946, 2.2716, 1.3783], + device='cuda:0'), covar=tensor([0.0541, 0.0439, 0.0858, 0.0244, 0.0719, 0.0492, 0.0334, 0.0660], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0035, 0.0034, 0.0034, 0.0035, 0.0039, 0.0032, 0.0033], + device='cuda:0'), out_proj_covar=tensor([8.3323e-05, 7.2577e-05, 7.4658e-05, 7.3212e-05, 7.7025e-05, 8.3373e-05, + 7.2561e-05, 6.9849e-05], device='cuda:0') +2023-03-20 19:13:23,150 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=13150.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:13:31,162 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-20 19:13:46,570 INFO [train.py:901] (0/2) Epoch 5, batch 1900, loss[loss=0.2315, simple_loss=0.2823, pruned_loss=0.0903, over 7246.00 frames. ], tot_loss[loss=0.2332, simple_loss=0.2881, pruned_loss=0.08919, over 1442942.38 frames. ], batch size: 64, lr: 2.71e-02, grad_scale: 32.0 +2023-03-20 19:13:48,478 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-20 19:13:50,871 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.920e+02 3.572e+02 4.381e+02 5.635e+02 1.189e+03, threshold=8.761e+02, percent-clipped=5.0 +2023-03-20 19:14:01,385 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6733, 3.6659, 3.5558, 3.5704, 3.2622, 3.6336, 3.8318, 3.9441], + device='cuda:0'), covar=tensor([0.0231, 0.0191, 0.0267, 0.0242, 0.0444, 0.0379, 0.0350, 0.0246], + device='cuda:0'), in_proj_covar=tensor([0.0083, 0.0088, 0.0084, 0.0096, 0.0092, 0.0070, 0.0070, 0.0072], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 19:14:13,361 INFO [train.py:901] (0/2) Epoch 5, batch 1950, loss[loss=0.2229, simple_loss=0.2871, pruned_loss=0.07934, over 7239.00 frames. ], tot_loss[loss=0.2335, simple_loss=0.2885, pruned_loss=0.0893, over 1443162.87 frames. ], batch size: 89, lr: 2.71e-02, grad_scale: 32.0 +2023-03-20 19:14:13,865 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-20 19:14:24,433 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-20 19:14:29,423 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-20 19:14:29,934 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-20 19:14:38,267 INFO [train.py:901] (0/2) Epoch 5, batch 2000, loss[loss=0.2427, simple_loss=0.2936, pruned_loss=0.09587, over 7320.00 frames. ], tot_loss[loss=0.2341, simple_loss=0.2889, pruned_loss=0.08961, over 1443977.83 frames. ], batch size: 61, lr: 2.71e-02, grad_scale: 32.0 +2023-03-20 19:14:43,470 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.925e+02 3.298e+02 4.123e+02 5.170e+02 8.355e+02, threshold=8.246e+02, percent-clipped=0.0 +2023-03-20 19:14:44,127 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1978, 4.0295, 3.6947, 3.5584, 3.5172, 2.6483, 2.0780, 4.1852], + device='cuda:0'), covar=tensor([0.0014, 0.0042, 0.0073, 0.0079, 0.0030, 0.0424, 0.0844, 0.0039], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0045, 0.0063, 0.0048, 0.0049, 0.0075, 0.0091, 0.0051], + device='cuda:0'), out_proj_covar=tensor([5.3197e-05, 7.0109e-05, 9.0020e-05, 6.9394e-05, 6.3864e-05, 1.1272e-04, + 1.3600e-04, 7.1653e-05], device='cuda:0') +2023-03-20 19:14:44,817 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.90 vs. limit=2.0 +2023-03-20 19:14:47,041 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-20 19:14:57,653 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-20 19:15:04,690 INFO [train.py:901] (0/2) Epoch 5, batch 2050, loss[loss=0.2302, simple_loss=0.2826, pruned_loss=0.08889, over 7330.00 frames. ], tot_loss[loss=0.2344, simple_loss=0.2896, pruned_loss=0.08958, over 1445468.24 frames. ], batch size: 51, lr: 2.70e-02, grad_scale: 32.0 +2023-03-20 19:15:05,228 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-20 19:15:25,742 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-03-20 19:15:29,574 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5385, 3.2324, 2.5910, 3.4665, 2.4471, 1.9742, 3.3655, 2.5759], + device='cuda:0'), covar=tensor([0.0086, 0.0084, 0.0247, 0.0031, 0.0242, 0.0433, 0.0093, 0.0440], + device='cuda:0'), in_proj_covar=tensor([0.0158, 0.0140, 0.0194, 0.0119, 0.0238, 0.0259, 0.0170, 0.0267], + device='cuda:0'), out_proj_covar=tensor([1.3958e-04, 1.2775e-04, 1.6245e-04, 9.9333e-05, 2.0221e-04, 2.1785e-04, + 1.5141e-04, 2.2755e-04], device='cuda:0') +2023-03-20 19:15:30,843 INFO [train.py:901] (0/2) Epoch 5, batch 2100, loss[loss=0.2341, simple_loss=0.2956, pruned_loss=0.08633, over 7282.00 frames. ], tot_loss[loss=0.2344, simple_loss=0.2897, pruned_loss=0.08957, over 1445171.07 frames. ], batch size: 57, lr: 2.70e-02, grad_scale: 32.0 +2023-03-20 19:15:35,084 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.118e+02 3.365e+02 4.362e+02 5.459e+02 9.984e+02, threshold=8.723e+02, percent-clipped=1.0 +2023-03-20 19:15:39,096 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-20 19:15:42,112 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-20 19:15:42,200 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=13419.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:15:56,129 INFO [train.py:901] (0/2) Epoch 5, batch 2150, loss[loss=0.2541, simple_loss=0.2994, pruned_loss=0.1044, over 7310.00 frames. ], tot_loss[loss=0.2336, simple_loss=0.289, pruned_loss=0.08908, over 1444532.55 frames. ], batch size: 83, lr: 2.69e-02, grad_scale: 32.0 +2023-03-20 19:16:06,801 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=13467.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:16:07,571 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.84 vs. limit=5.0 +2023-03-20 19:16:08,512 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-03-20 19:16:22,393 INFO [train.py:901] (0/2) Epoch 5, batch 2200, loss[loss=0.2233, simple_loss=0.2818, pruned_loss=0.08246, over 7278.00 frames. ], tot_loss[loss=0.2351, simple_loss=0.29, pruned_loss=0.09012, over 1443098.69 frames. ], batch size: 77, lr: 2.69e-02, grad_scale: 32.0 +2023-03-20 19:16:26,325 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.08 vs. limit=2.0 +2023-03-20 19:16:26,406 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.479e+02 3.587e+02 4.467e+02 5.823e+02 1.129e+03, threshold=8.934e+02, percent-clipped=3.0 +2023-03-20 19:16:26,447 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-20 19:16:40,863 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1537, 4.5248, 4.5425, 4.4781, 4.4238, 4.1884, 4.5541, 4.5255], + device='cuda:0'), covar=tensor([0.0403, 0.0370, 0.0568, 0.0498, 0.0327, 0.0292, 0.0383, 0.0401], + device='cuda:0'), in_proj_covar=tensor([0.0110, 0.0098, 0.0100, 0.0086, 0.0080, 0.0102, 0.0093, 0.0078], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 19:16:47,777 INFO [train.py:901] (0/2) Epoch 5, batch 2250, loss[loss=0.234, simple_loss=0.2913, pruned_loss=0.08838, over 7290.00 frames. ], tot_loss[loss=0.2346, simple_loss=0.2894, pruned_loss=0.08989, over 1442355.43 frames. ], batch size: 57, lr: 2.68e-02, grad_scale: 32.0 +2023-03-20 19:17:01,113 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-20 19:17:01,619 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-20 19:17:13,138 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-20 19:17:14,172 INFO [train.py:901] (0/2) Epoch 5, batch 2300, loss[loss=0.2147, simple_loss=0.2773, pruned_loss=0.076, over 7273.00 frames. ], tot_loss[loss=0.2342, simple_loss=0.2893, pruned_loss=0.08956, over 1444389.66 frames. ], batch size: 70, lr: 2.68e-02, grad_scale: 32.0 +2023-03-20 19:17:18,478 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.119e+02 3.634e+02 4.317e+02 5.619e+02 1.145e+03, threshold=8.633e+02, percent-clipped=2.0 +2023-03-20 19:17:26,950 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.05 vs. limit=2.0 +2023-03-20 19:17:40,828 INFO [train.py:901] (0/2) Epoch 5, batch 2350, loss[loss=0.2504, simple_loss=0.305, pruned_loss=0.09789, over 7348.00 frames. ], tot_loss[loss=0.2351, simple_loss=0.2901, pruned_loss=0.09004, over 1445606.43 frames. ], batch size: 63, lr: 2.68e-02, grad_scale: 32.0 +2023-03-20 19:18:00,345 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-20 19:18:05,788 INFO [train.py:901] (0/2) Epoch 5, batch 2400, loss[loss=0.2582, simple_loss=0.3062, pruned_loss=0.1051, over 7266.00 frames. ], tot_loss[loss=0.2341, simple_loss=0.289, pruned_loss=0.08957, over 1445355.59 frames. ], batch size: 89, lr: 2.67e-02, grad_scale: 32.0 +2023-03-20 19:18:05,810 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-20 19:18:09,737 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.037e+02 3.360e+02 4.208e+02 5.531e+02 1.498e+03, threshold=8.415e+02, percent-clipped=3.0 +2023-03-20 19:18:16,877 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-20 19:18:19,414 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-20 19:18:27,314 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=13737.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:18:32,093 INFO [train.py:901] (0/2) Epoch 5, batch 2450, loss[loss=0.2198, simple_loss=0.2817, pruned_loss=0.07898, over 7316.00 frames. ], tot_loss[loss=0.2345, simple_loss=0.2893, pruned_loss=0.08986, over 1442079.91 frames. ], batch size: 59, lr: 2.67e-02, grad_scale: 32.0 +2023-03-20 19:18:39,839 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.93 vs. limit=2.0 +2023-03-20 19:18:45,588 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-20 19:18:52,842 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.33 vs. limit=5.0 +2023-03-20 19:18:57,129 INFO [train.py:901] (0/2) Epoch 5, batch 2500, loss[loss=0.2935, simple_loss=0.3377, pruned_loss=0.1247, over 6830.00 frames. ], tot_loss[loss=0.2355, simple_loss=0.29, pruned_loss=0.09047, over 1443172.83 frames. ], batch size: 107, lr: 2.66e-02, grad_scale: 32.0 +2023-03-20 19:18:57,778 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=13798.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 19:19:01,982 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.868e+02 3.607e+02 4.140e+02 5.298e+02 1.184e+03, threshold=8.279e+02, percent-clipped=3.0 +2023-03-20 19:19:12,213 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-20 19:19:18,981 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.11 vs. limit=2.0 +2023-03-20 19:19:23,799 INFO [train.py:901] (0/2) Epoch 5, batch 2550, loss[loss=0.2338, simple_loss=0.2952, pruned_loss=0.08614, over 7213.00 frames. ], tot_loss[loss=0.2354, simple_loss=0.2903, pruned_loss=0.09023, over 1445284.92 frames. ], batch size: 93, lr: 2.66e-02, grad_scale: 32.0 +2023-03-20 19:19:48,490 INFO [train.py:901] (0/2) Epoch 5, batch 2600, loss[loss=0.1804, simple_loss=0.238, pruned_loss=0.06137, over 7147.00 frames. ], tot_loss[loss=0.235, simple_loss=0.2899, pruned_loss=0.09003, over 1446143.07 frames. ], batch size: 39, lr: 2.65e-02, grad_scale: 16.0 +2023-03-20 19:19:52,791 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.171e+02 3.332e+02 4.151e+02 5.370e+02 8.720e+02, threshold=8.301e+02, percent-clipped=2.0 +2023-03-20 19:20:13,531 INFO [train.py:901] (0/2) Epoch 5, batch 2650, loss[loss=0.2069, simple_loss=0.2656, pruned_loss=0.07408, over 7166.00 frames. ], tot_loss[loss=0.2341, simple_loss=0.2887, pruned_loss=0.08975, over 1444400.43 frames. ], batch size: 39, lr: 2.65e-02, grad_scale: 16.0 +2023-03-20 19:20:28,044 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7724, 3.8848, 3.5577, 3.1228, 3.3553, 2.0990, 2.2000, 3.9853], + device='cuda:0'), covar=tensor([0.0023, 0.0024, 0.0061, 0.0085, 0.0029, 0.0429, 0.0683, 0.0027], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0045, 0.0064, 0.0050, 0.0050, 0.0077, 0.0090, 0.0050], + device='cuda:0'), out_proj_covar=tensor([5.8695e-05, 6.9259e-05, 9.4088e-05, 7.4419e-05, 6.5468e-05, 1.1501e-04, + 1.3458e-04, 6.9381e-05], device='cuda:0') +2023-03-20 19:20:35,477 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.02 vs. limit=2.0 +2023-03-20 19:20:38,625 INFO [train.py:901] (0/2) Epoch 5, batch 2700, loss[loss=0.2339, simple_loss=0.2905, pruned_loss=0.0886, over 7353.00 frames. ], tot_loss[loss=0.2342, simple_loss=0.2892, pruned_loss=0.08957, over 1444930.23 frames. ], batch size: 63, lr: 2.65e-02, grad_scale: 16.0 +2023-03-20 19:20:43,427 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.392e+02 3.598e+02 4.306e+02 4.916e+02 1.230e+03, threshold=8.611e+02, percent-clipped=5.0 +2023-03-20 19:20:52,943 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2602, 1.3446, 1.2998, 1.1625, 0.9482, 1.3188, 1.0083, 1.1875], + device='cuda:0'), covar=tensor([0.0404, 0.0273, 0.0217, 0.0173, 0.0415, 0.0417, 0.0273, 0.0446], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0020, 0.0018, 0.0018, 0.0020, 0.0018, 0.0020, 0.0020], + device='cuda:0'), out_proj_covar=tensor([4.2751e-05, 4.1101e-05, 3.5663e-05, 3.6005e-05, 4.2463e-05, 3.9605e-05, + 4.1111e-05, 4.2989e-05], device='cuda:0') +2023-03-20 19:21:03,779 INFO [train.py:901] (0/2) Epoch 5, batch 2750, loss[loss=0.2374, simple_loss=0.2916, pruned_loss=0.0916, over 7271.00 frames. ], tot_loss[loss=0.2322, simple_loss=0.2875, pruned_loss=0.08841, over 1444440.45 frames. ], batch size: 57, lr: 2.64e-02, grad_scale: 16.0 +2023-03-20 19:21:16,269 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=14072.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:21:26,579 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=14093.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 19:21:28,450 INFO [train.py:901] (0/2) Epoch 5, batch 2800, loss[loss=0.2829, simple_loss=0.3239, pruned_loss=0.121, over 6652.00 frames. ], tot_loss[loss=0.2329, simple_loss=0.288, pruned_loss=0.08884, over 1443060.64 frames. ], batch size: 106, lr: 2.64e-02, grad_scale: 16.0 +2023-03-20 19:21:31,476 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2505, 4.7540, 4.8160, 4.6422, 4.5511, 4.2914, 4.8273, 4.6562], + device='cuda:0'), covar=tensor([0.0420, 0.0326, 0.0417, 0.0552, 0.0460, 0.0354, 0.0287, 0.0507], + device='cuda:0'), in_proj_covar=tensor([0.0110, 0.0100, 0.0097, 0.0087, 0.0083, 0.0107, 0.0092, 0.0078], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 19:21:31,529 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=14103.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:21:32,842 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.456e+02 3.620e+02 4.147e+02 5.352e+02 9.706e+02, threshold=8.293e+02, percent-clipped=5.0 +2023-03-20 19:21:40,908 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-5.pt +2023-03-20 19:21:58,153 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-20 19:21:58,207 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-20 19:21:58,232 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-20 19:21:58,251 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-20 19:21:58,762 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-20 19:21:58,778 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-20 19:21:58,789 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-20 19:21:58,888 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-20 19:21:58,962 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-20 19:21:58,978 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-20 19:21:58,994 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-20 19:21:59,157 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-20 19:21:59,163 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-20 19:21:59,170 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-20 19:21:59,309 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-20 19:21:59,312 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-20 19:21:59,371 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-20 19:22:01,133 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=14120.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 19:22:01,491 INFO [train.py:901] (0/2) Epoch 6, batch 0, loss[loss=0.2017, simple_loss=0.2658, pruned_loss=0.06878, over 7330.00 frames. ], tot_loss[loss=0.2017, simple_loss=0.2658, pruned_loss=0.06878, over 7330.00 frames. ], batch size: 54, lr: 2.53e-02, grad_scale: 16.0 +2023-03-20 19:22:01,492 INFO [train.py:926] (0/2) Computing validation loss +2023-03-20 19:22:17,872 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.9963, 1.1450, 0.7928, 0.8484, 1.0351, 0.8571, 0.8989, 0.9141], + device='cuda:0'), covar=tensor([0.0137, 0.0215, 0.0234, 0.0145, 0.0192, 0.0146, 0.0242, 0.0126], + device='cuda:0'), in_proj_covar=tensor([0.0017, 0.0019, 0.0016, 0.0016, 0.0017, 0.0017, 0.0016, 0.0017], + device='cuda:0'), out_proj_covar=tensor([2.0441e-05, 2.1802e-05, 2.2012e-05, 1.9686e-05, 1.9756e-05, 2.2285e-05, + 2.1194e-05, 2.4836e-05], device='cuda:0') +2023-03-20 19:22:21,110 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6795, 3.7677, 3.5518, 3.2682, 3.1388, 2.2212, 1.5496, 3.9273], + device='cuda:0'), covar=tensor([0.0039, 0.0041, 0.0066, 0.0090, 0.0044, 0.0637, 0.1077, 0.0063], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0047, 0.0066, 0.0052, 0.0054, 0.0081, 0.0096, 0.0053], + device='cuda:0'), out_proj_covar=tensor([6.2659e-05, 7.3514e-05, 9.5292e-05, 7.7542e-05, 7.1343e-05, 1.2097e-04, + 1.4199e-04, 7.3696e-05], device='cuda:0') +2023-03-20 19:22:27,063 INFO [train.py:935] (0/2) Epoch 6, validation: loss=0.1912, simple_loss=0.2765, pruned_loss=0.05298, over 1622729.00 frames. +2023-03-20 19:22:27,063 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12054MB +2023-03-20 19:22:32,876 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.00 vs. limit=2.0 +2023-03-20 19:22:33,235 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=14133.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:22:34,140 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-20 19:22:44,808 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-20 19:22:50,048 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=14164.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:22:51,894 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-20 19:22:53,346 INFO [train.py:901] (0/2) Epoch 6, batch 50, loss[loss=0.2363, simple_loss=0.2966, pruned_loss=0.08799, over 7331.00 frames. ], tot_loss[loss=0.2306, simple_loss=0.2863, pruned_loss=0.08742, over 326141.05 frames. ], batch size: 59, lr: 2.53e-02, grad_scale: 16.0 +2023-03-20 19:22:53,855 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-20 19:22:56,427 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-20 19:22:58,608 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=14181.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 19:23:11,286 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.057e+02 3.549e+02 4.505e+02 5.782e+02 1.054e+03, threshold=9.011e+02, percent-clipped=3.0 +2023-03-20 19:23:13,313 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-20 19:23:13,835 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-20 19:23:13,978 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2195, 2.2065, 2.1686, 3.2435, 2.7460, 2.8952, 2.5714, 2.5599], + device='cuda:0'), covar=tensor([0.0697, 0.0382, 0.1012, 0.0164, 0.0033, 0.0026, 0.0021, 0.0046], + device='cuda:0'), in_proj_covar=tensor([0.0267, 0.0194, 0.0245, 0.0158, 0.0094, 0.0101, 0.0096, 0.0105], + device='cuda:0'), out_proj_covar=tensor([2.3382e-04, 1.8780e-04, 2.2301e-04, 1.5584e-04, 9.2609e-05, 9.7062e-05, + 9.4612e-05, 1.0452e-04], device='cuda:0') +2023-03-20 19:23:14,580 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.96 vs. limit=2.0 +2023-03-20 19:23:18,732 INFO [train.py:901] (0/2) Epoch 6, batch 100, loss[loss=0.1644, simple_loss=0.2341, pruned_loss=0.04737, over 7159.00 frames. ], tot_loss[loss=0.2307, simple_loss=0.2862, pruned_loss=0.08758, over 571849.14 frames. ], batch size: 39, lr: 2.52e-02, grad_scale: 16.0 +2023-03-20 19:23:31,866 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-03-20 19:23:33,733 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.81 vs. limit=5.0 +2023-03-20 19:23:45,024 INFO [train.py:901] (0/2) Epoch 6, batch 150, loss[loss=0.2536, simple_loss=0.2997, pruned_loss=0.1038, over 7299.00 frames. ], tot_loss[loss=0.2292, simple_loss=0.2853, pruned_loss=0.08657, over 765348.51 frames. ], batch size: 68, lr: 2.52e-02, grad_scale: 16.0 +2023-03-20 19:24:02,566 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.799e+02 3.314e+02 3.960e+02 4.953e+02 9.074e+02, threshold=7.919e+02, percent-clipped=1.0 +2023-03-20 19:24:06,854 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.17 vs. limit=2.0 +2023-03-20 19:24:07,947 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.88 vs. limit=5.0 +2023-03-20 19:24:10,052 INFO [train.py:901] (0/2) Epoch 6, batch 200, loss[loss=0.2603, simple_loss=0.3052, pruned_loss=0.1077, over 7252.00 frames. ], tot_loss[loss=0.2274, simple_loss=0.2833, pruned_loss=0.08572, over 914316.47 frames. ], batch size: 89, lr: 2.52e-02, grad_scale: 16.0 +2023-03-20 19:24:16,767 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-20 19:24:21,808 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-20 19:24:27,248 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-20 19:24:36,319 INFO [train.py:901] (0/2) Epoch 6, batch 250, loss[loss=0.2299, simple_loss=0.2902, pruned_loss=0.08477, over 7246.00 frames. ], tot_loss[loss=0.2274, simple_loss=0.2836, pruned_loss=0.08562, over 1031513.50 frames. ], batch size: 55, lr: 2.51e-02, grad_scale: 16.0 +2023-03-20 19:24:39,912 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4431, 4.1621, 4.0074, 3.4591, 4.0295, 2.4532, 1.7668, 4.2124], + device='cuda:0'), covar=tensor([0.0014, 0.0032, 0.0071, 0.0087, 0.0018, 0.0394, 0.0815, 0.0046], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0045, 0.0066, 0.0053, 0.0052, 0.0079, 0.0096, 0.0053], + device='cuda:0'), out_proj_covar=tensor([5.9814e-05, 7.1011e-05, 9.6147e-05, 7.9041e-05, 6.9142e-05, 1.1836e-04, + 1.4140e-04, 7.4839e-05], device='cuda:0') +2023-03-20 19:24:40,308 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-20 19:24:42,889 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1665, 4.0916, 4.1400, 4.5504, 4.6154, 4.5230, 3.9573, 3.9210], + device='cuda:0'), covar=tensor([0.0641, 0.1732, 0.1778, 0.0750, 0.0466, 0.1137, 0.0643, 0.0860], + device='cuda:0'), in_proj_covar=tensor([0.0118, 0.0194, 0.0182, 0.0159, 0.0131, 0.0200, 0.0110, 0.0135], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 19:24:47,346 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=14393.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:24:54,059 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.668e+02 3.223e+02 4.250e+02 5.508e+02 1.036e+03, threshold=8.501e+02, percent-clipped=7.0 +2023-03-20 19:24:54,852 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.04 vs. limit=2.0 +2023-03-20 19:24:58,271 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=14413.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:25:01,767 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-20 19:25:02,711 INFO [train.py:901] (0/2) Epoch 6, batch 300, loss[loss=0.2404, simple_loss=0.2907, pruned_loss=0.09503, over 7218.00 frames. ], tot_loss[loss=0.2267, simple_loss=0.2833, pruned_loss=0.0851, over 1122879.91 frames. ], batch size: 45, lr: 2.51e-02, grad_scale: 16.0 +2023-03-20 19:25:05,084 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.11 vs. limit=2.0 +2023-03-20 19:25:06,315 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=14428.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:25:10,247 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-20 19:25:12,832 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=14441.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:25:13,432 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0368, 3.7719, 3.2895, 3.9614, 2.8456, 2.5993, 3.9479, 3.2669], + device='cuda:0'), covar=tensor([0.0074, 0.0043, 0.0119, 0.0022, 0.0156, 0.0267, 0.0082, 0.0250], + device='cuda:0'), in_proj_covar=tensor([0.0177, 0.0159, 0.0204, 0.0137, 0.0258, 0.0273, 0.0193, 0.0281], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0001, 0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 19:25:21,764 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=14459.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:25:27,671 INFO [train.py:901] (0/2) Epoch 6, batch 350, loss[loss=0.2362, simple_loss=0.2909, pruned_loss=0.09071, over 7281.00 frames. ], tot_loss[loss=0.2268, simple_loss=0.2837, pruned_loss=0.08492, over 1195661.49 frames. ], batch size: 66, lr: 2.50e-02, grad_scale: 16.0 +2023-03-20 19:25:29,280 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=14474.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:25:30,184 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=14476.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 19:25:35,719 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=14487.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:25:36,188 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0065, 4.1260, 3.9293, 4.1253, 3.7598, 4.3394, 4.4701, 4.5480], + device='cuda:0'), covar=tensor([0.0215, 0.0139, 0.0211, 0.0185, 0.0385, 0.0153, 0.0226, 0.0160], + device='cuda:0'), in_proj_covar=tensor([0.0085, 0.0085, 0.0083, 0.0093, 0.0087, 0.0068, 0.0071, 0.0070], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 19:25:36,723 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.8582, 1.0081, 1.0916, 1.2964, 1.3115, 0.8934, 0.6141, 1.1495], + device='cuda:0'), covar=tensor([0.0142, 0.0157, 0.0175, 0.0130, 0.0160, 0.0177, 0.0324, 0.0160], + device='cuda:0'), in_proj_covar=tensor([0.0017, 0.0018, 0.0016, 0.0017, 0.0016, 0.0018, 0.0018, 0.0019], + device='cuda:0'), out_proj_covar=tensor([2.1835e-05, 2.1224e-05, 2.2749e-05, 2.0853e-05, 1.8981e-05, 2.2617e-05, + 2.3303e-05, 2.6786e-05], device='cuda:0') +2023-03-20 19:25:43,316 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8126, 3.6148, 3.1372, 3.1274, 3.4637, 2.0314, 1.5268, 3.7341], + device='cuda:0'), covar=tensor([0.0017, 0.0024, 0.0138, 0.0066, 0.0041, 0.0429, 0.0744, 0.0046], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0045, 0.0066, 0.0052, 0.0053, 0.0078, 0.0094, 0.0051], + device='cuda:0'), out_proj_covar=tensor([5.9655e-05, 7.0571e-05, 9.7393e-05, 7.8141e-05, 7.0289e-05, 1.1677e-04, + 1.3952e-04, 7.2499e-05], device='cuda:0') +2023-03-20 19:25:43,734 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-20 19:25:46,321 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.128e+02 3.102e+02 3.888e+02 4.761e+02 1.038e+03, threshold=7.777e+02, percent-clipped=1.0 +2023-03-20 19:25:53,683 INFO [train.py:901] (0/2) Epoch 6, batch 400, loss[loss=0.2578, simple_loss=0.3066, pruned_loss=0.1045, over 7308.00 frames. ], tot_loss[loss=0.2278, simple_loss=0.2844, pruned_loss=0.08555, over 1251191.90 frames. ], batch size: 80, lr: 2.50e-02, grad_scale: 16.0 +2023-03-20 19:26:07,231 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=14548.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:26:18,440 INFO [train.py:901] (0/2) Epoch 6, batch 450, loss[loss=0.2289, simple_loss=0.2921, pruned_loss=0.08281, over 7347.00 frames. ], tot_loss[loss=0.2287, simple_loss=0.2853, pruned_loss=0.08608, over 1294103.87 frames. ], batch size: 61, lr: 2.50e-02, grad_scale: 16.0 +2023-03-20 19:26:22,526 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-20 19:26:23,041 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-20 19:26:37,410 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.357e+02 3.392e+02 4.278e+02 5.459e+02 9.889e+02, threshold=8.555e+02, percent-clipped=5.0 +2023-03-20 19:26:37,602 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9542, 2.1291, 2.0720, 3.0768, 2.8623, 3.0580, 3.1829, 2.7859], + device='cuda:0'), covar=tensor([0.0880, 0.0446, 0.1203, 0.0138, 0.0038, 0.0039, 0.0041, 0.0029], + device='cuda:0'), in_proj_covar=tensor([0.0258, 0.0187, 0.0235, 0.0152, 0.0094, 0.0099, 0.0095, 0.0097], + device='cuda:0'), out_proj_covar=tensor([2.2711e-04, 1.8121e-04, 2.1542e-04, 1.5018e-04, 9.2758e-05, 9.6299e-05, + 9.4287e-05, 9.6973e-05], device='cuda:0') +2023-03-20 19:26:45,003 INFO [train.py:901] (0/2) Epoch 6, batch 500, loss[loss=0.2377, simple_loss=0.2991, pruned_loss=0.08815, over 7327.00 frames. ], tot_loss[loss=0.2273, simple_loss=0.2841, pruned_loss=0.08522, over 1327236.27 frames. ], batch size: 83, lr: 2.49e-02, grad_scale: 16.0 +2023-03-20 19:26:45,184 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9202, 2.1347, 2.1606, 3.0053, 2.8159, 2.9244, 3.0246, 2.8854], + device='cuda:0'), covar=tensor([0.0788, 0.0346, 0.0848, 0.0137, 0.0036, 0.0030, 0.0028, 0.0031], + device='cuda:0'), in_proj_covar=tensor([0.0261, 0.0189, 0.0236, 0.0153, 0.0095, 0.0100, 0.0095, 0.0098], + device='cuda:0'), out_proj_covar=tensor([2.2993e-04, 1.8260e-04, 2.1668e-04, 1.5104e-04, 9.3698e-05, 9.7198e-05, + 9.4668e-05, 9.7593e-05], device='cuda:0') +2023-03-20 19:26:55,608 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-20 19:26:56,654 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-20 19:26:57,177 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-20 19:26:59,580 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-20 19:27:01,304 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5108, 3.7508, 2.5520, 3.8668, 3.0820, 3.7909, 2.3230, 2.1628], + device='cuda:0'), covar=tensor([0.0030, 0.0274, 0.0457, 0.0097, 0.0068, 0.0047, 0.0692, 0.0528], + device='cuda:0'), in_proj_covar=tensor([0.0104, 0.0137, 0.0248, 0.0119, 0.0127, 0.0123, 0.0242, 0.0237], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0003, 0.0001, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 19:27:04,205 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-20 19:27:09,430 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0427, 2.0790, 1.9248, 1.9192, 2.1839, 1.9439, 2.3247, 1.8980], + device='cuda:0'), covar=tensor([0.0776, 0.0413, 0.0454, 0.0375, 0.0373, 0.0449, 0.0391, 0.0585], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0036, 0.0036, 0.0036, 0.0034, 0.0037, 0.0034, 0.0034], + device='cuda:0'), out_proj_covar=tensor([9.1865e-05, 8.2600e-05, 8.3090e-05, 8.2121e-05, 8.1712e-05, 8.5086e-05, + 8.1491e-05, 7.7688e-05], device='cuda:0') +2023-03-20 19:27:11,366 INFO [train.py:901] (0/2) Epoch 6, batch 550, loss[loss=0.1943, simple_loss=0.26, pruned_loss=0.06431, over 7229.00 frames. ], tot_loss[loss=0.227, simple_loss=0.2838, pruned_loss=0.08516, over 1353325.51 frames. ], batch size: 45, lr: 2.49e-02, grad_scale: 16.0 +2023-03-20 19:27:16,884 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-20 19:27:25,493 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-20 19:27:28,538 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-20 19:27:29,015 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.843e+02 3.123e+02 4.272e+02 5.651e+02 1.008e+03, threshold=8.544e+02, percent-clipped=3.0 +2023-03-20 19:27:35,418 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3498, 2.7058, 2.2142, 1.9294, 2.6221, 2.7172, 2.0471, 2.7538], + device='cuda:0'), covar=tensor([0.1032, 0.0336, 0.2525, 0.3562, 0.0790, 0.0986, 0.2319, 0.1474], + device='cuda:0'), in_proj_covar=tensor([0.0024, 0.0026, 0.0028, 0.0028, 0.0023, 0.0027, 0.0033, 0.0028], + device='cuda:0'), out_proj_covar=tensor([6.5806e-05, 6.4444e-05, 7.3739e-05, 7.2802e-05, 6.4639e-05, 7.3413e-05, + 8.1381e-05, 7.2514e-05], device='cuda:0') +2023-03-20 19:27:36,284 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-20 19:27:36,786 INFO [train.py:901] (0/2) Epoch 6, batch 600, loss[loss=0.2251, simple_loss=0.2876, pruned_loss=0.0813, over 7299.00 frames. ], tot_loss[loss=0.2276, simple_loss=0.2843, pruned_loss=0.08546, over 1374500.18 frames. ], batch size: 86, lr: 2.49e-02, grad_scale: 16.0 +2023-03-20 19:27:40,418 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=14728.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:27:46,584 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=14740.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:27:52,425 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-20 19:27:56,615 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=14759.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:28:02,296 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=14769.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:28:03,284 INFO [train.py:901] (0/2) Epoch 6, batch 650, loss[loss=0.2182, simple_loss=0.2721, pruned_loss=0.08209, over 7233.00 frames. ], tot_loss[loss=0.2262, simple_loss=0.283, pruned_loss=0.08465, over 1391234.37 frames. ], batch size: 45, lr: 2.48e-02, grad_scale: 16.0 +2023-03-20 19:28:03,310 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-20 19:28:05,467 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7963, 2.3965, 2.1317, 2.3116, 1.5038, 2.6297, 1.5676, 2.9865], + device='cuda:0'), covar=tensor([0.0084, 0.0548, 0.1768, 0.0025, 0.4578, 0.0042, 0.0897, 0.0057], + device='cuda:0'), in_proj_covar=tensor([0.0131, 0.0218, 0.0309, 0.0115, 0.0311, 0.0127, 0.0245, 0.0136], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0003, 0.0001, 0.0003, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 19:28:05,854 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=14776.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:28:05,895 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=14776.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 19:28:18,889 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=14801.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:28:21,203 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.066e+02 3.284e+02 4.078e+02 5.127e+02 1.452e+03, threshold=8.156e+02, percent-clipped=2.0 +2023-03-20 19:28:21,253 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-20 19:28:21,786 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=14807.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:28:28,648 INFO [train.py:901] (0/2) Epoch 6, batch 700, loss[loss=0.2297, simple_loss=0.2819, pruned_loss=0.08871, over 7354.00 frames. ], tot_loss[loss=0.2256, simple_loss=0.2823, pruned_loss=0.08443, over 1400633.24 frames. ], batch size: 73, lr: 2.48e-02, grad_scale: 16.0 +2023-03-20 19:28:29,160 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-20 19:28:30,166 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=14824.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 19:28:34,324 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.41 vs. limit=2.0 +2023-03-20 19:28:35,652 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0671, 2.2706, 2.2951, 2.5075, 1.6740, 2.9461, 1.4073, 3.2002], + device='cuda:0'), covar=tensor([0.0063, 0.0664, 0.1477, 0.0026, 0.4163, 0.0087, 0.0985, 0.0074], + device='cuda:0'), in_proj_covar=tensor([0.0131, 0.0220, 0.0308, 0.0115, 0.0311, 0.0129, 0.0248, 0.0137], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0003, 0.0001, 0.0003, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 19:28:38,937 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.97 vs. limit=2.0 +2023-03-20 19:28:39,582 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=14843.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:28:42,760 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1430, 4.6617, 4.6677, 4.5625, 4.4766, 4.0821, 4.6555, 4.5167], + device='cuda:0'), covar=tensor([0.0439, 0.0328, 0.0408, 0.0459, 0.0412, 0.0349, 0.0294, 0.0514], + device='cuda:0'), in_proj_covar=tensor([0.0117, 0.0112, 0.0104, 0.0093, 0.0085, 0.0111, 0.0098, 0.0086], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 19:28:53,213 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-20 19:28:53,720 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-20 19:28:54,714 INFO [train.py:901] (0/2) Epoch 6, batch 750, loss[loss=0.2453, simple_loss=0.2992, pruned_loss=0.09572, over 7286.00 frames. ], tot_loss[loss=0.2256, simple_loss=0.2824, pruned_loss=0.08441, over 1408896.44 frames. ], batch size: 57, lr: 2.47e-02, grad_scale: 16.0 +2023-03-20 19:29:07,845 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-20 19:29:11,926 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-20 19:29:12,404 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.820e+02 3.169e+02 4.239e+02 5.428e+02 1.155e+03, threshold=8.477e+02, percent-clipped=3.0 +2023-03-20 19:29:17,410 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-20 19:29:18,894 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-20 19:29:19,952 INFO [train.py:901] (0/2) Epoch 6, batch 800, loss[loss=0.169, simple_loss=0.2109, pruned_loss=0.06355, over 5963.00 frames. ], tot_loss[loss=0.2255, simple_loss=0.2824, pruned_loss=0.08426, over 1414850.05 frames. ], batch size: 25, lr: 2.47e-02, grad_scale: 16.0 +2023-03-20 19:29:21,085 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1703, 4.6098, 4.6373, 4.4761, 4.4447, 4.0535, 4.5756, 4.5132], + device='cuda:0'), covar=tensor([0.0440, 0.0353, 0.0330, 0.0460, 0.0389, 0.0362, 0.0369, 0.0424], + device='cuda:0'), in_proj_covar=tensor([0.0120, 0.0116, 0.0107, 0.0094, 0.0087, 0.0114, 0.0100, 0.0087], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 19:29:30,144 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-20 19:29:33,947 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.14 vs. limit=5.0 +2023-03-20 19:29:46,278 INFO [train.py:901] (0/2) Epoch 6, batch 850, loss[loss=0.2435, simple_loss=0.301, pruned_loss=0.09295, over 7349.00 frames. ], tot_loss[loss=0.2258, simple_loss=0.2828, pruned_loss=0.08437, over 1421849.94 frames. ], batch size: 75, lr: 2.47e-02, grad_scale: 16.0 +2023-03-20 19:29:48,299 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-20 19:29:48,309 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-20 19:29:53,989 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-20 19:29:56,665 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3025, 2.1238, 2.3994, 3.4488, 2.7003, 2.9920, 3.1643, 2.3887], + device='cuda:0'), covar=tensor([0.0747, 0.0468, 0.0998, 0.0170, 0.0022, 0.0021, 0.0028, 0.0028], + device='cuda:0'), in_proj_covar=tensor([0.0267, 0.0193, 0.0239, 0.0158, 0.0098, 0.0097, 0.0096, 0.0102], + device='cuda:0'), out_proj_covar=tensor([2.3552e-04, 1.8749e-04, 2.2075e-04, 1.5599e-04, 9.7018e-05, 9.6935e-05, + 9.4167e-05, 1.0375e-04], device='cuda:0') +2023-03-20 19:29:57,521 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-20 19:30:04,433 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.946e+02 3.394e+02 4.189e+02 5.011e+02 1.297e+03, threshold=8.377e+02, percent-clipped=4.0 +2023-03-20 19:30:13,318 INFO [train.py:901] (0/2) Epoch 6, batch 900, loss[loss=0.2267, simple_loss=0.2839, pruned_loss=0.08478, over 7281.00 frames. ], tot_loss[loss=0.2254, simple_loss=0.2824, pruned_loss=0.0842, over 1428072.85 frames. ], batch size: 57, lr: 2.46e-02, grad_scale: 16.0 +2023-03-20 19:30:18,547 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=6.60 vs. limit=5.0 +2023-03-20 19:30:36,480 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-20 19:30:37,565 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=15069.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:30:38,481 INFO [train.py:901] (0/2) Epoch 6, batch 950, loss[loss=0.2205, simple_loss=0.274, pruned_loss=0.08352, over 7350.00 frames. ], tot_loss[loss=0.2255, simple_loss=0.2828, pruned_loss=0.08411, over 1434141.71 frames. ], batch size: 61, lr: 2.46e-02, grad_scale: 16.0 +2023-03-20 19:30:51,134 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=15096.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:30:56,597 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.209e+02 3.393e+02 3.946e+02 5.314e+02 1.457e+03, threshold=7.892e+02, percent-clipped=4.0 +2023-03-20 19:31:00,777 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-20 19:31:02,821 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=15117.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:31:04,790 INFO [train.py:901] (0/2) Epoch 6, batch 1000, loss[loss=0.2308, simple_loss=0.2907, pruned_loss=0.0855, over 7318.00 frames. ], tot_loss[loss=0.2249, simple_loss=0.2824, pruned_loss=0.08368, over 1437663.14 frames. ], batch size: 75, lr: 2.46e-02, grad_scale: 16.0 +2023-03-20 19:31:15,975 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=15143.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:31:22,299 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-20 19:31:29,714 INFO [train.py:901] (0/2) Epoch 6, batch 1050, loss[loss=0.1922, simple_loss=0.2537, pruned_loss=0.06542, over 7254.00 frames. ], tot_loss[loss=0.2241, simple_loss=0.2815, pruned_loss=0.08342, over 1437412.17 frames. ], batch size: 47, lr: 2.45e-02, grad_scale: 16.0 +2023-03-20 19:31:31,297 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8976, 4.1333, 4.0133, 3.6805, 3.9762, 3.7752, 4.0030, 3.4720], + device='cuda:0'), covar=tensor([0.0057, 0.0099, 0.0063, 0.0081, 0.0070, 0.0074, 0.0082, 0.0113], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0030, 0.0030, 0.0027, 0.0028, 0.0029, 0.0035, 0.0034], + device='cuda:0'), out_proj_covar=tensor([8.1055e-05, 9.8660e-05, 1.0265e-04, 7.9911e-05, 8.9764e-05, 9.2059e-05, + 1.1499e-04, 1.0932e-04], device='cuda:0') +2023-03-20 19:31:36,317 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0217, 3.7317, 3.7566, 3.7946, 3.8003, 3.8385, 4.0866, 3.9060], + device='cuda:0'), covar=tensor([0.0121, 0.0161, 0.0194, 0.0155, 0.0175, 0.0124, 0.0160, 0.0120], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0045, 0.0049, 0.0042, 0.0059, 0.0053, 0.0049, 0.0044], + device='cuda:0'), out_proj_covar=tensor([9.5934e-05, 1.1897e-04, 1.2263e-04, 1.0484e-04, 1.5325e-04, 1.3736e-04, + 1.3457e-04, 1.0446e-04], device='cuda:0') +2023-03-20 19:31:39,823 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=15191.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:31:42,574 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=15195.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:31:43,466 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-20 19:31:48,944 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.607e+02 3.249e+02 4.260e+02 5.162e+02 1.307e+03, threshold=8.520e+02, percent-clipped=8.0 +2023-03-20 19:31:48,978 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-20 19:31:56,453 INFO [train.py:901] (0/2) Epoch 6, batch 1100, loss[loss=0.2235, simple_loss=0.2844, pruned_loss=0.08133, over 7117.00 frames. ], tot_loss[loss=0.2239, simple_loss=0.2813, pruned_loss=0.08321, over 1437559.98 frames. ], batch size: 98, lr: 2.45e-02, grad_scale: 16.0 +2023-03-20 19:32:03,724 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.8575, 1.2673, 1.2817, 1.0984, 0.8159, 0.8685, 0.9732, 1.2432], + device='cuda:0'), covar=tensor([0.0478, 0.0248, 0.0141, 0.0351, 0.0096, 0.0187, 0.0284, 0.0125], + device='cuda:0'), in_proj_covar=tensor([0.0017, 0.0016, 0.0015, 0.0015, 0.0016, 0.0017, 0.0017, 0.0018], + device='cuda:0'), out_proj_covar=tensor([2.1603e-05, 1.9991e-05, 2.0651e-05, 1.8497e-05, 1.8823e-05, 2.0881e-05, + 2.3199e-05, 2.5898e-05], device='cuda:0') +2023-03-20 19:32:14,269 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=15256.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 19:32:17,146 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-20 19:32:17,610 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 19:32:21,168 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5956, 4.4448, 4.4413, 4.6762, 4.9452, 4.7840, 4.1623, 4.2802], + device='cuda:0'), covar=tensor([0.0601, 0.1454, 0.1710, 0.1226, 0.0405, 0.1033, 0.0541, 0.0724], + device='cuda:0'), in_proj_covar=tensor([0.0120, 0.0194, 0.0189, 0.0162, 0.0133, 0.0201, 0.0113, 0.0141], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 19:32:21,208 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5782, 3.5553, 3.4661, 3.4754, 3.2850, 3.2238, 3.3303, 3.1456], + device='cuda:0'), covar=tensor([0.0057, 0.0116, 0.0084, 0.0062, 0.0112, 0.0102, 0.0112, 0.0122], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0031, 0.0030, 0.0027, 0.0029, 0.0029, 0.0037, 0.0034], + device='cuda:0'), out_proj_covar=tensor([8.0309e-05, 9.9370e-05, 1.0487e-04, 8.1524e-05, 9.4838e-05, 9.3649e-05, + 1.2046e-04, 1.1261e-04], device='cuda:0') +2023-03-20 19:32:21,573 INFO [train.py:901] (0/2) Epoch 6, batch 1150, loss[loss=0.2313, simple_loss=0.294, pruned_loss=0.08433, over 7304.00 frames. ], tot_loss[loss=0.2226, simple_loss=0.2797, pruned_loss=0.08273, over 1436248.46 frames. ], batch size: 68, lr: 2.44e-02, grad_scale: 16.0 +2023-03-20 19:32:30,758 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-20 19:32:32,352 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-20 19:32:40,307 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.354e+02 3.501e+02 4.043e+02 4.936e+02 1.554e+03, threshold=8.087e+02, percent-clipped=2.0 +2023-03-20 19:32:42,170 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-03-20 19:32:47,934 INFO [train.py:901] (0/2) Epoch 6, batch 1200, loss[loss=0.2085, simple_loss=0.2727, pruned_loss=0.07213, over 7327.00 frames. ], tot_loss[loss=0.2231, simple_loss=0.2803, pruned_loss=0.08296, over 1438711.31 frames. ], batch size: 75, lr: 2.44e-02, grad_scale: 16.0 +2023-03-20 19:33:03,367 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-20 19:33:13,671 INFO [train.py:901] (0/2) Epoch 6, batch 1250, loss[loss=0.2189, simple_loss=0.2793, pruned_loss=0.07927, over 7296.00 frames. ], tot_loss[loss=0.2237, simple_loss=0.2811, pruned_loss=0.08311, over 1440507.27 frames. ], batch size: 80, lr: 2.44e-02, grad_scale: 16.0 +2023-03-20 19:33:26,916 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-20 19:33:27,034 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=15396.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:33:30,791 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-20 19:33:32,276 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.950e+02 3.210e+02 3.988e+02 5.290e+02 1.673e+03, threshold=7.977e+02, percent-clipped=3.0 +2023-03-20 19:33:32,326 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-20 19:33:34,524 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8732, 2.1189, 1.8306, 2.6507, 1.4552, 2.8006, 1.3661, 2.7962], + device='cuda:0'), covar=tensor([0.0061, 0.0568, 0.2018, 0.0045, 0.4044, 0.0072, 0.1048, 0.0070], + device='cuda:0'), in_proj_covar=tensor([0.0128, 0.0219, 0.0308, 0.0117, 0.0301, 0.0127, 0.0243, 0.0130], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0003, 0.0001, 0.0003, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 19:33:39,484 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1589, 1.8091, 1.1766, 1.4633, 1.0619, 0.9039, 0.9483, 1.1437], + device='cuda:0'), covar=tensor([0.0303, 0.0125, 0.0106, 0.0146, 0.0476, 0.0302, 0.0190, 0.0186], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0017, 0.0017, 0.0018, 0.0018, 0.0017, 0.0019, 0.0020], + device='cuda:0'), out_proj_covar=tensor([4.1942e-05, 3.6660e-05, 3.3750e-05, 3.4319e-05, 4.0351e-05, 3.7516e-05, + 4.0364e-05, 4.3026e-05], device='cuda:0') +2023-03-20 19:33:39,848 INFO [train.py:901] (0/2) Epoch 6, batch 1300, loss[loss=0.2542, simple_loss=0.3122, pruned_loss=0.09814, over 7134.00 frames. ], tot_loss[loss=0.2239, simple_loss=0.2815, pruned_loss=0.08317, over 1440315.67 frames. ], batch size: 98, lr: 2.43e-02, grad_scale: 16.0 +2023-03-20 19:33:49,479 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1541, 2.0614, 1.3047, 1.9274, 1.0456, 0.6879, 1.1091, 1.0198], + device='cuda:0'), covar=tensor([0.0770, 0.0256, 0.0243, 0.0129, 0.0856, 0.0833, 0.0223, 0.0383], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0017, 0.0017, 0.0018, 0.0018, 0.0017, 0.0019, 0.0020], + device='cuda:0'), out_proj_covar=tensor([4.2922e-05, 3.7532e-05, 3.4511e-05, 3.5122e-05, 4.0853e-05, 3.8316e-05, + 4.0714e-05, 4.3603e-05], device='cuda:0') +2023-03-20 19:33:51,383 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=15444.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:33:54,848 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-20 19:33:57,394 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-20 19:34:01,573 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-20 19:34:06,036 INFO [train.py:901] (0/2) Epoch 6, batch 1350, loss[loss=0.2184, simple_loss=0.2789, pruned_loss=0.07893, over 7283.00 frames. ], tot_loss[loss=0.2238, simple_loss=0.2818, pruned_loss=0.08291, over 1441305.19 frames. ], batch size: 57, lr: 2.43e-02, grad_scale: 16.0 +2023-03-20 19:34:12,040 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-20 19:34:21,311 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0154, 3.4366, 3.8411, 3.7797, 3.8931, 3.7840, 3.9612, 3.4980], + device='cuda:0'), covar=tensor([0.0119, 0.0207, 0.0153, 0.0152, 0.0165, 0.0120, 0.0160, 0.0209], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0044, 0.0048, 0.0043, 0.0058, 0.0051, 0.0048, 0.0044], + device='cuda:0'), out_proj_covar=tensor([9.5174e-05, 1.1586e-04, 1.1896e-04, 1.0717e-04, 1.5329e-04, 1.3307e-04, + 1.3207e-04, 1.0570e-04], device='cuda:0') +2023-03-20 19:34:23,683 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.201e+02 3.158e+02 3.725e+02 4.655e+02 8.945e+02, threshold=7.450e+02, percent-clipped=1.0 +2023-03-20 19:34:24,851 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.7282, 1.0251, 1.1072, 1.1544, 0.7715, 0.8121, 0.8299, 1.0002], + device='cuda:0'), covar=tensor([0.0939, 0.0677, 0.0361, 0.0423, 0.1152, 0.1173, 0.0325, 0.1100], + device='cuda:0'), in_proj_covar=tensor([0.0029, 0.0035, 0.0029, 0.0029, 0.0031, 0.0029, 0.0034, 0.0030], + device='cuda:0'), out_proj_covar=tensor([5.1776e-05, 6.7881e-05, 4.4180e-05, 4.5239e-05, 5.2930e-05, 5.2394e-05, + 5.7154e-05, 5.3999e-05], device='cuda:0') +2023-03-20 19:34:31,276 INFO [train.py:901] (0/2) Epoch 6, batch 1400, loss[loss=0.2377, simple_loss=0.2992, pruned_loss=0.08808, over 7260.00 frames. ], tot_loss[loss=0.2232, simple_loss=0.2812, pruned_loss=0.0826, over 1440897.74 frames. ], batch size: 77, lr: 2.43e-02, grad_scale: 16.0 +2023-03-20 19:34:45,799 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-20 19:34:47,830 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=15551.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 19:34:57,767 INFO [train.py:901] (0/2) Epoch 6, batch 1450, loss[loss=0.2304, simple_loss=0.2862, pruned_loss=0.0873, over 7332.00 frames. ], tot_loss[loss=0.2238, simple_loss=0.2818, pruned_loss=0.08289, over 1442916.05 frames. ], batch size: 54, lr: 2.42e-02, grad_scale: 16.0 +2023-03-20 19:35:08,883 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-20 19:35:15,836 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.277e+02 3.171e+02 3.854e+02 4.989e+02 1.002e+03, threshold=7.708e+02, percent-clipped=5.0 +2023-03-20 19:35:24,027 INFO [train.py:901] (0/2) Epoch 6, batch 1500, loss[loss=0.2233, simple_loss=0.2828, pruned_loss=0.08187, over 7296.00 frames. ], tot_loss[loss=0.2238, simple_loss=0.2819, pruned_loss=0.08291, over 1442840.30 frames. ], batch size: 86, lr: 2.42e-02, grad_scale: 16.0 +2023-03-20 19:35:25,021 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-20 19:35:36,754 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0615, 2.4109, 2.7931, 2.3494, 2.3875, 2.4300, 1.7250, 2.1719], + device='cuda:0'), covar=tensor([0.1491, 0.0275, 0.0901, 0.2801, 0.0985, 0.1622, 0.3250, 0.1783], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0023, 0.0025, 0.0028, 0.0024, 0.0024, 0.0033, 0.0025], + device='cuda:0'), out_proj_covar=tensor([7.0901e-05, 6.2099e-05, 7.1921e-05, 7.6543e-05, 6.7242e-05, 7.1970e-05, + 8.5552e-05, 7.0782e-05], device='cuda:0') +2023-03-20 19:35:48,797 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-20 19:35:49,782 INFO [train.py:901] (0/2) Epoch 6, batch 1550, loss[loss=0.2423, simple_loss=0.3, pruned_loss=0.09233, over 7322.00 frames. ], tot_loss[loss=0.2244, simple_loss=0.2824, pruned_loss=0.08318, over 1444944.88 frames. ], batch size: 49, lr: 2.42e-02, grad_scale: 16.0 +2023-03-20 19:36:07,855 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.699e+02 3.513e+02 4.071e+02 5.180e+02 1.061e+03, threshold=8.142e+02, percent-clipped=4.0 +2023-03-20 19:36:12,337 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.22 vs. limit=5.0 +2023-03-20 19:36:16,049 INFO [train.py:901] (0/2) Epoch 6, batch 1600, loss[loss=0.2191, simple_loss=0.2788, pruned_loss=0.07969, over 7321.00 frames. ], tot_loss[loss=0.2233, simple_loss=0.2815, pruned_loss=0.08256, over 1445947.20 frames. ], batch size: 80, lr: 2.41e-02, grad_scale: 16.0 +2023-03-20 19:36:16,358 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.99 vs. limit=2.0 +2023-03-20 19:36:20,516 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-20 19:36:21,018 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-20 19:36:24,556 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-20 19:36:34,102 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-20 19:36:34,861 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.11 vs. limit=2.0 +2023-03-20 19:36:38,570 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.89 vs. limit=2.0 +2023-03-20 19:36:38,722 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-20 19:36:41,170 INFO [train.py:901] (0/2) Epoch 6, batch 1650, loss[loss=0.217, simple_loss=0.2748, pruned_loss=0.0796, over 7313.00 frames. ], tot_loss[loss=0.2234, simple_loss=0.2818, pruned_loss=0.08253, over 1446185.08 frames. ], batch size: 49, lr: 2.41e-02, grad_scale: 16.0 +2023-03-20 19:36:46,695 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-20 19:36:51,916 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2409, 4.1647, 3.8945, 3.4969, 4.0574, 2.5772, 1.7481, 4.2336], + device='cuda:0'), covar=tensor([0.0016, 0.0072, 0.0063, 0.0072, 0.0026, 0.0387, 0.0777, 0.0038], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0049, 0.0070, 0.0055, 0.0056, 0.0083, 0.0096, 0.0055], + device='cuda:0'), out_proj_covar=tensor([6.2334e-05, 7.7771e-05, 1.0136e-04, 8.1126e-05, 7.3897e-05, 1.2243e-04, + 1.4317e-04, 7.6029e-05], device='cuda:0') +2023-03-20 19:36:59,800 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.745e+02 3.076e+02 3.798e+02 5.090e+02 1.024e+03, threshold=7.597e+02, percent-clipped=1.0 +2023-03-20 19:37:05,406 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 19:37:07,891 INFO [train.py:901] (0/2) Epoch 6, batch 1700, loss[loss=0.1958, simple_loss=0.2544, pruned_loss=0.06855, over 7203.00 frames. ], tot_loss[loss=0.2222, simple_loss=0.2804, pruned_loss=0.08197, over 1446365.22 frames. ], batch size: 39, lr: 2.41e-02, grad_scale: 16.0 +2023-03-20 19:37:09,394 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-20 19:37:13,184 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.06 vs. limit=2.0 +2023-03-20 19:37:19,461 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-20 19:37:23,033 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=15851.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:37:32,977 INFO [train.py:901] (0/2) Epoch 6, batch 1750, loss[loss=0.1785, simple_loss=0.2209, pruned_loss=0.06811, over 6395.00 frames. ], tot_loss[loss=0.2223, simple_loss=0.2803, pruned_loss=0.08214, over 1443467.23 frames. ], batch size: 28, lr: 2.40e-02, grad_scale: 32.0 +2023-03-20 19:37:39,712 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2999, 4.2960, 4.1901, 4.6292, 4.6647, 4.6279, 4.1384, 4.1256], + device='cuda:0'), covar=tensor([0.0639, 0.1216, 0.1542, 0.0775, 0.0447, 0.0962, 0.0505, 0.0690], + device='cuda:0'), in_proj_covar=tensor([0.0129, 0.0197, 0.0188, 0.0164, 0.0136, 0.0211, 0.0114, 0.0141], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 19:37:39,788 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=15883.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:37:43,107 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-20 19:37:44,114 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-20 19:37:48,244 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=15899.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:37:51,728 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.174e+02 3.400e+02 4.133e+02 5.525e+02 1.160e+03, threshold=8.266e+02, percent-clipped=5.0 +2023-03-20 19:37:59,240 INFO [train.py:901] (0/2) Epoch 6, batch 1800, loss[loss=0.2246, simple_loss=0.2846, pruned_loss=0.08232, over 7255.00 frames. ], tot_loss[loss=0.2225, simple_loss=0.2806, pruned_loss=0.08226, over 1444865.15 frames. ], batch size: 47, lr: 2.40e-02, grad_scale: 32.0 +2023-03-20 19:38:03,452 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1266, 4.2310, 3.7344, 3.3432, 3.7708, 2.3762, 2.0450, 4.0910], + device='cuda:0'), covar=tensor([0.0016, 0.0016, 0.0064, 0.0102, 0.0053, 0.0392, 0.0648, 0.0033], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0047, 0.0069, 0.0055, 0.0056, 0.0082, 0.0094, 0.0056], + device='cuda:0'), out_proj_covar=tensor([6.3017e-05, 7.5720e-05, 1.0051e-04, 8.2677e-05, 7.6071e-05, 1.2087e-04, + 1.4056e-04, 7.6950e-05], device='cuda:0') +2023-03-20 19:38:05,886 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-20 19:38:11,001 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=15944.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:38:19,485 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-20 19:38:25,100 INFO [train.py:901] (0/2) Epoch 6, batch 1850, loss[loss=0.2436, simple_loss=0.3023, pruned_loss=0.09245, over 7266.00 frames. ], tot_loss[loss=0.2214, simple_loss=0.2799, pruned_loss=0.08151, over 1444428.11 frames. ], batch size: 64, lr: 2.40e-02, grad_scale: 32.0 +2023-03-20 19:38:30,081 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-20 19:38:39,114 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.75 vs. limit=5.0 +2023-03-20 19:38:40,549 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-16000.pt +2023-03-20 19:38:44,623 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=16001.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:38:46,992 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.936e+02 3.006e+02 3.753e+02 4.751e+02 8.909e+02, threshold=7.505e+02, percent-clipped=1.0 +2023-03-20 19:38:50,552 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-20 19:38:51,620 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9123, 3.2210, 3.4496, 3.4795, 3.5237, 3.5162, 3.6628, 3.5771], + device='cuda:0'), covar=tensor([0.0084, 0.0230, 0.0165, 0.0192, 0.0210, 0.0135, 0.0174, 0.0127], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0046, 0.0050, 0.0042, 0.0061, 0.0052, 0.0048, 0.0045], + device='cuda:0'), out_proj_covar=tensor([9.3009e-05, 1.2126e-04, 1.2599e-04, 1.0608e-04, 1.5964e-04, 1.3689e-04, + 1.3480e-04, 1.0798e-04], device='cuda:0') +2023-03-20 19:38:54,483 INFO [train.py:901] (0/2) Epoch 6, batch 1900, loss[loss=0.2231, simple_loss=0.2844, pruned_loss=0.08087, over 7257.00 frames. ], tot_loss[loss=0.2214, simple_loss=0.2801, pruned_loss=0.08136, over 1444949.26 frames. ], batch size: 55, lr: 2.39e-02, grad_scale: 32.0 +2023-03-20 19:39:15,749 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-20 19:39:15,846 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=16062.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:39:20,794 INFO [train.py:901] (0/2) Epoch 6, batch 1950, loss[loss=0.2342, simple_loss=0.2892, pruned_loss=0.08962, over 7343.00 frames. ], tot_loss[loss=0.2223, simple_loss=0.2803, pruned_loss=0.08213, over 1439186.04 frames. ], batch size: 73, lr: 2.39e-02, grad_scale: 16.0 +2023-03-20 19:39:27,833 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-20 19:39:27,964 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=16085.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:39:32,271 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-20 19:39:32,775 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-20 19:39:38,700 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.218e+02 3.583e+02 4.318e+02 5.667e+02 1.031e+03, threshold=8.635e+02, percent-clipped=6.0 +2023-03-20 19:39:45,758 INFO [train.py:901] (0/2) Epoch 6, batch 2000, loss[loss=0.2395, simple_loss=0.2938, pruned_loss=0.09266, over 7292.00 frames. ], tot_loss[loss=0.2221, simple_loss=0.2801, pruned_loss=0.08203, over 1439383.33 frames. ], batch size: 57, lr: 2.39e-02, grad_scale: 16.0 +2023-03-20 19:39:46,441 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1307, 2.6845, 2.3831, 3.4263, 2.8722, 3.2757, 2.9067, 2.4848], + device='cuda:0'), covar=tensor([0.0876, 0.0324, 0.1033, 0.0153, 0.0021, 0.0022, 0.0039, 0.0044], + device='cuda:0'), in_proj_covar=tensor([0.0286, 0.0202, 0.0258, 0.0179, 0.0101, 0.0103, 0.0104, 0.0109], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 19:39:48,264 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-20 19:39:59,037 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5252, 3.6019, 3.5009, 3.5291, 3.4597, 3.5134, 3.8933, 3.9353], + device='cuda:0'), covar=tensor([0.0227, 0.0191, 0.0231, 0.0198, 0.0331, 0.0424, 0.0273, 0.0219], + device='cuda:0'), in_proj_covar=tensor([0.0081, 0.0085, 0.0077, 0.0088, 0.0087, 0.0069, 0.0066, 0.0068], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 19:39:59,102 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=16146.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:39:59,575 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9529, 4.1374, 3.9969, 3.6027, 3.6902, 3.7400, 3.7789, 3.9027], + device='cuda:0'), covar=tensor([0.0047, 0.0062, 0.0055, 0.0062, 0.0067, 0.0063, 0.0078, 0.0064], + device='cuda:0'), in_proj_covar=tensor([0.0026, 0.0031, 0.0030, 0.0027, 0.0028, 0.0029, 0.0036, 0.0033], + device='cuda:0'), out_proj_covar=tensor([7.8704e-05, 9.9221e-05, 1.0478e-04, 7.9487e-05, 8.8409e-05, 9.0406e-05, + 1.1904e-04, 1.0653e-04], device='cuda:0') +2023-03-20 19:39:59,968 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-20 19:40:00,605 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=16149.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:40:09,661 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-20 19:40:12,236 INFO [train.py:901] (0/2) Epoch 6, batch 2050, loss[loss=0.205, simple_loss=0.2746, pruned_loss=0.06772, over 7339.00 frames. ], tot_loss[loss=0.2207, simple_loss=0.279, pruned_loss=0.08123, over 1440133.11 frames. ], batch size: 54, lr: 2.38e-02, grad_scale: 16.0 +2023-03-20 19:40:30,494 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.959e+02 3.318e+02 4.048e+02 5.137e+02 9.969e+02, threshold=8.096e+02, percent-clipped=1.0 +2023-03-20 19:40:32,172 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=16210.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:40:37,595 INFO [train.py:901] (0/2) Epoch 6, batch 2100, loss[loss=0.2317, simple_loss=0.2805, pruned_loss=0.09144, over 7289.00 frames. ], tot_loss[loss=0.2198, simple_loss=0.2783, pruned_loss=0.08062, over 1442359.26 frames. ], batch size: 68, lr: 2.38e-02, grad_scale: 16.0 +2023-03-20 19:40:42,789 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-20 19:40:45,279 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-20 19:40:47,408 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=16239.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:40:52,874 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.14 vs. limit=5.0 +2023-03-20 19:41:04,149 INFO [train.py:901] (0/2) Epoch 6, batch 2150, loss[loss=0.1699, simple_loss=0.2239, pruned_loss=0.05797, over 7085.00 frames. ], tot_loss[loss=0.2204, simple_loss=0.279, pruned_loss=0.08088, over 1442677.31 frames. ], batch size: 35, lr: 2.38e-02, grad_scale: 16.0 +2023-03-20 19:41:09,793 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5434, 4.9407, 5.0075, 4.8624, 4.7785, 4.4490, 4.9995, 4.8053], + device='cuda:0'), covar=tensor([0.0383, 0.0360, 0.0387, 0.0422, 0.0321, 0.0333, 0.0335, 0.0481], + device='cuda:0'), in_proj_covar=tensor([0.0121, 0.0121, 0.0103, 0.0096, 0.0087, 0.0119, 0.0109, 0.0085], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 19:41:09,855 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0529, 2.0495, 2.0367, 2.3168, 1.8217, 2.1240, 2.0245, 2.1565], + device='cuda:0'), covar=tensor([0.0624, 0.0696, 0.0594, 0.0410, 0.1815, 0.0611, 0.0821, 0.0963], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0034, 0.0035, 0.0035, 0.0033, 0.0036, 0.0033, 0.0035], + device='cuda:0'), out_proj_covar=tensor([9.5947e-05, 8.4025e-05, 8.5177e-05, 8.5166e-05, 8.4187e-05, 8.8478e-05, + 8.3896e-05, 8.6188e-05], device='cuda:0') +2023-03-20 19:41:11,795 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2821, 3.4450, 3.2870, 2.6823, 2.9375, 3.2523, 1.9749, 2.3809], + device='cuda:0'), covar=tensor([0.0940, 0.0196, 0.1141, 0.3697, 0.1004, 0.1716, 0.3013, 0.2056], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0025, 0.0028, 0.0030, 0.0025, 0.0026, 0.0034, 0.0027], + device='cuda:0'), out_proj_covar=tensor([7.4271e-05, 6.9149e-05, 7.8796e-05, 8.3427e-05, 7.2452e-05, 7.8042e-05, + 9.3021e-05, 7.7256e-05], device='cuda:0') +2023-03-20 19:41:16,029 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.01 vs. limit=2.0 +2023-03-20 19:41:22,164 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.745e+02 3.153e+02 4.020e+02 4.904e+02 9.769e+02, threshold=8.040e+02, percent-clipped=2.0 +2023-03-20 19:41:29,915 INFO [train.py:901] (0/2) Epoch 6, batch 2200, loss[loss=0.2087, simple_loss=0.2672, pruned_loss=0.0751, over 7206.00 frames. ], tot_loss[loss=0.22, simple_loss=0.2785, pruned_loss=0.0807, over 1441452.43 frames. ], batch size: 45, lr: 2.37e-02, grad_scale: 16.0 +2023-03-20 19:41:33,927 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-20 19:41:48,565 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=16357.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:41:48,578 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0034, 4.2178, 4.1512, 4.1967, 4.0335, 4.4240, 4.4626, 4.6721], + device='cuda:0'), covar=tensor([0.0209, 0.0157, 0.0231, 0.0142, 0.0311, 0.0144, 0.0312, 0.0186], + device='cuda:0'), in_proj_covar=tensor([0.0085, 0.0088, 0.0082, 0.0095, 0.0089, 0.0072, 0.0069, 0.0072], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:0') +2023-03-20 19:41:54,604 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9966, 2.0901, 1.8683, 1.9957, 1.9867, 1.8928, 1.7390, 2.0821], + device='cuda:0'), covar=tensor([0.0590, 0.0342, 0.0850, 0.0515, 0.0625, 0.0748, 0.0949, 0.0820], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0034, 0.0035, 0.0036, 0.0033, 0.0036, 0.0034, 0.0035], + device='cuda:0'), out_proj_covar=tensor([9.6862e-05, 8.4490e-05, 8.6197e-05, 8.7021e-05, 8.5336e-05, 8.9348e-05, + 8.5916e-05, 8.7424e-05], device='cuda:0') +2023-03-20 19:41:55,489 INFO [train.py:901] (0/2) Epoch 6, batch 2250, loss[loss=0.1968, simple_loss=0.2586, pruned_loss=0.0675, over 7237.00 frames. ], tot_loss[loss=0.2201, simple_loss=0.2784, pruned_loss=0.08088, over 1443039.84 frames. ], batch size: 55, lr: 2.37e-02, grad_scale: 16.0 +2023-03-20 19:41:55,599 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=16371.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:42:07,521 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-20 19:42:08,033 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-20 19:42:14,074 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.726e+02 3.600e+02 4.424e+02 5.691e+02 1.633e+03, threshold=8.848e+02, percent-clipped=4.0 +2023-03-20 19:42:20,707 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-20 19:42:21,195 INFO [train.py:901] (0/2) Epoch 6, batch 2300, loss[loss=0.2377, simple_loss=0.297, pruned_loss=0.08916, over 7314.00 frames. ], tot_loss[loss=0.2209, simple_loss=0.2792, pruned_loss=0.08132, over 1440866.81 frames. ], batch size: 80, lr: 2.37e-02, grad_scale: 16.0 +2023-03-20 19:42:27,514 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=16432.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:42:32,005 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=16441.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:42:32,552 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7683, 3.8489, 4.3640, 4.0851, 4.2646, 4.3520, 4.6670, 4.1713], + device='cuda:0'), covar=tensor([0.0061, 0.0157, 0.0106, 0.0121, 0.0145, 0.0078, 0.0110, 0.0114], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0042, 0.0048, 0.0041, 0.0057, 0.0051, 0.0045, 0.0043], + device='cuda:0'), out_proj_covar=tensor([9.1529e-05, 1.1161e-04, 1.1999e-04, 1.0443e-04, 1.5036e-04, 1.3525e-04, + 1.2757e-04, 1.0549e-04], device='cuda:0') +2023-03-20 19:42:46,882 INFO [train.py:901] (0/2) Epoch 6, batch 2350, loss[loss=0.2259, simple_loss=0.2823, pruned_loss=0.0847, over 7354.00 frames. ], tot_loss[loss=0.2207, simple_loss=0.279, pruned_loss=0.08116, over 1441214.66 frames. ], batch size: 63, lr: 2.36e-02, grad_scale: 16.0 +2023-03-20 19:43:04,547 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=16505.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:43:05,504 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.961e+02 3.138e+02 3.927e+02 4.655e+02 8.908e+02, threshold=7.854e+02, percent-clipped=1.0 +2023-03-20 19:43:06,545 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-20 19:43:13,380 INFO [train.py:901] (0/2) Epoch 6, batch 2400, loss[loss=0.2174, simple_loss=0.2767, pruned_loss=0.07905, over 7256.00 frames. ], tot_loss[loss=0.2188, simple_loss=0.2774, pruned_loss=0.08007, over 1439590.64 frames. ], batch size: 52, lr: 2.36e-02, grad_scale: 16.0 +2023-03-20 19:43:13,929 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-20 19:43:20,283 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.93 vs. limit=2.0 +2023-03-20 19:43:22,470 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-20 19:43:22,549 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=16539.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:43:25,410 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-20 19:43:28,941 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5961, 4.5610, 4.3507, 4.8886, 4.8926, 4.9631, 4.4147, 4.3274], + device='cuda:0'), covar=tensor([0.0727, 0.1449, 0.1853, 0.1034, 0.0512, 0.1015, 0.0519, 0.0842], + device='cuda:0'), in_proj_covar=tensor([0.0124, 0.0194, 0.0182, 0.0165, 0.0131, 0.0211, 0.0115, 0.0138], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 19:43:38,663 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9349, 3.1128, 2.7634, 2.4686, 3.0088, 2.9881, 2.4255, 2.5358], + device='cuda:0'), covar=tensor([0.1054, 0.0265, 0.1651, 0.3827, 0.0528, 0.1063, 0.2336, 0.1925], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0024, 0.0027, 0.0029, 0.0024, 0.0025, 0.0033, 0.0026], + device='cuda:0'), out_proj_covar=tensor([7.5484e-05, 6.8371e-05, 7.7906e-05, 8.1477e-05, 7.1063e-05, 7.6556e-05, + 9.1182e-05, 7.7936e-05], device='cuda:0') +2023-03-20 19:43:39,047 INFO [train.py:901] (0/2) Epoch 6, batch 2450, loss[loss=0.2306, simple_loss=0.2825, pruned_loss=0.08939, over 7231.00 frames. ], tot_loss[loss=0.2192, simple_loss=0.2776, pruned_loss=0.08033, over 1440144.26 frames. ], batch size: 45, lr: 2.36e-02, grad_scale: 16.0 +2023-03-20 19:43:47,154 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=16587.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:43:49,640 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5310, 4.6946, 4.5951, 4.9418, 5.1183, 5.0781, 4.4841, 4.5639], + device='cuda:0'), covar=tensor([0.0844, 0.1359, 0.1742, 0.0847, 0.0467, 0.1032, 0.0606, 0.0710], + device='cuda:0'), in_proj_covar=tensor([0.0126, 0.0194, 0.0185, 0.0167, 0.0134, 0.0217, 0.0117, 0.0139], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 19:43:51,938 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.01 vs. limit=2.0 +2023-03-20 19:43:52,634 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-20 19:43:58,204 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.525e+02 3.302e+02 4.000e+02 4.919e+02 9.001e+02, threshold=8.000e+02, percent-clipped=2.0 +2023-03-20 19:44:05,300 INFO [train.py:901] (0/2) Epoch 6, batch 2500, loss[loss=0.2266, simple_loss=0.2888, pruned_loss=0.0822, over 7246.00 frames. ], tot_loss[loss=0.2196, simple_loss=0.278, pruned_loss=0.08061, over 1440412.80 frames. ], batch size: 55, lr: 2.35e-02, grad_scale: 16.0 +2023-03-20 19:44:18,943 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-20 19:44:24,213 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=16657.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:44:31,137 INFO [train.py:901] (0/2) Epoch 6, batch 2550, loss[loss=0.2071, simple_loss=0.2785, pruned_loss=0.06784, over 7265.00 frames. ], tot_loss[loss=0.2188, simple_loss=0.2775, pruned_loss=0.08009, over 1442137.04 frames. ], batch size: 52, lr: 2.35e-02, grad_scale: 16.0 +2023-03-20 19:44:32,590 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.89 vs. limit=2.0 +2023-03-20 19:44:49,002 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=16705.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:44:49,939 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.726e+02 3.249e+02 3.789e+02 4.913e+02 1.049e+03, threshold=7.579e+02, percent-clipped=3.0 +2023-03-20 19:44:56,968 INFO [train.py:901] (0/2) Epoch 6, batch 2600, loss[loss=0.2092, simple_loss=0.2714, pruned_loss=0.07352, over 7221.00 frames. ], tot_loss[loss=0.2181, simple_loss=0.2771, pruned_loss=0.0795, over 1442515.53 frames. ], batch size: 93, lr: 2.35e-02, grad_scale: 16.0 +2023-03-20 19:44:57,219 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.56 vs. limit=5.0 +2023-03-20 19:44:59,923 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=16727.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:45:06,796 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=16741.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:45:08,784 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9651, 1.7716, 1.8667, 2.0707, 1.6887, 1.8031, 2.1174, 1.9492], + device='cuda:0'), covar=tensor([0.0598, 0.1099, 0.0759, 0.0607, 0.1042, 0.0500, 0.0848, 0.0525], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0036, 0.0034, 0.0037, 0.0034, 0.0035, 0.0032, 0.0036], + device='cuda:0'), out_proj_covar=tensor([9.8487e-05, 8.9944e-05, 8.5102e-05, 9.0104e-05, 8.7469e-05, 8.7752e-05, + 8.2006e-05, 9.0201e-05], device='cuda:0') +2023-03-20 19:45:21,582 INFO [train.py:901] (0/2) Epoch 6, batch 2650, loss[loss=0.2375, simple_loss=0.2932, pruned_loss=0.09092, over 7348.00 frames. ], tot_loss[loss=0.2178, simple_loss=0.2768, pruned_loss=0.07942, over 1442392.95 frames. ], batch size: 63, lr: 2.34e-02, grad_scale: 16.0 +2023-03-20 19:45:30,428 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=16789.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:45:39,174 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=16805.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:45:40,048 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.977e+02 3.511e+02 4.230e+02 5.351e+02 9.415e+02, threshold=8.460e+02, percent-clipped=2.0 +2023-03-20 19:45:46,931 INFO [train.py:901] (0/2) Epoch 6, batch 2700, loss[loss=0.2478, simple_loss=0.2918, pruned_loss=0.1019, over 7225.00 frames. ], tot_loss[loss=0.2183, simple_loss=0.277, pruned_loss=0.07982, over 1441840.37 frames. ], batch size: 93, lr: 2.34e-02, grad_scale: 16.0 +2023-03-20 19:45:54,080 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0047, 0.7478, 0.7670, 1.0193, 0.7130, 0.8559, 0.9094, 0.8086], + device='cuda:0'), covar=tensor([0.0113, 0.0094, 0.0214, 0.0073, 0.0122, 0.0139, 0.0165, 0.0230], + device='cuda:0'), in_proj_covar=tensor([0.0017, 0.0016, 0.0015, 0.0015, 0.0016, 0.0017, 0.0016, 0.0019], + device='cuda:0'), out_proj_covar=tensor([2.2010e-05, 1.8714e-05, 2.0513e-05, 1.7826e-05, 1.8986e-05, 2.0791e-05, + 2.1752e-05, 2.7015e-05], device='cuda:0') +2023-03-20 19:45:58,855 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3402, 2.1132, 2.1333, 1.8166, 2.3252, 2.1934, 2.5503, 1.9476], + device='cuda:0'), covar=tensor([0.0521, 0.1571, 0.0699, 0.1033, 0.1615, 0.0437, 0.1702, 0.1516], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0034, 0.0034, 0.0036, 0.0031, 0.0033, 0.0031, 0.0035], + device='cuda:0'), out_proj_covar=tensor([9.5501e-05, 8.6265e-05, 8.4192e-05, 9.0036e-05, 8.2255e-05, 8.4520e-05, + 8.0241e-05, 8.8162e-05], device='cuda:0') +2023-03-20 19:46:02,720 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=16853.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:46:10,336 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=16868.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 19:46:11,761 INFO [train.py:901] (0/2) Epoch 6, batch 2750, loss[loss=0.2004, simple_loss=0.267, pruned_loss=0.0669, over 7273.00 frames. ], tot_loss[loss=0.2177, simple_loss=0.2764, pruned_loss=0.07948, over 1440973.90 frames. ], batch size: 52, lr: 2.34e-02, grad_scale: 16.0 +2023-03-20 19:46:28,958 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=16906.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:46:29,264 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.247e+02 3.398e+02 4.341e+02 5.354e+02 9.122e+02, threshold=8.682e+02, percent-clipped=2.0 +2023-03-20 19:46:30,839 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1343, 2.4310, 2.3130, 2.1782, 2.5662, 2.5795, 1.9483, 2.3447], + device='cuda:0'), covar=tensor([0.2345, 0.0330, 0.2156, 0.3062, 0.0928, 0.1563, 0.2865, 0.2851], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0024, 0.0028, 0.0029, 0.0024, 0.0026, 0.0035, 0.0028], + device='cuda:0'), out_proj_covar=tensor([8.0144e-05, 7.0479e-05, 8.3451e-05, 8.5684e-05, 7.4988e-05, 8.0174e-05, + 9.7627e-05, 8.3011e-05], device='cuda:0') +2023-03-20 19:46:36,490 INFO [train.py:901] (0/2) Epoch 6, batch 2800, loss[loss=0.224, simple_loss=0.2866, pruned_loss=0.08069, over 7262.00 frames. ], tot_loss[loss=0.218, simple_loss=0.2767, pruned_loss=0.07968, over 1442905.78 frames. ], batch size: 64, lr: 2.33e-02, grad_scale: 16.0 +2023-03-20 19:46:40,548 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=16929.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 19:46:45,437 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.46 vs. limit=5.0 +2023-03-20 19:46:48,924 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-6.pt +2023-03-20 19:47:06,721 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-20 19:47:06,777 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-20 19:47:06,805 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-20 19:47:06,824 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-20 19:47:07,008 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-20 19:47:07,024 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-20 19:47:07,035 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-20 19:47:07,143 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-20 19:47:07,222 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-20 19:47:07,239 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-20 19:47:07,257 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-20 19:47:07,774 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-20 19:47:07,780 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-20 19:47:07,788 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-20 19:47:07,936 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-20 19:47:07,939 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-20 19:47:08,004 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-20 19:47:10,295 INFO [train.py:901] (0/2) Epoch 7, batch 0, loss[loss=0.1969, simple_loss=0.2586, pruned_loss=0.06764, over 7348.00 frames. ], tot_loss[loss=0.1969, simple_loss=0.2586, pruned_loss=0.06764, over 7348.00 frames. ], batch size: 44, lr: 2.24e-02, grad_scale: 16.0 +2023-03-20 19:47:10,296 INFO [train.py:926] (0/2) Computing validation loss +2023-03-20 19:47:26,779 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.9529, 1.2771, 1.5532, 1.2527, 1.1361, 1.4132, 1.0012, 1.3168], + device='cuda:0'), covar=tensor([0.0864, 0.0594, 0.0409, 0.0188, 0.0288, 0.0344, 0.0378, 0.0503], + device='cuda:0'), in_proj_covar=tensor([0.0021, 0.0020, 0.0020, 0.0020, 0.0021, 0.0019, 0.0020, 0.0023], + device='cuda:0'), out_proj_covar=tensor([4.8013e-05, 4.3185e-05, 4.0949e-05, 3.9016e-05, 4.7489e-05, 4.3493e-05, + 4.5166e-05, 5.0997e-05], device='cuda:0') +2023-03-20 19:47:36,469 INFO [train.py:935] (0/2) Epoch 7, validation: loss=0.1866, simple_loss=0.272, pruned_loss=0.05057, over 1622729.00 frames. +2023-03-20 19:47:36,470 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-20 19:47:43,608 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-20 19:47:47,777 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=16967.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 19:47:54,604 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-20 19:48:01,237 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-20 19:48:01,730 INFO [train.py:901] (0/2) Epoch 7, batch 50, loss[loss=0.1877, simple_loss=0.2549, pruned_loss=0.06024, over 7266.00 frames. ], tot_loss[loss=0.211, simple_loss=0.271, pruned_loss=0.0755, over 325976.48 frames. ], batch size: 47, lr: 2.23e-02, grad_scale: 16.0 +2023-03-20 19:48:03,651 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-20 19:48:07,128 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-20 19:48:07,443 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.95 vs. limit=2.0 +2023-03-20 19:48:08,746 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.771e+02 3.154e+02 3.992e+02 5.174e+02 1.123e+03, threshold=7.983e+02, percent-clipped=2.0 +2023-03-20 19:48:18,915 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=17027.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:48:25,434 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-20 19:48:25,889 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-20 19:48:28,410 INFO [train.py:901] (0/2) Epoch 7, batch 100, loss[loss=0.2256, simple_loss=0.2873, pruned_loss=0.082, over 7331.00 frames. ], tot_loss[loss=0.2159, simple_loss=0.2754, pruned_loss=0.07819, over 574575.63 frames. ], batch size: 61, lr: 2.23e-02, grad_scale: 16.0 +2023-03-20 19:48:43,318 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=17075.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:48:46,011 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.52 vs. limit=2.0 +2023-03-20 19:48:53,196 INFO [train.py:901] (0/2) Epoch 7, batch 150, loss[loss=0.2282, simple_loss=0.2846, pruned_loss=0.08587, over 7320.00 frames. ], tot_loss[loss=0.2182, simple_loss=0.2775, pruned_loss=0.07949, over 768059.80 frames. ], batch size: 61, lr: 2.23e-02, grad_scale: 16.0 +2023-03-20 19:48:59,721 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.829e+02 3.433e+02 4.214e+02 5.230e+02 1.024e+03, threshold=8.429e+02, percent-clipped=1.0 +2023-03-20 19:49:19,115 INFO [train.py:901] (0/2) Epoch 7, batch 200, loss[loss=0.2221, simple_loss=0.269, pruned_loss=0.08763, over 7228.00 frames. ], tot_loss[loss=0.2192, simple_loss=0.278, pruned_loss=0.08017, over 916319.96 frames. ], batch size: 45, lr: 2.22e-02, grad_scale: 16.0 +2023-03-20 19:49:25,201 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-20 19:49:29,203 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-20 19:49:35,111 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-20 19:49:37,035 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.92 vs. limit=2.0 +2023-03-20 19:49:41,672 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=17189.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:49:44,513 INFO [train.py:901] (0/2) Epoch 7, batch 250, loss[loss=0.2426, simple_loss=0.2977, pruned_loss=0.09378, over 7278.00 frames. ], tot_loss[loss=0.2191, simple_loss=0.2784, pruned_loss=0.07993, over 1033869.55 frames. ], batch size: 57, lr: 2.22e-02, grad_scale: 16.0 +2023-03-20 19:49:49,480 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-20 19:49:51,486 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.722e+02 3.314e+02 4.064e+02 4.801e+02 1.041e+03, threshold=8.128e+02, percent-clipped=1.0 +2023-03-20 19:49:58,490 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9821, 4.5374, 4.4750, 4.4301, 4.4146, 3.9694, 4.5311, 4.4177], + device='cuda:0'), covar=tensor([0.0446, 0.0396, 0.0487, 0.0469, 0.0329, 0.0382, 0.0336, 0.0435], + device='cuda:0'), in_proj_covar=tensor([0.0125, 0.0125, 0.0106, 0.0097, 0.0087, 0.0120, 0.0113, 0.0087], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 19:49:59,954 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=17224.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 19:50:08,341 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-20 19:50:10,274 INFO [train.py:901] (0/2) Epoch 7, batch 300, loss[loss=0.2213, simple_loss=0.2801, pruned_loss=0.08126, over 7327.00 frames. ], tot_loss[loss=0.2179, simple_loss=0.2774, pruned_loss=0.07919, over 1124136.77 frames. ], batch size: 61, lr: 2.22e-02, grad_scale: 16.0 +2023-03-20 19:50:13,010 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=17250.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:50:16,921 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-20 19:50:19,596 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=17262.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 19:50:20,146 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2021, 1.9263, 2.0393, 2.1975, 2.0088, 1.9270, 2.1409, 2.0846], + device='cuda:0'), covar=tensor([0.0684, 0.0659, 0.0715, 0.0573, 0.0633, 0.0385, 0.0660, 0.1130], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0033, 0.0033, 0.0035, 0.0032, 0.0034, 0.0032, 0.0035], + device='cuda:0'), out_proj_covar=tensor([9.7951e-05, 8.6623e-05, 8.4507e-05, 8.8059e-05, 8.4760e-05, 8.5345e-05, + 8.2840e-05, 8.7410e-05], device='cuda:0') +2023-03-20 19:50:20,646 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6250, 4.2864, 4.0951, 3.6358, 4.2178, 2.6786, 2.0712, 4.4427], + device='cuda:0'), covar=tensor([0.0012, 0.0079, 0.0055, 0.0061, 0.0024, 0.0369, 0.0637, 0.0032], + device='cuda:0'), in_proj_covar=tensor([0.0052, 0.0050, 0.0071, 0.0055, 0.0057, 0.0088, 0.0097, 0.0057], + device='cuda:0'), out_proj_covar=tensor([6.6851e-05, 7.9758e-05, 1.0180e-04, 8.1580e-05, 7.6586e-05, 1.3007e-04, + 1.4366e-04, 8.1472e-05], device='cuda:0') +2023-03-20 19:50:27,310 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.36 vs. limit=5.0 +2023-03-20 19:50:27,764 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2108, 4.0641, 3.2207, 3.9587, 2.7102, 2.5553, 4.0741, 3.2664], + device='cuda:0'), covar=tensor([0.0073, 0.0058, 0.0196, 0.0037, 0.0264, 0.0418, 0.0122, 0.0372], + device='cuda:0'), in_proj_covar=tensor([0.0209, 0.0196, 0.0216, 0.0163, 0.0276, 0.0287, 0.0215, 0.0288], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-20 19:50:36,767 INFO [train.py:901] (0/2) Epoch 7, batch 350, loss[loss=0.2098, simple_loss=0.2838, pruned_loss=0.06794, over 7299.00 frames. ], tot_loss[loss=0.2153, simple_loss=0.2754, pruned_loss=0.07765, over 1192956.77 frames. ], batch size: 86, lr: 2.22e-02, grad_scale: 16.0 +2023-03-20 19:50:42,826 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.792e+02 3.275e+02 4.052e+02 4.742e+02 1.008e+03, threshold=8.103e+02, percent-clipped=2.0 +2023-03-20 19:50:44,543 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0802, 3.2548, 2.1369, 3.5097, 2.8667, 3.2944, 2.1099, 1.9889], + device='cuda:0'), covar=tensor([0.0033, 0.0335, 0.0561, 0.0073, 0.0069, 0.0135, 0.0757, 0.0555], + device='cuda:0'), in_proj_covar=tensor([0.0123, 0.0160, 0.0269, 0.0140, 0.0154, 0.0148, 0.0257, 0.0257], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 19:50:44,979 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9759, 4.3822, 4.0225, 3.9287, 3.7605, 4.0988, 4.1242, 3.9041], + device='cuda:0'), covar=tensor([0.0046, 0.0051, 0.0056, 0.0060, 0.0056, 0.0051, 0.0064, 0.0080], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0033, 0.0033, 0.0029, 0.0030, 0.0031, 0.0040, 0.0036], + device='cuda:0'), out_proj_covar=tensor([8.1449e-05, 1.0456e-04, 1.1803e-04, 8.5441e-05, 9.2677e-05, 9.8532e-05, + 1.3356e-04, 1.1450e-04], device='cuda:0') +2023-03-20 19:50:47,718 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.01 vs. limit=2.0 +2023-03-20 19:50:52,931 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-20 19:51:01,930 INFO [train.py:901] (0/2) Epoch 7, batch 400, loss[loss=0.2337, simple_loss=0.2968, pruned_loss=0.08527, over 7290.00 frames. ], tot_loss[loss=0.2155, simple_loss=0.2757, pruned_loss=0.07765, over 1249109.84 frames. ], batch size: 57, lr: 2.21e-02, grad_scale: 16.0 +2023-03-20 19:51:11,160 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=17362.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:51:22,174 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=17383.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 19:51:24,167 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1343, 1.8370, 1.6022, 1.1887, 1.1517, 1.3511, 1.3161, 1.1414], + device='cuda:0'), covar=tensor([0.0460, 0.0191, 0.0184, 0.0094, 0.0384, 0.0648, 0.0518, 0.0286], + device='cuda:0'), in_proj_covar=tensor([0.0021, 0.0019, 0.0019, 0.0020, 0.0021, 0.0019, 0.0021, 0.0022], + device='cuda:0'), out_proj_covar=tensor([4.8531e-05, 4.3209e-05, 4.0541e-05, 3.9607e-05, 4.8038e-05, 4.3411e-05, + 4.6484e-05, 5.0252e-05], device='cuda:0') +2023-03-20 19:51:25,137 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3665, 4.2802, 4.1381, 4.5407, 4.6479, 4.6252, 4.1245, 4.1043], + device='cuda:0'), covar=tensor([0.0650, 0.1461, 0.1698, 0.0826, 0.0440, 0.0984, 0.0563, 0.0780], + device='cuda:0'), in_proj_covar=tensor([0.0126, 0.0203, 0.0187, 0.0166, 0.0130, 0.0216, 0.0120, 0.0143], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 19:51:28,075 INFO [train.py:901] (0/2) Epoch 7, batch 450, loss[loss=0.2653, simple_loss=0.3198, pruned_loss=0.1054, over 6777.00 frames. ], tot_loss[loss=0.2159, simple_loss=0.2763, pruned_loss=0.07774, over 1294303.38 frames. ], batch size: 106, lr: 2.21e-02, grad_scale: 16.0 +2023-03-20 19:51:34,463 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.675e+02 3.337e+02 4.059e+02 5.111e+02 8.541e+02, threshold=8.117e+02, percent-clipped=3.0 +2023-03-20 19:51:34,503 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-20 19:51:34,986 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-20 19:51:38,600 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8623, 2.4687, 1.9567, 2.8845, 1.5227, 2.6530, 1.2143, 2.5810], + device='cuda:0'), covar=tensor([0.0042, 0.0660, 0.1608, 0.0035, 0.3743, 0.0053, 0.0941, 0.0076], + device='cuda:0'), in_proj_covar=tensor([0.0136, 0.0245, 0.0322, 0.0122, 0.0305, 0.0138, 0.0265, 0.0146], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0003, 0.0001, 0.0003, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 19:51:42,618 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=17423.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:51:43,071 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0607, 4.5520, 4.5776, 4.4310, 4.3510, 3.9451, 4.5621, 4.3804], + device='cuda:0'), covar=tensor([0.0416, 0.0331, 0.0362, 0.0465, 0.0338, 0.0399, 0.0332, 0.0424], + device='cuda:0'), in_proj_covar=tensor([0.0124, 0.0124, 0.0106, 0.0098, 0.0086, 0.0120, 0.0114, 0.0089], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 19:51:46,578 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2951, 4.0141, 3.4977, 3.6059, 3.7974, 2.3431, 1.9480, 4.1412], + device='cuda:0'), covar=tensor([0.0014, 0.0023, 0.0088, 0.0051, 0.0037, 0.0409, 0.0613, 0.0031], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0051, 0.0072, 0.0054, 0.0059, 0.0092, 0.0099, 0.0059], + device='cuda:0'), out_proj_covar=tensor([6.8254e-05, 8.0381e-05, 1.0309e-04, 8.1078e-05, 7.9144e-05, 1.3563e-04, + 1.4667e-04, 8.4219e-05], device='cuda:0') +2023-03-20 19:51:53,710 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=17444.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 19:51:54,035 INFO [train.py:901] (0/2) Epoch 7, batch 500, loss[loss=0.1715, simple_loss=0.2268, pruned_loss=0.05811, over 6994.00 frames. ], tot_loss[loss=0.2159, simple_loss=0.2761, pruned_loss=0.07781, over 1327198.22 frames. ], batch size: 35, lr: 2.21e-02, grad_scale: 16.0 +2023-03-20 19:52:07,021 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-20 19:52:08,546 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-20 19:52:09,059 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-20 19:52:11,453 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-20 19:52:12,595 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2464, 3.4635, 2.3871, 3.5755, 2.9918, 3.4679, 2.2831, 2.0187], + device='cuda:0'), covar=tensor([0.0039, 0.0185, 0.0509, 0.0089, 0.0087, 0.0117, 0.0739, 0.0560], + device='cuda:0'), in_proj_covar=tensor([0.0121, 0.0155, 0.0265, 0.0138, 0.0155, 0.0145, 0.0248, 0.0253], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 19:52:15,809 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-20 19:52:19,304 INFO [train.py:901] (0/2) Epoch 7, batch 550, loss[loss=0.2321, simple_loss=0.2897, pruned_loss=0.0872, over 7278.00 frames. ], tot_loss[loss=0.216, simple_loss=0.2764, pruned_loss=0.0778, over 1353417.45 frames. ], batch size: 52, lr: 2.20e-02, grad_scale: 16.0 +2023-03-20 19:52:25,330 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.616e+02 3.083e+02 3.892e+02 4.518e+02 9.581e+02, threshold=7.783e+02, percent-clipped=1.0 +2023-03-20 19:52:25,849 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-20 19:52:29,432 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1190, 1.9545, 1.7838, 1.6595, 2.0499, 1.8653, 2.2715, 1.7192], + device='cuda:0'), covar=tensor([0.0379, 0.0667, 0.0858, 0.0754, 0.0453, 0.0419, 0.0395, 0.1055], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0034, 0.0034, 0.0036, 0.0033, 0.0036, 0.0032, 0.0036], + device='cuda:0'), out_proj_covar=tensor([1.0094e-04, 8.9052e-05, 8.7956e-05, 9.0703e-05, 8.7360e-05, 9.0916e-05, + 8.3495e-05, 9.1565e-05], device='cuda:0') +2023-03-20 19:52:34,476 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=17524.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 19:52:34,861 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-20 19:52:38,320 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-20 19:52:45,415 INFO [train.py:901] (0/2) Epoch 7, batch 600, loss[loss=0.2676, simple_loss=0.3161, pruned_loss=0.1095, over 6632.00 frames. ], tot_loss[loss=0.2169, simple_loss=0.277, pruned_loss=0.07837, over 1371102.56 frames. ], batch size: 106, lr: 2.20e-02, grad_scale: 8.0 +2023-03-20 19:52:45,439 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-20 19:52:45,500 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=17545.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:52:48,118 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6427, 3.8914, 2.3863, 3.8446, 3.2247, 3.8407, 2.5834, 2.1674], + device='cuda:0'), covar=tensor([0.0040, 0.0153, 0.0663, 0.0076, 0.0115, 0.0062, 0.0821, 0.0628], + device='cuda:0'), in_proj_covar=tensor([0.0124, 0.0162, 0.0275, 0.0144, 0.0159, 0.0146, 0.0256, 0.0261], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 19:52:54,212 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=17562.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 19:52:59,183 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=17572.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 19:53:02,027 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-20 19:53:09,978 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-20 19:53:10,473 INFO [train.py:901] (0/2) Epoch 7, batch 650, loss[loss=0.2067, simple_loss=0.273, pruned_loss=0.0702, over 7213.00 frames. ], tot_loss[loss=0.2164, simple_loss=0.2764, pruned_loss=0.0782, over 1386420.82 frames. ], batch size: 50, lr: 2.20e-02, grad_scale: 8.0 +2023-03-20 19:53:17,887 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.063e+02 3.125e+02 3.787e+02 4.523e+02 1.116e+03, threshold=7.575e+02, percent-clipped=2.0 +2023-03-20 19:53:18,972 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=17610.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:53:26,851 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-20 19:53:37,129 INFO [train.py:901] (0/2) Epoch 7, batch 700, loss[loss=0.2188, simple_loss=0.2781, pruned_loss=0.07975, over 7283.00 frames. ], tot_loss[loss=0.2156, simple_loss=0.2755, pruned_loss=0.07781, over 1399177.58 frames. ], batch size: 57, lr: 2.20e-02, grad_scale: 8.0 +2023-03-20 19:53:37,151 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-20 19:53:41,848 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1350, 3.6839, 3.7662, 3.7664, 3.7065, 3.7923, 4.1460, 3.3821], + device='cuda:0'), covar=tensor([0.0196, 0.0139, 0.0185, 0.0124, 0.0191, 0.0114, 0.0160, 0.0201], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0043, 0.0051, 0.0041, 0.0059, 0.0052, 0.0047, 0.0044], + device='cuda:0'), out_proj_covar=tensor([9.6327e-05, 1.1430e-04, 1.2723e-04, 1.0473e-04, 1.5269e-04, 1.3846e-04, + 1.3238e-04, 1.0777e-04], device='cuda:0') +2023-03-20 19:53:43,077 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.96 vs. limit=2.0 +2023-03-20 19:53:45,429 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4296, 3.6143, 3.3895, 3.5935, 3.5805, 3.5906, 3.8318, 3.9300], + device='cuda:0'), covar=tensor([0.0259, 0.0175, 0.0232, 0.0202, 0.0319, 0.0285, 0.0278, 0.0206], + device='cuda:0'), in_proj_covar=tensor([0.0090, 0.0090, 0.0083, 0.0094, 0.0092, 0.0073, 0.0073, 0.0073], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 19:54:02,015 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-20 19:54:02,502 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-20 19:54:02,976 INFO [train.py:901] (0/2) Epoch 7, batch 750, loss[loss=0.2161, simple_loss=0.2805, pruned_loss=0.07588, over 7260.00 frames. ], tot_loss[loss=0.2149, simple_loss=0.2744, pruned_loss=0.07769, over 1404559.26 frames. ], batch size: 64, lr: 2.19e-02, grad_scale: 8.0 +2023-03-20 19:54:05,521 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=17700.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:54:07,221 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.95 vs. limit=2.0 +2023-03-20 19:54:09,381 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.593e+02 3.256e+02 3.901e+02 5.020e+02 1.278e+03, threshold=7.802e+02, percent-clipped=1.0 +2023-03-20 19:54:14,990 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=17718.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:54:16,934 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-20 19:54:20,954 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-20 19:54:21,054 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=17730.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:54:25,498 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=17739.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 19:54:27,458 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-20 19:54:28,405 INFO [train.py:901] (0/2) Epoch 7, batch 800, loss[loss=0.2023, simple_loss=0.2521, pruned_loss=0.07624, over 7158.00 frames. ], tot_loss[loss=0.2152, simple_loss=0.2746, pruned_loss=0.07789, over 1413706.93 frames. ], batch size: 39, lr: 2.19e-02, grad_scale: 8.0 +2023-03-20 19:54:28,429 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-20 19:54:32,511 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8968, 4.0361, 4.0878, 3.7980, 3.8646, 4.0872, 4.0487, 3.8262], + device='cuda:0'), covar=tensor([0.0047, 0.0065, 0.0037, 0.0049, 0.0055, 0.0043, 0.0049, 0.0063], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0033, 0.0032, 0.0029, 0.0030, 0.0031, 0.0039, 0.0036], + device='cuda:0'), out_proj_covar=tensor([8.3192e-05, 1.0432e-04, 1.1067e-04, 8.9151e-05, 9.2612e-05, 9.9447e-05, + 1.2865e-04, 1.1439e-04], device='cuda:0') +2023-03-20 19:54:36,501 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=17761.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:54:38,312 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-20 19:54:51,911 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=17791.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:54:53,739 INFO [train.py:901] (0/2) Epoch 7, batch 850, loss[loss=0.2066, simple_loss=0.2701, pruned_loss=0.07153, over 7330.00 frames. ], tot_loss[loss=0.2137, simple_loss=0.2736, pruned_loss=0.07684, over 1418475.23 frames. ], batch size: 49, lr: 2.19e-02, grad_scale: 8.0 +2023-03-20 19:54:57,096 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-20 19:54:57,105 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-20 19:55:01,228 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.727e+02 3.034e+02 3.726e+02 4.781e+02 1.153e+03, threshold=7.452e+02, percent-clipped=1.0 +2023-03-20 19:55:03,218 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-20 19:55:06,746 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-20 19:55:17,841 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.78 vs. limit=5.0 +2023-03-20 19:55:19,515 INFO [train.py:901] (0/2) Epoch 7, batch 900, loss[loss=0.2206, simple_loss=0.2824, pruned_loss=0.07941, over 7302.00 frames. ], tot_loss[loss=0.2137, simple_loss=0.2733, pruned_loss=0.077, over 1422025.44 frames. ], batch size: 83, lr: 2.18e-02, grad_scale: 8.0 +2023-03-20 19:55:19,614 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=17845.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:55:28,861 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.52 vs. limit=5.0 +2023-03-20 19:55:44,938 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-20 19:55:44,970 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=17893.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:55:45,954 INFO [train.py:901] (0/2) Epoch 7, batch 950, loss[loss=0.2114, simple_loss=0.2698, pruned_loss=0.07644, over 7332.00 frames. ], tot_loss[loss=0.2137, simple_loss=0.2738, pruned_loss=0.0768, over 1427992.00 frames. ], batch size: 49, lr: 2.18e-02, grad_scale: 8.0 +2023-03-20 19:55:52,442 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.338e+02 3.363e+02 3.920e+02 4.922e+02 1.038e+03, threshold=7.839e+02, percent-clipped=2.0 +2023-03-20 19:55:59,839 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.90 vs. limit=2.0 +2023-03-20 19:56:08,896 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-20 19:56:10,873 INFO [train.py:901] (0/2) Epoch 7, batch 1000, loss[loss=0.1758, simple_loss=0.242, pruned_loss=0.05478, over 7144.00 frames. ], tot_loss[loss=0.2127, simple_loss=0.2729, pruned_loss=0.07627, over 1431440.95 frames. ], batch size: 41, lr: 2.18e-02, grad_scale: 8.0 +2023-03-20 19:56:19,235 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=17960.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:56:27,745 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0332, 0.8621, 1.1592, 1.2777, 0.8891, 1.1731, 0.9902, 1.2111], + device='cuda:0'), covar=tensor([0.0991, 0.1583, 0.0508, 0.0438, 0.2182, 0.0996, 0.0431, 0.1216], + device='cuda:0'), in_proj_covar=tensor([0.0026, 0.0033, 0.0029, 0.0028, 0.0032, 0.0028, 0.0031, 0.0030], + device='cuda:0'), out_proj_covar=tensor([4.9505e-05, 6.8494e-05, 4.8335e-05, 4.7305e-05, 5.8089e-05, 5.3728e-05, + 5.7282e-05, 5.7195e-05], device='cuda:0') +2023-03-20 19:56:27,924 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.05 vs. limit=2.0 +2023-03-20 19:56:28,630 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-20 19:56:37,048 INFO [train.py:901] (0/2) Epoch 7, batch 1050, loss[loss=0.2453, simple_loss=0.3023, pruned_loss=0.09421, over 7290.00 frames. ], tot_loss[loss=0.2128, simple_loss=0.2731, pruned_loss=0.07622, over 1435668.58 frames. ], batch size: 68, lr: 2.18e-02, grad_scale: 8.0 +2023-03-20 19:56:38,647 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3602, 2.3815, 1.9185, 1.9820, 2.2792, 2.1193, 2.4780, 1.9421], + device='cuda:0'), covar=tensor([0.0610, 0.0420, 0.0663, 0.0651, 0.1128, 0.0438, 0.0671, 0.1077], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0033, 0.0036, 0.0035, 0.0033, 0.0035, 0.0034, 0.0035], + device='cuda:0'), out_proj_covar=tensor([1.0422e-04, 8.9422e-05, 9.1178e-05, 8.9383e-05, 8.9933e-05, 9.1199e-05, + 8.9028e-05, 9.0720e-05], device='cuda:0') +2023-03-20 19:56:42,166 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.64 vs. limit=2.0 +2023-03-20 19:56:43,871 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.919e+02 3.096e+02 3.604e+02 4.422e+02 1.016e+03, threshold=7.208e+02, percent-clipped=3.0 +2023-03-20 19:56:43,987 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7481, 3.9436, 3.6143, 3.9156, 3.5212, 3.9719, 4.1724, 4.1031], + device='cuda:0'), covar=tensor([0.0255, 0.0181, 0.0298, 0.0218, 0.0443, 0.0200, 0.0290, 0.0250], + device='cuda:0'), in_proj_covar=tensor([0.0093, 0.0095, 0.0089, 0.0099, 0.0095, 0.0077, 0.0076, 0.0078], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 19:56:49,037 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=18018.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:56:50,593 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=18021.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:56:51,972 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-20 19:56:55,510 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-20 19:56:56,642 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8342, 2.0184, 1.8094, 3.2233, 1.5590, 2.6888, 1.0534, 3.0161], + device='cuda:0'), covar=tensor([0.0040, 0.0805, 0.1739, 0.0038, 0.3844, 0.0065, 0.0953, 0.0085], + device='cuda:0'), in_proj_covar=tensor([0.0134, 0.0247, 0.0321, 0.0126, 0.0305, 0.0140, 0.0263, 0.0156], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0003, 0.0001, 0.0003, 0.0001, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 19:56:59,623 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=18039.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 19:57:03,128 INFO [train.py:901] (0/2) Epoch 7, batch 1100, loss[loss=0.2435, simple_loss=0.2995, pruned_loss=0.0938, over 7340.00 frames. ], tot_loss[loss=0.2129, simple_loss=0.2731, pruned_loss=0.07634, over 1436263.77 frames. ], batch size: 54, lr: 2.17e-02, grad_scale: 8.0 +2023-03-20 19:57:08,652 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=18056.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:57:13,601 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=18066.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:57:24,263 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-20 19:57:24,324 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=18086.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:57:24,786 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 19:57:24,833 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=18087.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 19:57:28,657 INFO [train.py:901] (0/2) Epoch 7, batch 1150, loss[loss=0.2128, simple_loss=0.279, pruned_loss=0.07334, over 7195.00 frames. ], tot_loss[loss=0.2134, simple_loss=0.2738, pruned_loss=0.07649, over 1439584.07 frames. ], batch size: 50, lr: 2.17e-02, grad_scale: 8.0 +2023-03-20 19:57:35,150 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.225e+02 2.927e+02 3.944e+02 4.726e+02 8.317e+02, threshold=7.887e+02, percent-clipped=4.0 +2023-03-20 19:57:36,645 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-20 19:57:37,159 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-20 19:57:39,942 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.36 vs. limit=5.0 +2023-03-20 19:57:54,428 INFO [train.py:901] (0/2) Epoch 7, batch 1200, loss[loss=0.2123, simple_loss=0.2705, pruned_loss=0.07705, over 7236.00 frames. ], tot_loss[loss=0.2128, simple_loss=0.2733, pruned_loss=0.07614, over 1439097.45 frames. ], batch size: 45, lr: 2.17e-02, grad_scale: 8.0 +2023-03-20 19:58:01,772 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.9771, 0.7902, 0.9968, 1.1689, 0.8099, 1.0481, 1.1272, 1.1377], + device='cuda:0'), covar=tensor([0.0653, 0.0773, 0.0197, 0.0398, 0.0743, 0.0656, 0.0456, 0.0683], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0032, 0.0029, 0.0030, 0.0033, 0.0029, 0.0033, 0.0032], + device='cuda:0'), out_proj_covar=tensor([5.1758e-05, 6.8594e-05, 4.9616e-05, 5.1064e-05, 6.0080e-05, 5.6135e-05, + 5.9870e-05, 6.0317e-05], device='cuda:0') +2023-03-20 19:58:09,728 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=18174.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:58:10,238 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=18175.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:58:11,114 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-20 19:58:11,747 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5206, 3.6338, 2.3554, 3.4674, 3.0609, 3.4352, 2.0182, 1.9643], + device='cuda:0'), covar=tensor([0.0054, 0.0264, 0.0787, 0.0235, 0.0082, 0.0146, 0.1167, 0.0820], + device='cuda:0'), in_proj_covar=tensor([0.0129, 0.0169, 0.0282, 0.0151, 0.0168, 0.0152, 0.0257, 0.0260], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 19:58:20,169 INFO [train.py:901] (0/2) Epoch 7, batch 1250, loss[loss=0.2365, simple_loss=0.2942, pruned_loss=0.08933, over 7302.00 frames. ], tot_loss[loss=0.2127, simple_loss=0.2735, pruned_loss=0.07594, over 1441039.40 frames. ], batch size: 68, lr: 2.16e-02, grad_scale: 8.0 +2023-03-20 19:58:23,282 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.09 vs. limit=2.0 +2023-03-20 19:58:27,031 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.616e+02 3.110e+02 3.917e+02 4.843e+02 1.048e+03, threshold=7.833e+02, percent-clipped=3.0 +2023-03-20 19:58:35,830 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-20 19:58:38,049 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.57 vs. limit=5.0 +2023-03-20 19:58:40,302 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-20 19:58:41,301 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-20 19:58:41,436 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=18235.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:58:41,919 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=18236.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:58:42,462 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9379, 3.7370, 3.1900, 3.6383, 2.9100, 2.5967, 3.8530, 3.0968], + device='cuda:0'), covar=tensor([0.0083, 0.0085, 0.0146, 0.0059, 0.0195, 0.0282, 0.0104, 0.0314], + device='cuda:0'), in_proj_covar=tensor([0.0217, 0.0205, 0.0223, 0.0175, 0.0278, 0.0283, 0.0217, 0.0289], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-20 19:58:43,925 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2001, 2.9971, 2.6451, 2.9775, 2.9924, 2.6312, 2.4130, 2.9641], + device='cuda:0'), covar=tensor([0.1162, 0.0258, 0.1790, 0.1860, 0.1026, 0.1470, 0.2727, 0.1805], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0023, 0.0028, 0.0028, 0.0025, 0.0027, 0.0036, 0.0027], + device='cuda:0'), out_proj_covar=tensor([8.5914e-05, 7.3562e-05, 8.7554e-05, 8.9929e-05, 8.3238e-05, 8.8476e-05, + 1.0599e-04, 8.7289e-05], device='cuda:0') +2023-03-20 19:58:46,313 INFO [train.py:901] (0/2) Epoch 7, batch 1300, loss[loss=0.2375, simple_loss=0.2947, pruned_loss=0.09013, over 7361.00 frames. ], tot_loss[loss=0.211, simple_loss=0.2718, pruned_loss=0.0751, over 1440716.28 frames. ], batch size: 63, lr: 2.16e-02, grad_scale: 8.0 +2023-03-20 19:58:59,585 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.59 vs. limit=5.0 +2023-03-20 19:59:04,329 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-20 19:59:06,320 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-20 19:59:09,724 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-20 19:59:11,660 INFO [train.py:901] (0/2) Epoch 7, batch 1350, loss[loss=0.222, simple_loss=0.2888, pruned_loss=0.07756, over 7340.00 frames. ], tot_loss[loss=0.2124, simple_loss=0.2731, pruned_loss=0.07582, over 1443379.26 frames. ], batch size: 73, lr: 2.16e-02, grad_scale: 8.0 +2023-03-20 19:59:18,812 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.866e+02 3.114e+02 3.862e+02 4.620e+02 8.762e+02, threshold=7.723e+02, percent-clipped=2.0 +2023-03-20 19:59:20,349 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-20 19:59:22,912 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=18316.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:59:36,042 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9589, 1.7935, 1.7550, 3.1278, 1.4420, 2.7011, 1.0165, 2.7266], + device='cuda:0'), covar=tensor([0.0065, 0.0778, 0.1738, 0.0046, 0.3949, 0.0053, 0.1061, 0.0071], + device='cuda:0'), in_proj_covar=tensor([0.0139, 0.0255, 0.0330, 0.0128, 0.0314, 0.0142, 0.0271, 0.0160], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0003, 0.0001, 0.0003, 0.0001, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 19:59:37,880 INFO [train.py:901] (0/2) Epoch 7, batch 1400, loss[loss=0.1699, simple_loss=0.2371, pruned_loss=0.05139, over 7186.00 frames. ], tot_loss[loss=0.2124, simple_loss=0.2732, pruned_loss=0.07578, over 1441958.53 frames. ], batch size: 39, lr: 2.16e-02, grad_scale: 8.0 +2023-03-20 19:59:43,580 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=18356.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 19:59:51,590 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-20 19:59:51,721 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.9076, 0.8307, 1.2287, 1.1460, 1.0290, 1.2072, 1.0731, 1.1925], + device='cuda:0'), covar=tensor([0.0759, 0.2025, 0.0726, 0.0345, 0.0862, 0.0780, 0.0424, 0.0700], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0034, 0.0031, 0.0029, 0.0031, 0.0030, 0.0032, 0.0031], + device='cuda:0'), out_proj_covar=tensor([5.2563e-05, 7.0556e-05, 5.1456e-05, 5.1239e-05, 5.8802e-05, 5.6939e-05, + 5.9665e-05, 5.9499e-05], device='cuda:0') +2023-03-20 19:59:58,729 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=18386.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:00:03,669 INFO [train.py:901] (0/2) Epoch 7, batch 1450, loss[loss=0.1973, simple_loss=0.2671, pruned_loss=0.06377, over 7260.00 frames. ], tot_loss[loss=0.2121, simple_loss=0.2732, pruned_loss=0.07553, over 1442114.29 frames. ], batch size: 47, lr: 2.15e-02, grad_scale: 8.0 +2023-03-20 20:00:08,532 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=18404.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:00:10,481 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.072e+02 3.218e+02 3.702e+02 4.494e+02 7.594e+02, threshold=7.404e+02, percent-clipped=0.0 +2023-03-20 20:00:16,816 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-20 20:00:23,959 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4906, 3.6703, 2.3493, 3.6225, 3.0381, 3.4238, 2.1063, 1.9476], + device='cuda:0'), covar=tensor([0.0061, 0.0145, 0.0654, 0.0162, 0.0062, 0.0084, 0.0994, 0.0696], + device='cuda:0'), in_proj_covar=tensor([0.0131, 0.0166, 0.0279, 0.0151, 0.0166, 0.0155, 0.0259, 0.0263], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 20:00:24,324 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=18434.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:00:29,767 INFO [train.py:901] (0/2) Epoch 7, batch 1500, loss[loss=0.2741, simple_loss=0.3197, pruned_loss=0.1142, over 6647.00 frames. ], tot_loss[loss=0.213, simple_loss=0.274, pruned_loss=0.07597, over 1442115.11 frames. ], batch size: 106, lr: 2.15e-02, grad_scale: 8.0 +2023-03-20 20:00:31,838 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-20 20:00:53,297 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=18490.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:00:55,599 INFO [train.py:901] (0/2) Epoch 7, batch 1550, loss[loss=0.2019, simple_loss=0.2691, pruned_loss=0.06738, over 7360.00 frames. ], tot_loss[loss=0.2129, simple_loss=0.2742, pruned_loss=0.07586, over 1441888.99 frames. ], batch size: 63, lr: 2.15e-02, grad_scale: 8.0 +2023-03-20 20:00:56,641 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-20 20:01:02,705 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.878e+02 3.000e+02 3.800e+02 4.612e+02 7.440e+02, threshold=7.601e+02, percent-clipped=1.0 +2023-03-20 20:01:02,856 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=18508.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 20:01:13,793 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=18530.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:01:14,310 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=18531.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:01:14,651 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.22 vs. limit=5.0 +2023-03-20 20:01:15,175 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.92 vs. limit=2.0 +2023-03-20 20:01:18,944 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9918, 2.1190, 2.1170, 2.0656, 2.1916, 1.9950, 2.2334, 2.1976], + device='cuda:0'), covar=tensor([0.1733, 0.0501, 0.0539, 0.0518, 0.0779, 0.0427, 0.0608, 0.0798], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0032, 0.0035, 0.0034, 0.0033, 0.0033, 0.0031, 0.0033], + device='cuda:0'), out_proj_covar=tensor([1.0501e-04, 8.8467e-05, 9.1559e-05, 9.0254e-05, 9.0567e-05, 8.8345e-05, + 8.4484e-05, 8.7044e-05], device='cuda:0') +2023-03-20 20:01:21,257 INFO [train.py:901] (0/2) Epoch 7, batch 1600, loss[loss=0.211, simple_loss=0.2646, pruned_loss=0.07875, over 7321.00 frames. ], tot_loss[loss=0.2132, simple_loss=0.2741, pruned_loss=0.07618, over 1441988.13 frames. ], batch size: 61, lr: 2.15e-02, grad_scale: 8.0 +2023-03-20 20:01:21,752 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.1231, 4.8950, 4.8863, 5.1991, 5.2747, 5.2533, 4.7363, 4.7501], + device='cuda:0'), covar=tensor([0.0621, 0.1471, 0.1454, 0.0989, 0.0438, 0.0845, 0.0524, 0.0633], + device='cuda:0'), in_proj_covar=tensor([0.0139, 0.0217, 0.0193, 0.0176, 0.0147, 0.0230, 0.0131, 0.0152], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:0') +2023-03-20 20:01:24,340 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=18551.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:01:27,667 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-20 20:01:28,682 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-20 20:01:32,290 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-20 20:01:33,885 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=18569.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 20:01:41,140 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-20 20:01:45,082 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-20 20:01:46,356 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.14 vs. limit=2.0 +2023-03-20 20:01:46,556 INFO [train.py:901] (0/2) Epoch 7, batch 1650, loss[loss=0.2137, simple_loss=0.2767, pruned_loss=0.07534, over 7323.00 frames. ], tot_loss[loss=0.2129, simple_loss=0.2737, pruned_loss=0.0761, over 1442811.57 frames. ], batch size: 61, lr: 2.14e-02, grad_scale: 8.0 +2023-03-20 20:01:54,021 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.964e+02 3.211e+02 3.996e+02 4.881e+02 1.178e+03, threshold=7.992e+02, percent-clipped=2.0 +2023-03-20 20:01:54,043 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-20 20:01:58,126 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=18616.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:02:10,310 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 20:02:12,339 INFO [train.py:901] (0/2) Epoch 7, batch 1700, loss[loss=0.2076, simple_loss=0.2711, pruned_loss=0.07209, over 7263.00 frames. ], tot_loss[loss=0.2122, simple_loss=0.2732, pruned_loss=0.07554, over 1440681.10 frames. ], batch size: 77, lr: 2.14e-02, grad_scale: 8.0 +2023-03-20 20:02:14,422 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-20 20:02:22,610 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=18664.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:02:25,663 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-20 20:02:27,800 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=18674.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:02:37,901 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0260, 1.0897, 1.1629, 0.7778, 0.9352, 0.8489, 1.0124, 0.7441], + device='cuda:0'), covar=tensor([0.0139, 0.0080, 0.0123, 0.0084, 0.0159, 0.0100, 0.0179, 0.0243], + device='cuda:0'), in_proj_covar=tensor([0.0017, 0.0016, 0.0016, 0.0015, 0.0018, 0.0016, 0.0016, 0.0020], + device='cuda:0'), out_proj_covar=tensor([2.0985e-05, 1.8431e-05, 2.1538e-05, 1.7345e-05, 2.0393e-05, 1.9254e-05, + 2.0664e-05, 2.8249e-05], device='cuda:0') +2023-03-20 20:02:38,759 INFO [train.py:901] (0/2) Epoch 7, batch 1750, loss[loss=0.2152, simple_loss=0.2739, pruned_loss=0.07829, over 7363.00 frames. ], tot_loss[loss=0.2129, simple_loss=0.2736, pruned_loss=0.07605, over 1440427.39 frames. ], batch size: 73, lr: 2.14e-02, grad_scale: 8.0 +2023-03-20 20:02:45,509 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.854e+02 2.856e+02 3.595e+02 4.544e+02 1.273e+03, threshold=7.190e+02, percent-clipped=1.0 +2023-03-20 20:02:47,310 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.93 vs. limit=2.0 +2023-03-20 20:02:50,981 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-20 20:02:51,993 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-20 20:02:54,855 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.52 vs. limit=5.0 +2023-03-20 20:02:59,056 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=18735.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:03:00,580 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2162, 4.0471, 3.5783, 3.3516, 4.1243, 2.2793, 1.7321, 4.1536], + device='cuda:0'), covar=tensor([0.0013, 0.0013, 0.0081, 0.0073, 0.0020, 0.0385, 0.0606, 0.0035], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0049, 0.0073, 0.0058, 0.0063, 0.0091, 0.0099, 0.0062], + device='cuda:0'), out_proj_covar=tensor([6.8853e-05, 7.8195e-05, 1.0529e-04, 8.5676e-05, 8.4825e-05, 1.3225e-04, + 1.4582e-04, 8.6294e-05], device='cuda:0') +2023-03-20 20:03:04,580 INFO [train.py:901] (0/2) Epoch 7, batch 1800, loss[loss=0.2335, simple_loss=0.2969, pruned_loss=0.08505, over 7312.00 frames. ], tot_loss[loss=0.2128, simple_loss=0.2737, pruned_loss=0.07595, over 1440405.57 frames. ], batch size: 59, lr: 2.13e-02, grad_scale: 8.0 +2023-03-20 20:03:10,620 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2444, 4.1960, 3.7613, 3.4267, 4.1417, 2.4372, 1.7080, 4.2265], + device='cuda:0'), covar=tensor([0.0013, 0.0013, 0.0060, 0.0060, 0.0019, 0.0383, 0.0624, 0.0036], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0050, 0.0074, 0.0058, 0.0063, 0.0091, 0.0100, 0.0062], + device='cuda:0'), out_proj_covar=tensor([6.8870e-05, 7.8864e-05, 1.0557e-04, 8.6328e-05, 8.4587e-05, 1.3259e-04, + 1.4645e-04, 8.6527e-05], device='cuda:0') +2023-03-20 20:03:15,019 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-20 20:03:20,924 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.91 vs. limit=2.0 +2023-03-20 20:03:21,285 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9219, 4.1054, 3.8248, 3.8434, 3.6349, 4.0379, 4.3275, 4.3610], + device='cuda:0'), covar=tensor([0.0205, 0.0148, 0.0198, 0.0220, 0.0422, 0.0206, 0.0283, 0.0209], + device='cuda:0'), in_proj_covar=tensor([0.0090, 0.0091, 0.0085, 0.0095, 0.0093, 0.0073, 0.0074, 0.0074], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 20:03:29,202 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-20 20:03:30,211 INFO [train.py:901] (0/2) Epoch 7, batch 1850, loss[loss=0.2107, simple_loss=0.2682, pruned_loss=0.07663, over 7251.00 frames. ], tot_loss[loss=0.212, simple_loss=0.2731, pruned_loss=0.07543, over 1441968.82 frames. ], batch size: 52, lr: 2.13e-02, grad_scale: 8.0 +2023-03-20 20:03:37,095 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.910e+02 3.015e+02 3.933e+02 4.902e+02 8.513e+02, threshold=7.867e+02, percent-clipped=4.0 +2023-03-20 20:03:37,790 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0999, 1.0976, 1.1807, 1.3745, 1.1987, 1.4488, 1.3196, 1.5469], + device='cuda:0'), covar=tensor([0.1154, 0.0889, 0.0514, 0.0936, 0.2663, 0.0753, 0.0455, 0.1502], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0034, 0.0031, 0.0030, 0.0030, 0.0029, 0.0034, 0.0031], + device='cuda:0'), out_proj_covar=tensor([5.2869e-05, 7.1109e-05, 5.2777e-05, 5.2577e-05, 5.7968e-05, 5.6207e-05, + 6.3105e-05, 5.9735e-05], device='cuda:0') +2023-03-20 20:03:39,175 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-20 20:03:48,825 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=18830.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:03:49,308 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=18831.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:03:54,302 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4764, 3.3410, 3.3588, 3.4972, 3.2510, 3.4351, 3.1352, 3.0568], + device='cuda:0'), covar=tensor([0.0064, 0.0123, 0.0079, 0.0060, 0.0084, 0.0084, 0.0119, 0.0117], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0033, 0.0031, 0.0029, 0.0030, 0.0031, 0.0038, 0.0036], + device='cuda:0'), out_proj_covar=tensor([7.9505e-05, 1.0761e-04, 1.0500e-04, 8.6694e-05, 9.4909e-05, 9.6973e-05, + 1.2501e-04, 1.1280e-04], device='cuda:0') +2023-03-20 20:03:54,564 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.10 vs. limit=2.0 +2023-03-20 20:03:55,666 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-20 20:03:56,141 INFO [train.py:901] (0/2) Epoch 7, batch 1900, loss[loss=0.2158, simple_loss=0.2749, pruned_loss=0.0784, over 7268.00 frames. ], tot_loss[loss=0.2113, simple_loss=0.2725, pruned_loss=0.07505, over 1443637.94 frames. ], batch size: 64, lr: 2.13e-02, grad_scale: 8.0 +2023-03-20 20:03:56,743 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=18846.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:03:59,741 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3753, 4.4366, 4.4369, 4.7973, 4.8102, 4.7647, 3.9564, 4.2396], + device='cuda:0'), covar=tensor([0.0736, 0.1574, 0.1830, 0.1177, 0.0528, 0.1218, 0.0681, 0.0972], + device='cuda:0'), in_proj_covar=tensor([0.0137, 0.0216, 0.0195, 0.0181, 0.0142, 0.0231, 0.0131, 0.0154], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0001, 0.0002], + device='cuda:0') +2023-03-20 20:04:06,314 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=18864.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 20:04:07,518 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.86 vs. limit=2.0 +2023-03-20 20:04:11,091 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.05 vs. limit=2.0 +2023-03-20 20:04:13,354 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=18878.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:04:13,820 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=18879.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:04:19,811 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-20 20:04:21,786 INFO [train.py:901] (0/2) Epoch 7, batch 1950, loss[loss=0.2188, simple_loss=0.2864, pruned_loss=0.0756, over 7289.00 frames. ], tot_loss[loss=0.2108, simple_loss=0.2721, pruned_loss=0.07473, over 1444935.75 frames. ], batch size: 86, lr: 2.13e-02, grad_scale: 8.0 +2023-03-20 20:04:28,239 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.263e+02 3.431e+02 3.888e+02 4.453e+02 9.116e+02, threshold=7.775e+02, percent-clipped=1.0 +2023-03-20 20:04:30,814 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-20 20:04:33,516 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2204, 1.5549, 2.5561, 1.7653, 1.1917, 1.1732, 1.7548, 1.8769], + device='cuda:0'), covar=tensor([0.0361, 0.0107, 0.0079, 0.0065, 0.0419, 0.0373, 0.0175, 0.0149], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0018, 0.0018, 0.0018, 0.0018, 0.0017, 0.0018, 0.0018], + device='cuda:0'), out_proj_covar=tensor([4.5702e-05, 4.0165e-05, 3.8925e-05, 3.5226e-05, 4.3617e-05, 3.9419e-05, + 4.1451e-05, 4.4099e-05], device='cuda:0') +2023-03-20 20:04:35,423 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-20 20:04:35,934 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-20 20:04:40,013 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=18930.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:04:47,374 INFO [train.py:901] (0/2) Epoch 7, batch 2000, loss[loss=0.1549, simple_loss=0.2146, pruned_loss=0.04763, over 6968.00 frames. ], tot_loss[loss=0.2107, simple_loss=0.272, pruned_loss=0.07469, over 1442026.19 frames. ], batch size: 35, lr: 2.12e-02, grad_scale: 8.0 +2023-03-20 20:04:53,126 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-20 20:04:57,788 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7438, 3.9513, 3.6058, 3.7035, 3.6674, 3.7545, 4.1491, 4.1102], + device='cuda:0'), covar=tensor([0.0207, 0.0162, 0.0202, 0.0199, 0.0380, 0.0209, 0.0237, 0.0226], + device='cuda:0'), in_proj_covar=tensor([0.0088, 0.0089, 0.0083, 0.0093, 0.0091, 0.0072, 0.0072, 0.0072], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 20:05:04,225 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-20 20:05:11,468 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=18991.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:05:12,800 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-20 20:05:13,299 INFO [train.py:901] (0/2) Epoch 7, batch 2050, loss[loss=0.252, simple_loss=0.317, pruned_loss=0.0935, over 6692.00 frames. ], tot_loss[loss=0.2102, simple_loss=0.2716, pruned_loss=0.07442, over 1440076.99 frames. ], batch size: 106, lr: 2.12e-02, grad_scale: 8.0 +2023-03-20 20:05:20,081 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.785e+02 3.151e+02 3.634e+02 4.204e+02 9.564e+02, threshold=7.267e+02, percent-clipped=4.0 +2023-03-20 20:05:31,882 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=19030.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:05:39,803 INFO [train.py:901] (0/2) Epoch 7, batch 2100, loss[loss=0.2074, simple_loss=0.2711, pruned_loss=0.07187, over 7319.00 frames. ], tot_loss[loss=0.21, simple_loss=0.2711, pruned_loss=0.07445, over 1438184.04 frames. ], batch size: 61, lr: 2.12e-02, grad_scale: 8.0 +2023-03-20 20:05:40,647 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.01 vs. limit=2.0 +2023-03-20 20:05:46,905 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-20 20:05:49,907 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-20 20:05:50,199 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-20 20:06:05,015 INFO [train.py:901] (0/2) Epoch 7, batch 2150, loss[loss=0.2068, simple_loss=0.2539, pruned_loss=0.07987, over 6975.00 frames. ], tot_loss[loss=0.2104, simple_loss=0.2716, pruned_loss=0.07465, over 1438795.41 frames. ], batch size: 35, lr: 2.12e-02, grad_scale: 8.0 +2023-03-20 20:06:12,198 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.027e+02 3.142e+02 3.714e+02 4.816e+02 1.137e+03, threshold=7.428e+02, percent-clipped=3.0 +2023-03-20 20:06:23,645 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=19129.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 20:06:26,704 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.01 vs. limit=5.0 +2023-03-20 20:06:31,460 INFO [train.py:901] (0/2) Epoch 7, batch 2200, loss[loss=0.1647, simple_loss=0.23, pruned_loss=0.04969, over 7135.00 frames. ], tot_loss[loss=0.2108, simple_loss=0.2721, pruned_loss=0.07475, over 1440906.06 frames. ], batch size: 41, lr: 2.11e-02, grad_scale: 8.0 +2023-03-20 20:06:32,073 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=19146.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:06:36,559 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-20 20:06:41,197 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=19164.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 20:06:54,914 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=19190.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 20:06:56,850 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=19194.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:06:57,305 INFO [train.py:901] (0/2) Epoch 7, batch 2250, loss[loss=0.2135, simple_loss=0.2728, pruned_loss=0.07715, over 7283.00 frames. ], tot_loss[loss=0.2108, simple_loss=0.2721, pruned_loss=0.07476, over 1439017.10 frames. ], batch size: 70, lr: 2.11e-02, grad_scale: 8.0 +2023-03-20 20:06:57,735 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.87 vs. limit=2.0 +2023-03-20 20:07:02,335 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2635, 2.2040, 1.8745, 1.7581, 2.0315, 1.8794, 2.2911, 2.0011], + device='cuda:0'), covar=tensor([0.0471, 0.0399, 0.0816, 0.1106, 0.1313, 0.0576, 0.0636, 0.1088], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0033, 0.0038, 0.0036, 0.0033, 0.0036, 0.0031, 0.0035], + device='cuda:0'), out_proj_covar=tensor([1.0902e-04, 9.2473e-05, 1.0061e-04, 9.6352e-05, 9.3317e-05, 9.5107e-05, + 8.7488e-05, 9.4410e-05], device='cuda:0') +2023-03-20 20:07:04,214 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.690e+02 2.891e+02 3.634e+02 4.674e+02 1.266e+03, threshold=7.268e+02, percent-clipped=2.0 +2023-03-20 20:07:06,339 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=19212.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 20:07:11,015 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-20 20:07:11,503 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-20 20:07:20,163 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4374, 2.3639, 2.1248, 3.3694, 1.5379, 3.2414, 1.0024, 3.4273], + device='cuda:0'), covar=tensor([0.0075, 0.0612, 0.1600, 0.0046, 0.4452, 0.0064, 0.1120, 0.0059], + device='cuda:0'), in_proj_covar=tensor([0.0136, 0.0254, 0.0322, 0.0123, 0.0304, 0.0138, 0.0268, 0.0155], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0003, 0.0001, 0.0003, 0.0001, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 20:07:23,530 INFO [train.py:901] (0/2) Epoch 7, batch 2300, loss[loss=0.199, simple_loss=0.2688, pruned_loss=0.06464, over 7276.00 frames. ], tot_loss[loss=0.2106, simple_loss=0.2721, pruned_loss=0.07457, over 1438327.56 frames. ], batch size: 77, lr: 2.11e-02, grad_scale: 8.0 +2023-03-20 20:07:23,552 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-20 20:07:34,257 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3386, 4.4037, 3.9493, 4.1600, 4.0090, 4.2951, 4.5761, 4.6038], + device='cuda:0'), covar=tensor([0.0113, 0.0118, 0.0171, 0.0148, 0.0285, 0.0129, 0.0192, 0.0169], + device='cuda:0'), in_proj_covar=tensor([0.0088, 0.0091, 0.0082, 0.0095, 0.0092, 0.0072, 0.0073, 0.0072], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 20:07:44,929 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=19286.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:07:49,332 INFO [train.py:901] (0/2) Epoch 7, batch 2350, loss[loss=0.2319, simple_loss=0.2854, pruned_loss=0.08916, over 7312.00 frames. ], tot_loss[loss=0.2107, simple_loss=0.2723, pruned_loss=0.0746, over 1439693.05 frames. ], batch size: 49, lr: 2.11e-02, grad_scale: 8.0 +2023-03-20 20:07:56,473 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.976e+02 3.198e+02 3.776e+02 4.552e+02 1.396e+03, threshold=7.552e+02, percent-clipped=7.0 +2023-03-20 20:08:01,062 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7877, 4.7540, 4.6848, 5.1340, 5.1302, 5.1867, 4.4121, 4.6074], + device='cuda:0'), covar=tensor([0.0826, 0.1682, 0.1915, 0.0898, 0.0585, 0.0989, 0.0713, 0.0909], + device='cuda:0'), in_proj_covar=tensor([0.0140, 0.0221, 0.0198, 0.0184, 0.0144, 0.0239, 0.0131, 0.0158], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002], + device='cuda:0') +2023-03-20 20:08:07,673 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=19330.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:08:09,624 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-20 20:08:15,102 INFO [train.py:901] (0/2) Epoch 7, batch 2400, loss[loss=0.2053, simple_loss=0.2642, pruned_loss=0.07321, over 7317.00 frames. ], tot_loss[loss=0.2104, simple_loss=0.2718, pruned_loss=0.07455, over 1438254.09 frames. ], batch size: 80, lr: 2.10e-02, grad_scale: 8.0 +2023-03-20 20:08:15,627 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-20 20:08:25,850 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.3390, 5.6197, 5.6544, 5.5180, 5.3908, 5.2171, 5.7019, 5.4974], + device='cuda:0'), covar=tensor([0.0275, 0.0281, 0.0304, 0.0432, 0.0258, 0.0253, 0.0239, 0.0331], + device='cuda:0'), in_proj_covar=tensor([0.0130, 0.0136, 0.0109, 0.0105, 0.0090, 0.0128, 0.0116, 0.0091], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 20:08:26,317 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-20 20:08:28,857 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-20 20:08:32,455 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=19378.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:08:41,191 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=19394.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:08:41,566 INFO [train.py:901] (0/2) Epoch 7, batch 2450, loss[loss=0.1904, simple_loss=0.2561, pruned_loss=0.06236, over 7356.00 frames. ], tot_loss[loss=0.2104, simple_loss=0.2721, pruned_loss=0.0744, over 1440070.46 frames. ], batch size: 44, lr: 2.10e-02, grad_scale: 8.0 +2023-03-20 20:08:48,358 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.646e+02 3.092e+02 3.904e+02 5.234e+02 1.110e+03, threshold=7.808e+02, percent-clipped=4.0 +2023-03-20 20:08:54,837 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-20 20:09:07,599 INFO [train.py:901] (0/2) Epoch 7, batch 2500, loss[loss=0.2499, simple_loss=0.3141, pruned_loss=0.09288, over 6784.00 frames. ], tot_loss[loss=0.2114, simple_loss=0.2729, pruned_loss=0.07488, over 1439716.90 frames. ], batch size: 106, lr: 2.10e-02, grad_scale: 8.0 +2023-03-20 20:09:12,815 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=19455.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:09:21,288 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-20 20:09:28,465 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=19485.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 20:09:33,350 INFO [train.py:901] (0/2) Epoch 7, batch 2550, loss[loss=0.1957, simple_loss=0.2649, pruned_loss=0.06327, over 7270.00 frames. ], tot_loss[loss=0.2112, simple_loss=0.2731, pruned_loss=0.07469, over 1441362.88 frames. ], batch size: 70, lr: 2.10e-02, grad_scale: 8.0 +2023-03-20 20:09:39,865 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.969e+02 3.253e+02 3.755e+02 4.518e+02 7.754e+02, threshold=7.511e+02, percent-clipped=0.0 +2023-03-20 20:09:50,284 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8464, 3.8076, 3.5843, 3.7427, 3.5944, 3.8526, 4.0431, 4.0403], + device='cuda:0'), covar=tensor([0.0159, 0.0149, 0.0225, 0.0182, 0.0325, 0.0153, 0.0328, 0.0241], + device='cuda:0'), in_proj_covar=tensor([0.0088, 0.0090, 0.0083, 0.0095, 0.0093, 0.0070, 0.0072, 0.0072], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 20:09:59,144 INFO [train.py:901] (0/2) Epoch 7, batch 2600, loss[loss=0.2214, simple_loss=0.28, pruned_loss=0.0814, over 7309.00 frames. ], tot_loss[loss=0.21, simple_loss=0.2721, pruned_loss=0.07397, over 1442077.53 frames. ], batch size: 49, lr: 2.09e-02, grad_scale: 16.0 +2023-03-20 20:10:00,238 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1733, 2.3110, 2.2021, 2.2203, 2.4200, 2.0312, 2.3188, 2.2597], + device='cuda:0'), covar=tensor([0.1886, 0.0973, 0.0507, 0.0698, 0.0519, 0.0655, 0.0853, 0.0720], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0033, 0.0037, 0.0035, 0.0034, 0.0037, 0.0033, 0.0035], + device='cuda:0'), out_proj_covar=tensor([1.1399e-04, 9.4708e-05, 1.0054e-04, 9.5937e-05, 9.6085e-05, 1.0038e-04, + 9.1791e-05, 9.4612e-05], device='cuda:0') +2023-03-20 20:10:19,609 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=19586.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:10:23,822 INFO [train.py:901] (0/2) Epoch 7, batch 2650, loss[loss=0.2058, simple_loss=0.2716, pruned_loss=0.06994, over 7318.00 frames. ], tot_loss[loss=0.2103, simple_loss=0.2726, pruned_loss=0.07395, over 1441203.83 frames. ], batch size: 83, lr: 2.09e-02, grad_scale: 16.0 +2023-03-20 20:10:31,138 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.231e+02 3.103e+02 3.938e+02 4.730e+02 1.209e+03, threshold=7.875e+02, percent-clipped=4.0 +2023-03-20 20:10:36,192 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=19618.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:10:44,138 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=19634.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:10:49,452 INFO [train.py:901] (0/2) Epoch 7, batch 2700, loss[loss=0.1993, simple_loss=0.2667, pruned_loss=0.06595, over 7245.00 frames. ], tot_loss[loss=0.2096, simple_loss=0.272, pruned_loss=0.07365, over 1441498.25 frames. ], batch size: 47, lr: 2.09e-02, grad_scale: 16.0 +2023-03-20 20:11:06,100 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=19679.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:11:08,692 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.99 vs. limit=2.0 +2023-03-20 20:11:13,900 INFO [train.py:901] (0/2) Epoch 7, batch 2750, loss[loss=0.2204, simple_loss=0.2885, pruned_loss=0.07608, over 7332.00 frames. ], tot_loss[loss=0.2087, simple_loss=0.2709, pruned_loss=0.07319, over 1441401.90 frames. ], batch size: 54, lr: 2.09e-02, grad_scale: 16.0 +2023-03-20 20:11:20,353 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.753e+02 2.910e+02 3.544e+02 4.508e+02 1.425e+03, threshold=7.087e+02, percent-clipped=5.0 +2023-03-20 20:11:29,828 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.91 vs. limit=2.0 +2023-03-20 20:11:38,345 INFO [train.py:901] (0/2) Epoch 7, batch 2800, loss[loss=0.1971, simple_loss=0.2475, pruned_loss=0.07332, over 7085.00 frames. ], tot_loss[loss=0.209, simple_loss=0.2713, pruned_loss=0.07334, over 1444282.89 frames. ], batch size: 35, lr: 2.08e-02, grad_scale: 16.0 +2023-03-20 20:11:40,812 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=19750.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:11:47,617 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.07 vs. limit=5.0 +2023-03-20 20:11:51,078 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-7.pt +2023-03-20 20:12:08,851 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-20 20:12:08,907 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-20 20:12:08,932 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-20 20:12:08,951 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-20 20:12:09,133 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-20 20:12:09,149 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-20 20:12:09,160 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-20 20:12:09,596 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-20 20:12:09,676 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-20 20:12:09,692 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-20 20:12:09,710 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-20 20:12:09,886 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-20 20:12:09,892 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-20 20:12:09,902 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-20 20:12:10,051 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-20 20:12:10,054 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-20 20:12:10,115 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-20 20:12:12,427 INFO [train.py:901] (0/2) Epoch 8, batch 0, loss[loss=0.2135, simple_loss=0.274, pruned_loss=0.07655, over 7286.00 frames. ], tot_loss[loss=0.2135, simple_loss=0.274, pruned_loss=0.07655, over 7286.00 frames. ], batch size: 70, lr: 2.00e-02, grad_scale: 16.0 +2023-03-20 20:12:12,428 INFO [train.py:926] (0/2) Computing validation loss +2023-03-20 20:12:24,905 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5887, 2.2194, 2.4531, 2.4026, 2.9006, 2.1091, 1.8847, 2.5161], + device='cuda:0'), covar=tensor([0.1508, 0.0775, 0.3020, 0.3119, 0.0843, 0.2961, 0.4957, 0.2257], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0026, 0.0030, 0.0029, 0.0027, 0.0029, 0.0037, 0.0028], + device='cuda:0'), out_proj_covar=tensor([9.6256e-05, 8.8191e-05, 1.0168e-04, 9.9615e-05, 9.4147e-05, 9.8971e-05, + 1.1571e-04, 9.9657e-05], device='cuda:0') +2023-03-20 20:12:24,916 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2568, 1.2918, 1.7832, 0.8576, 1.3501, 1.1871, 1.2338, 1.1590], + device='cuda:0'), covar=tensor([0.0398, 0.0233, 0.0142, 0.0095, 0.0341, 0.0320, 0.0155, 0.0334], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0019, 0.0018, 0.0018, 0.0020, 0.0018, 0.0019, 0.0019], + device='cuda:0'), out_proj_covar=tensor([4.6160e-05, 4.2350e-05, 4.0312e-05, 3.6544e-05, 4.5967e-05, 4.1084e-05, + 4.3107e-05, 4.6509e-05], device='cuda:0') +2023-03-20 20:12:38,316 INFO [train.py:935] (0/2) Epoch 8, validation: loss=0.1827, simple_loss=0.2687, pruned_loss=0.0484, over 1622729.00 frames. +2023-03-20 20:12:38,316 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-20 20:12:44,331 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-20 20:12:46,487 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=19785.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 20:12:54,672 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-20 20:12:58,033 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.946e+02 2.948e+02 3.559e+02 4.646e+02 7.849e+02, threshold=7.118e+02, percent-clipped=2.0 +2023-03-20 20:13:01,511 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-20 20:13:03,544 INFO [train.py:901] (0/2) Epoch 8, batch 50, loss[loss=0.1848, simple_loss=0.2582, pruned_loss=0.05563, over 7339.00 frames. ], tot_loss[loss=0.2137, simple_loss=0.275, pruned_loss=0.07618, over 325909.81 frames. ], batch size: 44, lr: 1.99e-02, grad_scale: 16.0 +2023-03-20 20:13:03,558 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-20 20:13:07,320 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-20 20:13:11,219 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=19833.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 20:13:21,097 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2461, 1.1819, 1.7221, 0.8614, 1.1706, 0.7748, 1.2676, 1.0906], + device='cuda:0'), covar=tensor([0.0274, 0.0186, 0.0085, 0.0068, 0.0379, 0.0407, 0.0172, 0.0283], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0018, 0.0018, 0.0019, 0.0019, 0.0018, 0.0020, 0.0020], + device='cuda:0'), out_proj_covar=tensor([4.5884e-05, 4.1108e-05, 3.9804e-05, 3.6647e-05, 4.4932e-05, 4.1851e-05, + 4.3504e-05, 4.7046e-05], device='cuda:0') +2023-03-20 20:13:24,056 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-20 20:13:24,579 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-20 20:13:25,347 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.98 vs. limit=2.0 +2023-03-20 20:13:29,644 INFO [train.py:901] (0/2) Epoch 8, batch 100, loss[loss=0.2002, simple_loss=0.2585, pruned_loss=0.07093, over 7252.00 frames. ], tot_loss[loss=0.2087, simple_loss=0.2707, pruned_loss=0.07333, over 572546.90 frames. ], batch size: 55, lr: 1.99e-02, grad_scale: 16.0 +2023-03-20 20:13:49,792 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.917e+02 2.926e+02 3.845e+02 4.661e+02 1.011e+03, threshold=7.690e+02, percent-clipped=6.0 +2023-03-20 20:13:55,387 INFO [train.py:901] (0/2) Epoch 8, batch 150, loss[loss=0.1897, simple_loss=0.2513, pruned_loss=0.06409, over 7172.00 frames. ], tot_loss[loss=0.2093, simple_loss=0.2713, pruned_loss=0.07364, over 766305.39 frames. ], batch size: 41, lr: 1.99e-02, grad_scale: 16.0 +2023-03-20 20:14:11,926 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7890, 3.6864, 3.2854, 3.5004, 2.8374, 2.4789, 3.6883, 3.0067], + device='cuda:0'), covar=tensor([0.0104, 0.0120, 0.0136, 0.0076, 0.0211, 0.0318, 0.0133, 0.0377], + device='cuda:0'), in_proj_covar=tensor([0.0222, 0.0212, 0.0223, 0.0199, 0.0283, 0.0286, 0.0229, 0.0292], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-20 20:14:16,142 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-03-20 20:14:21,439 INFO [train.py:901] (0/2) Epoch 8, batch 200, loss[loss=0.2891, simple_loss=0.3262, pruned_loss=0.126, over 6682.00 frames. ], tot_loss[loss=0.2088, simple_loss=0.2708, pruned_loss=0.07339, over 915358.59 frames. ], batch size: 106, lr: 1.99e-02, grad_scale: 8.0 +2023-03-20 20:14:24,080 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=19974.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:14:25,580 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-20 20:14:30,054 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-20 20:14:35,618 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8917, 3.9081, 3.2680, 3.3070, 3.6919, 2.3858, 1.5958, 3.9256], + device='cuda:0'), covar=tensor([0.0017, 0.0038, 0.0072, 0.0052, 0.0043, 0.0364, 0.0591, 0.0033], + device='cuda:0'), in_proj_covar=tensor([0.0056, 0.0050, 0.0073, 0.0058, 0.0064, 0.0092, 0.0102, 0.0061], + device='cuda:0'), out_proj_covar=tensor([7.1727e-05, 7.9982e-05, 1.0504e-04, 8.5996e-05, 8.5947e-05, 1.3429e-04, + 1.4808e-04, 8.4683e-05], device='cuda:0') +2023-03-20 20:14:37,101 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-20 20:14:37,955 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-20000.pt +2023-03-20 20:14:46,014 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.800e+02 3.019e+02 3.456e+02 4.343e+02 8.927e+02, threshold=6.913e+02, percent-clipped=3.0 +2023-03-20 20:14:51,096 INFO [train.py:901] (0/2) Epoch 8, batch 250, loss[loss=0.2323, simple_loss=0.2782, pruned_loss=0.09318, over 7327.00 frames. ], tot_loss[loss=0.2079, simple_loss=0.2703, pruned_loss=0.07275, over 1031458.83 frames. ], batch size: 75, lr: 1.99e-02, grad_scale: 8.0 +2023-03-20 20:14:51,803 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=20020.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:14:53,666 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-20 20:15:07,218 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=20050.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:15:13,775 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-20 20:15:14,904 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0724, 1.5502, 1.8560, 1.0754, 1.1820, 0.9892, 1.5237, 1.1234], + device='cuda:0'), covar=tensor([0.0230, 0.0134, 0.0072, 0.0044, 0.0285, 0.0331, 0.0083, 0.0178], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0019, 0.0018, 0.0019, 0.0019, 0.0018, 0.0019, 0.0020], + device='cuda:0'), out_proj_covar=tensor([4.5048e-05, 4.2992e-05, 3.9988e-05, 3.6220e-05, 4.3972e-05, 4.0540e-05, + 4.2040e-05, 4.7235e-05], device='cuda:0') +2023-03-20 20:15:16,851 INFO [train.py:901] (0/2) Epoch 8, batch 300, loss[loss=0.1952, simple_loss=0.2613, pruned_loss=0.06453, over 7294.00 frames. ], tot_loss[loss=0.2066, simple_loss=0.2696, pruned_loss=0.07179, over 1122979.78 frames. ], batch size: 68, lr: 1.98e-02, grad_scale: 8.0 +2023-03-20 20:15:19,314 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.79 vs. limit=2.0 +2023-03-20 20:15:22,664 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-20 20:15:23,812 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=20081.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:15:27,743 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=20089.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:15:32,206 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=20098.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:15:37,677 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.998e+02 3.160e+02 3.716e+02 4.730e+02 1.016e+03, threshold=7.433e+02, percent-clipped=7.0 +2023-03-20 20:15:39,319 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=20112.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:15:43,566 INFO [train.py:901] (0/2) Epoch 8, batch 350, loss[loss=0.2394, simple_loss=0.2979, pruned_loss=0.09038, over 7224.00 frames. ], tot_loss[loss=0.2078, simple_loss=0.2709, pruned_loss=0.07235, over 1194945.72 frames. ], batch size: 93, lr: 1.98e-02, grad_scale: 8.0 +2023-03-20 20:15:54,999 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-20 20:15:59,108 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=20150.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:16:03,656 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3752, 4.4050, 3.9354, 3.6431, 4.0975, 2.2912, 1.6090, 4.3855], + device='cuda:0'), covar=tensor([0.0017, 0.0014, 0.0059, 0.0056, 0.0026, 0.0447, 0.0629, 0.0029], + device='cuda:0'), in_proj_covar=tensor([0.0057, 0.0050, 0.0074, 0.0059, 0.0065, 0.0093, 0.0101, 0.0062], + device='cuda:0'), out_proj_covar=tensor([7.4371e-05, 7.9415e-05, 1.0702e-04, 8.7131e-05, 8.7362e-05, 1.3517e-04, + 1.4765e-04, 8.6124e-05], device='cuda:0') +2023-03-20 20:16:08,569 INFO [train.py:901] (0/2) Epoch 8, batch 400, loss[loss=0.2038, simple_loss=0.2782, pruned_loss=0.06474, over 7335.00 frames. ], tot_loss[loss=0.2066, simple_loss=0.2695, pruned_loss=0.07189, over 1248895.80 frames. ], batch size: 54, lr: 1.98e-02, grad_scale: 8.0 +2023-03-20 20:16:11,248 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=20173.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:16:14,136 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6579, 5.1059, 5.1494, 4.9510, 4.8540, 4.5067, 5.0640, 4.9173], + device='cuda:0'), covar=tensor([0.0271, 0.0232, 0.0267, 0.0367, 0.0261, 0.0266, 0.0276, 0.0424], + device='cuda:0'), in_proj_covar=tensor([0.0129, 0.0136, 0.0109, 0.0106, 0.0090, 0.0128, 0.0114, 0.0095], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 20:16:27,608 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=20204.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:16:29,994 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.659e+02 3.295e+02 4.120e+02 4.990e+02 9.338e+02, threshold=8.240e+02, percent-clipped=6.0 +2023-03-20 20:16:35,077 INFO [train.py:901] (0/2) Epoch 8, batch 450, loss[loss=0.2006, simple_loss=0.2659, pruned_loss=0.06767, over 7248.00 frames. ], tot_loss[loss=0.2063, simple_loss=0.2691, pruned_loss=0.07174, over 1292583.34 frames. ], batch size: 55, lr: 1.98e-02, grad_scale: 8.0 +2023-03-20 20:16:37,529 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-20 20:16:38,042 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-20 20:16:58,779 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=20265.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:17:00,643 INFO [train.py:901] (0/2) Epoch 8, batch 500, loss[loss=0.2758, simple_loss=0.3196, pruned_loss=0.116, over 6734.00 frames. ], tot_loss[loss=0.2067, simple_loss=0.2697, pruned_loss=0.07191, over 1324881.58 frames. ], batch size: 106, lr: 1.97e-02, grad_scale: 8.0 +2023-03-20 20:17:03,282 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=20274.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:17:03,508 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.02 vs. limit=2.0 +2023-03-20 20:17:10,319 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-20 20:17:12,868 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-20 20:17:13,390 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-20 20:17:15,349 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-20 20:17:19,858 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-20 20:17:21,309 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.055e+02 2.899e+02 3.508e+02 4.098e+02 7.899e+02, threshold=7.016e+02, percent-clipped=0.0 +2023-03-20 20:17:26,335 INFO [train.py:901] (0/2) Epoch 8, batch 550, loss[loss=0.2312, simple_loss=0.2885, pruned_loss=0.08688, over 7311.00 frames. ], tot_loss[loss=0.2054, simple_loss=0.2684, pruned_loss=0.0712, over 1351301.49 frames. ], batch size: 80, lr: 1.97e-02, grad_scale: 8.0 +2023-03-20 20:17:27,866 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=20322.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:17:30,824 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-20 20:17:36,578 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4984, 2.3244, 2.6504, 2.6380, 2.5406, 2.0519, 2.0311, 2.7096], + device='cuda:0'), covar=tensor([0.1340, 0.0220, 0.0932, 0.1137, 0.0778, 0.1512, 0.2116, 0.0776], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0025, 0.0030, 0.0029, 0.0026, 0.0026, 0.0036, 0.0026], + device='cuda:0'), out_proj_covar=tensor([9.7208e-05, 8.8940e-05, 1.0177e-04, 9.9604e-05, 9.4602e-05, 9.5458e-05, + 1.1512e-04, 9.6548e-05], device='cuda:0') +2023-03-20 20:17:39,598 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-20 20:17:42,621 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-20 20:17:49,171 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-20 20:17:50,111 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-20 20:17:52,184 INFO [train.py:901] (0/2) Epoch 8, batch 600, loss[loss=0.2116, simple_loss=0.273, pruned_loss=0.07509, over 7320.00 frames. ], tot_loss[loss=0.2066, simple_loss=0.2697, pruned_loss=0.07175, over 1371140.56 frames. ], batch size: 75, lr: 1.97e-02, grad_scale: 8.0 +2023-03-20 20:17:56,298 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=20376.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:18:02,369 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=20388.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:18:04,806 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-20 20:18:13,129 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.802e+02 3.123e+02 3.615e+02 4.494e+02 7.871e+02, threshold=7.229e+02, percent-clipped=1.0 +2023-03-20 20:18:13,172 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-20 20:18:17,991 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=20418.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:18:18,372 INFO [train.py:901] (0/2) Epoch 8, batch 650, loss[loss=0.1724, simple_loss=0.225, pruned_loss=0.0599, over 7025.00 frames. ], tot_loss[loss=0.2057, simple_loss=0.2693, pruned_loss=0.07101, over 1388061.32 frames. ], batch size: 35, lr: 1.97e-02, grad_scale: 8.0 +2023-03-20 20:18:30,958 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-20 20:18:32,024 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=20445.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:18:34,108 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=20449.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:18:41,216 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-20 20:18:44,349 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=20468.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:18:44,830 INFO [train.py:901] (0/2) Epoch 8, batch 700, loss[loss=0.2204, simple_loss=0.2787, pruned_loss=0.08106, over 7352.00 frames. ], tot_loss[loss=0.2054, simple_loss=0.2689, pruned_loss=0.07094, over 1397992.96 frames. ], batch size: 54, lr: 1.96e-02, grad_scale: 8.0 +2023-03-20 20:18:45,489 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9832, 2.5326, 2.0268, 3.2896, 2.7165, 3.1737, 3.0459, 2.9944], + device='cuda:0'), covar=tensor([0.1200, 0.0468, 0.1579, 0.0325, 0.0019, 0.0053, 0.0034, 0.0029], + device='cuda:0'), in_proj_covar=tensor([0.0293, 0.0216, 0.0267, 0.0211, 0.0111, 0.0113, 0.0118, 0.0115], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 20:18:49,965 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=20479.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:19:03,854 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-20 20:19:04,393 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-20 20:19:04,878 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.681e+02 2.952e+02 3.612e+02 4.322e+02 1.243e+03, threshold=7.225e+02, percent-clipped=3.0 +2023-03-20 20:19:10,609 INFO [train.py:901] (0/2) Epoch 8, batch 750, loss[loss=0.2116, simple_loss=0.2768, pruned_loss=0.07321, over 7340.00 frames. ], tot_loss[loss=0.2051, simple_loss=0.2686, pruned_loss=0.07081, over 1409851.27 frames. ], batch size: 75, lr: 1.96e-02, grad_scale: 8.0 +2023-03-20 20:19:18,570 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-20 20:19:18,651 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7898, 5.2547, 5.3030, 5.1196, 5.0160, 4.7939, 5.2945, 5.0875], + device='cuda:0'), covar=tensor([0.0289, 0.0297, 0.0282, 0.0393, 0.0293, 0.0260, 0.0263, 0.0456], + device='cuda:0'), in_proj_covar=tensor([0.0135, 0.0141, 0.0110, 0.0107, 0.0095, 0.0130, 0.0116, 0.0095], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 20:19:23,737 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-20 20:19:23,884 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8749, 2.1669, 2.1042, 3.1989, 1.4099, 2.9187, 1.1963, 3.1596], + device='cuda:0'), covar=tensor([0.0046, 0.0854, 0.1680, 0.0042, 0.4162, 0.0056, 0.1227, 0.0084], + device='cuda:0'), in_proj_covar=tensor([0.0133, 0.0257, 0.0320, 0.0125, 0.0302, 0.0136, 0.0263, 0.0164], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0003, 0.0001, 0.0003, 0.0001, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 20:19:29,762 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-20 20:19:30,829 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-20 20:19:31,932 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=20560.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:19:36,347 INFO [train.py:901] (0/2) Epoch 8, batch 800, loss[loss=0.214, simple_loss=0.2792, pruned_loss=0.07438, over 7356.00 frames. ], tot_loss[loss=0.205, simple_loss=0.2687, pruned_loss=0.07071, over 1415640.42 frames. ], batch size: 63, lr: 1.96e-02, grad_scale: 8.0 +2023-03-20 20:19:41,464 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-20 20:19:46,135 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=20588.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:19:57,367 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.523e+02 2.783e+02 3.356e+02 4.383e+02 1.008e+03, threshold=6.712e+02, percent-clipped=2.0 +2023-03-20 20:20:00,948 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-20 20:20:01,462 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-20 20:20:02,446 INFO [train.py:901] (0/2) Epoch 8, batch 850, loss[loss=0.211, simple_loss=0.2756, pruned_loss=0.07316, over 7279.00 frames. ], tot_loss[loss=0.2041, simple_loss=0.268, pruned_loss=0.07011, over 1422770.42 frames. ], batch size: 86, lr: 1.96e-02, grad_scale: 8.0 +2023-03-20 20:20:07,060 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-20 20:20:08,922 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.84 vs. limit=2.0 +2023-03-20 20:20:11,175 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-20 20:20:18,255 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=20649.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:20:27,910 INFO [train.py:901] (0/2) Epoch 8, batch 900, loss[loss=0.2053, simple_loss=0.2735, pruned_loss=0.06856, over 7286.00 frames. ], tot_loss[loss=0.2041, simple_loss=0.2677, pruned_loss=0.07031, over 1424349.11 frames. ], batch size: 80, lr: 1.96e-02, grad_scale: 8.0 +2023-03-20 20:20:31,469 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=20676.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:20:48,348 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.807e+02 3.423e+02 4.064e+02 5.075e+02 9.468e+02, threshold=8.127e+02, percent-clipped=9.0 +2023-03-20 20:20:48,381 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-20 20:20:49,034 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4796, 1.2146, 1.5448, 1.2445, 1.6280, 1.5113, 1.3340, 1.0547], + device='cuda:0'), covar=tensor([0.0084, 0.0162, 0.0165, 0.0066, 0.0075, 0.0117, 0.0185, 0.0202], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0018, 0.0017, 0.0017, 0.0020, 0.0018, 0.0018, 0.0021], + device='cuda:0'), out_proj_covar=tensor([2.5357e-05, 2.0701e-05, 2.1962e-05, 1.8471e-05, 2.2283e-05, 2.1970e-05, + 2.2245e-05, 2.9748e-05], device='cuda:0') +2023-03-20 20:20:54,018 INFO [train.py:901] (0/2) Epoch 8, batch 950, loss[loss=0.207, simple_loss=0.2685, pruned_loss=0.07276, over 7271.00 frames. ], tot_loss[loss=0.2046, simple_loss=0.2682, pruned_loss=0.07055, over 1429933.86 frames. ], batch size: 52, lr: 1.95e-02, grad_scale: 8.0 +2023-03-20 20:20:56,543 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=20724.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:21:03,082 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0604, 3.5973, 3.7260, 3.5725, 3.5958, 3.6670, 3.9964, 3.5387], + device='cuda:0'), covar=tensor([0.0078, 0.0137, 0.0149, 0.0151, 0.0176, 0.0101, 0.0123, 0.0139], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0047, 0.0053, 0.0041, 0.0065, 0.0056, 0.0050, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 20:21:06,592 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=20744.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:21:07,144 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=20745.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:21:11,505 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-20 20:21:18,651 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=20768.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:21:19,034 INFO [train.py:901] (0/2) Epoch 8, batch 1000, loss[loss=0.1935, simple_loss=0.2607, pruned_loss=0.06316, over 7261.00 frames. ], tot_loss[loss=0.205, simple_loss=0.2682, pruned_loss=0.07088, over 1432829.02 frames. ], batch size: 52, lr: 1.95e-02, grad_scale: 8.0 +2023-03-20 20:21:21,368 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.86 vs. limit=2.0 +2023-03-20 20:21:21,628 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=20774.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:21:31,703 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=20793.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:21:32,156 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-20 20:21:36,602 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6405, 2.4974, 2.7493, 2.6973, 2.7503, 2.4256, 1.8931, 2.4560], + device='cuda:0'), covar=tensor([0.1209, 0.0223, 0.1228, 0.1160, 0.0761, 0.1133, 0.2373, 0.1684], + device='cuda:0'), in_proj_covar=tensor([0.0030, 0.0027, 0.0033, 0.0034, 0.0030, 0.0028, 0.0041, 0.0030], + device='cuda:0'), out_proj_covar=tensor([1.0840e-04, 9.9331e-05, 1.1478e-04, 1.1705e-04, 1.0706e-04, 1.0623e-04, + 1.3317e-04, 1.1094e-04], device='cuda:0') +2023-03-20 20:21:40,615 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.591e+02 3.000e+02 3.544e+02 4.329e+02 8.832e+02, threshold=7.088e+02, percent-clipped=1.0 +2023-03-20 20:21:44,206 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=20816.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:21:45,677 INFO [train.py:901] (0/2) Epoch 8, batch 1050, loss[loss=0.2167, simple_loss=0.2798, pruned_loss=0.07675, over 7329.00 frames. ], tot_loss[loss=0.2045, simple_loss=0.2679, pruned_loss=0.0705, over 1433793.09 frames. ], batch size: 75, lr: 1.95e-02, grad_scale: 8.0 +2023-03-20 20:21:54,656 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-20 20:21:58,685 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-20 20:22:06,484 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=20860.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:22:10,334 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.02 vs. limit=2.0 +2023-03-20 20:22:11,561 INFO [train.py:901] (0/2) Epoch 8, batch 1100, loss[loss=0.222, simple_loss=0.288, pruned_loss=0.07803, over 7317.00 frames. ], tot_loss[loss=0.2039, simple_loss=0.2675, pruned_loss=0.07011, over 1437154.65 frames. ], batch size: 59, lr: 1.95e-02, grad_scale: 8.0 +2023-03-20 20:22:28,658 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-20 20:22:29,130 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 20:22:31,670 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=20908.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:22:32,090 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.584e+02 2.881e+02 3.698e+02 4.714e+02 1.346e+03, threshold=7.395e+02, percent-clipped=2.0 +2023-03-20 20:22:37,180 INFO [train.py:901] (0/2) Epoch 8, batch 1150, loss[loss=0.2184, simple_loss=0.2787, pruned_loss=0.07899, over 7251.00 frames. ], tot_loss[loss=0.204, simple_loss=0.2673, pruned_loss=0.0704, over 1438874.40 frames. ], batch size: 89, lr: 1.94e-02, grad_scale: 8.0 +2023-03-20 20:22:42,239 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-20 20:22:42,255 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-20 20:22:44,909 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1715, 0.9350, 0.7740, 1.2197, 1.2527, 1.1824, 1.0542, 1.0153], + device='cuda:0'), covar=tensor([0.0640, 0.3159, 0.0721, 0.1006, 0.0814, 0.1255, 0.0786, 0.1732], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0036, 0.0031, 0.0029, 0.0028, 0.0027, 0.0035, 0.0034], + device='cuda:0'), out_proj_covar=tensor([5.5576e-05, 7.6010e-05, 5.4379e-05, 5.4632e-05, 5.7798e-05, 5.6192e-05, + 6.7353e-05, 6.6320e-05], device='cuda:0') +2023-03-20 20:22:49,929 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=20944.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:23:02,934 INFO [train.py:901] (0/2) Epoch 8, batch 1200, loss[loss=0.255, simple_loss=0.3062, pruned_loss=0.1019, over 6802.00 frames. ], tot_loss[loss=0.2034, simple_loss=0.2665, pruned_loss=0.07015, over 1439637.07 frames. ], batch size: 107, lr: 1.94e-02, grad_scale: 8.0 +2023-03-20 20:23:16,149 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-20 20:23:24,113 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.032e+02 2.800e+02 3.354e+02 4.128e+02 7.316e+02, threshold=6.709e+02, percent-clipped=0.0 +2023-03-20 20:23:28,993 INFO [train.py:901] (0/2) Epoch 8, batch 1250, loss[loss=0.2164, simple_loss=0.2727, pruned_loss=0.08002, over 7268.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2659, pruned_loss=0.0695, over 1439745.29 frames. ], batch size: 57, lr: 1.94e-02, grad_scale: 8.0 +2023-03-20 20:23:29,779 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.34 vs. limit=5.0 +2023-03-20 20:23:39,944 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-20 20:23:40,764 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.32 vs. limit=5.0 +2023-03-20 20:23:42,151 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=21044.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:23:44,449 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-20 20:23:45,483 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-20 20:23:46,097 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=21052.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:23:53,717 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4059, 2.3872, 2.1805, 3.4365, 1.4805, 3.2603, 1.2109, 3.4030], + device='cuda:0'), covar=tensor([0.0044, 0.0799, 0.1618, 0.0022, 0.5055, 0.0064, 0.1104, 0.0104], + device='cuda:0'), in_proj_covar=tensor([0.0138, 0.0262, 0.0323, 0.0126, 0.0316, 0.0140, 0.0276, 0.0170], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0003, 0.0003, 0.0001, 0.0003, 0.0001, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 20:23:54,572 INFO [train.py:901] (0/2) Epoch 8, batch 1300, loss[loss=0.1658, simple_loss=0.2355, pruned_loss=0.04799, over 7322.00 frames. ], tot_loss[loss=0.2029, simple_loss=0.2663, pruned_loss=0.0697, over 1441728.85 frames. ], batch size: 44, lr: 1.94e-02, grad_scale: 8.0 +2023-03-20 20:23:57,702 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=21074.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:24:06,735 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=21092.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:24:08,755 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-20 20:24:11,234 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-20 20:24:13,817 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-20 20:24:15,293 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.647e+02 3.023e+02 3.781e+02 4.755e+02 9.380e+02, threshold=7.562e+02, percent-clipped=5.0 +2023-03-20 20:24:17,469 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=21113.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:24:19,975 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=21118.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:24:20,340 INFO [train.py:901] (0/2) Epoch 8, batch 1350, loss[loss=0.2167, simple_loss=0.2746, pruned_loss=0.07937, over 7138.00 frames. ], tot_loss[loss=0.2039, simple_loss=0.2669, pruned_loss=0.07044, over 1439957.51 frames. ], batch size: 41, lr: 1.94e-02, grad_scale: 8.0 +2023-03-20 20:24:21,897 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=21122.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:24:23,866 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-20 20:24:46,754 INFO [train.py:901] (0/2) Epoch 8, batch 1400, loss[loss=0.2034, simple_loss=0.2608, pruned_loss=0.07298, over 7292.00 frames. ], tot_loss[loss=0.2031, simple_loss=0.266, pruned_loss=0.07008, over 1438691.75 frames. ], batch size: 68, lr: 1.93e-02, grad_scale: 8.0 +2023-03-20 20:24:47,920 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0623, 3.3669, 2.7725, 2.7883, 3.2906, 2.3244, 3.1639, 2.2460], + device='cuda:0'), covar=tensor([0.1029, 0.0713, 0.0282, 0.0502, 0.0984, 0.1005, 0.0259, 0.1469], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0033, 0.0041, 0.0036, 0.0037, 0.0038, 0.0034, 0.0035], + device='cuda:0'), out_proj_covar=tensor([1.2033e-04, 9.8836e-05, 1.1364e-04, 1.0352e-04, 1.0771e-04, 1.0901e-04, + 9.9218e-05, 1.0192e-04], device='cuda:0') +2023-03-20 20:24:51,902 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=21179.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:24:58,772 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-20 20:25:07,012 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.034e+02 2.789e+02 3.546e+02 4.120e+02 7.896e+02, threshold=7.092e+02, percent-clipped=1.0 +2023-03-20 20:25:11,981 INFO [train.py:901] (0/2) Epoch 8, batch 1450, loss[loss=0.2039, simple_loss=0.2672, pruned_loss=0.07028, over 7287.00 frames. ], tot_loss[loss=0.2027, simple_loss=0.2658, pruned_loss=0.06985, over 1437841.71 frames. ], batch size: 66, lr: 1.93e-02, grad_scale: 8.0 +2023-03-20 20:25:18,470 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.63 vs. limit=2.0 +2023-03-20 20:25:20,333 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6695, 3.3296, 3.3630, 3.1889, 3.2799, 3.1641, 3.5101, 3.3000], + device='cuda:0'), covar=tensor([0.0086, 0.0164, 0.0181, 0.0208, 0.0242, 0.0147, 0.0167, 0.0149], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0049, 0.0055, 0.0043, 0.0069, 0.0059, 0.0054, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 20:25:22,291 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-20 20:25:25,308 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=21244.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:25:37,958 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-20 20:25:38,430 INFO [train.py:901] (0/2) Epoch 8, batch 1500, loss[loss=0.1979, simple_loss=0.2706, pruned_loss=0.06266, over 7291.00 frames. ], tot_loss[loss=0.2032, simple_loss=0.2668, pruned_loss=0.06982, over 1437890.17 frames. ], batch size: 68, lr: 1.93e-02, grad_scale: 8.0 +2023-03-20 20:25:50,159 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=21292.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:25:59,226 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.558e+02 2.932e+02 3.502e+02 4.369e+02 7.358e+02, threshold=7.004e+02, percent-clipped=1.0 +2023-03-20 20:26:00,871 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=21312.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:26:01,769 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-20 20:26:04,319 INFO [train.py:901] (0/2) Epoch 8, batch 1550, loss[loss=0.2112, simple_loss=0.2735, pruned_loss=0.07444, over 7281.00 frames. ], tot_loss[loss=0.2029, simple_loss=0.2668, pruned_loss=0.06946, over 1439720.41 frames. ], batch size: 86, lr: 1.93e-02, grad_scale: 8.0 +2023-03-20 20:26:06,941 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2646, 1.5561, 2.0763, 1.6909, 1.2360, 1.4463, 0.8906, 1.1731], + device='cuda:0'), covar=tensor([0.0313, 0.0282, 0.0077, 0.0056, 0.0361, 0.0201, 0.0377, 0.0275], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0019, 0.0020, 0.0019, 0.0019, 0.0017, 0.0020, 0.0020], + device='cuda:0'), out_proj_covar=tensor([4.7686e-05, 4.4286e-05, 4.3827e-05, 3.6921e-05, 4.4566e-05, 3.8933e-05, + 4.4709e-05, 4.7456e-05], device='cuda:0') +2023-03-20 20:26:08,922 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=21328.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:26:30,052 INFO [train.py:901] (0/2) Epoch 8, batch 1600, loss[loss=0.2219, simple_loss=0.2807, pruned_loss=0.0816, over 7249.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2664, pruned_loss=0.06918, over 1439815.87 frames. ], batch size: 64, lr: 1.93e-02, grad_scale: 8.0 +2023-03-20 20:26:32,188 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=21373.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:26:33,042 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-20 20:26:34,116 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-20 20:26:34,277 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1109, 2.4714, 2.3441, 3.6877, 1.4265, 2.9376, 1.3890, 3.2516], + device='cuda:0'), covar=tensor([0.0071, 0.0816, 0.1613, 0.0036, 0.4857, 0.0057, 0.1091, 0.0075], + device='cuda:0'), in_proj_covar=tensor([0.0142, 0.0263, 0.0325, 0.0126, 0.0306, 0.0141, 0.0278, 0.0170], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0003, 0.0003, 0.0001, 0.0003, 0.0001, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 20:26:36,706 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-20 20:26:38,320 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1744, 4.0256, 3.9607, 4.4568, 4.4344, 4.4426, 3.9315, 3.9237], + device='cuda:0'), covar=tensor([0.0944, 0.2295, 0.2229, 0.1176, 0.0812, 0.1407, 0.0735, 0.1095], + device='cuda:0'), in_proj_covar=tensor([0.0142, 0.0236, 0.0197, 0.0190, 0.0146, 0.0248, 0.0133, 0.0161], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0001, 0.0002], + device='cuda:0') +2023-03-20 20:26:40,375 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=21389.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:26:47,117 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-20 20:26:50,737 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=21408.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:26:51,134 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.861e+02 3.255e+02 3.633e+02 4.410e+02 1.218e+03, threshold=7.266e+02, percent-clipped=6.0 +2023-03-20 20:26:51,175 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-20 20:26:54,823 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6042, 1.7256, 2.3598, 1.8666, 1.0327, 1.4922, 1.1427, 1.3711], + device='cuda:0'), covar=tensor([0.0122, 0.0214, 0.0110, 0.0051, 0.0446, 0.0372, 0.0290, 0.0314], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0020, 0.0020, 0.0019, 0.0020, 0.0018, 0.0020, 0.0020], + device='cuda:0'), out_proj_covar=tensor([4.7478e-05, 4.6056e-05, 4.3941e-05, 3.7271e-05, 4.5793e-05, 4.1078e-05, + 4.5022e-05, 4.8643e-05], device='cuda:0') +2023-03-20 20:26:56,169 INFO [train.py:901] (0/2) Epoch 8, batch 1650, loss[loss=0.2135, simple_loss=0.2817, pruned_loss=0.07267, over 7130.00 frames. ], tot_loss[loss=0.2032, simple_loss=0.267, pruned_loss=0.06968, over 1441519.69 frames. ], batch size: 98, lr: 1.92e-02, grad_scale: 8.0 +2023-03-20 20:27:00,383 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-20 20:27:11,272 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.90 vs. limit=5.0 +2023-03-20 20:27:15,895 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 20:27:20,400 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-20 20:27:21,867 INFO [train.py:901] (0/2) Epoch 8, batch 1700, loss[loss=0.2043, simple_loss=0.273, pruned_loss=0.06783, over 7322.00 frames. ], tot_loss[loss=0.2028, simple_loss=0.2667, pruned_loss=0.06948, over 1441770.96 frames. ], batch size: 59, lr: 1.92e-02, grad_scale: 8.0 +2023-03-20 20:27:24,461 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=21474.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:27:31,045 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-20 20:27:32,885 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.89 vs. limit=2.0 +2023-03-20 20:27:43,084 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.770e+02 2.755e+02 3.361e+02 4.411e+02 8.082e+02, threshold=6.722e+02, percent-clipped=3.0 +2023-03-20 20:27:48,216 INFO [train.py:901] (0/2) Epoch 8, batch 1750, loss[loss=0.2122, simple_loss=0.2771, pruned_loss=0.07368, over 7278.00 frames. ], tot_loss[loss=0.2019, simple_loss=0.266, pruned_loss=0.0689, over 1442925.26 frames. ], batch size: 57, lr: 1.92e-02, grad_scale: 8.0 +2023-03-20 20:27:56,273 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-20 20:27:57,275 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-20 20:28:13,430 INFO [train.py:901] (0/2) Epoch 8, batch 1800, loss[loss=0.1429, simple_loss=0.1964, pruned_loss=0.04465, over 5757.00 frames. ], tot_loss[loss=0.2016, simple_loss=0.2658, pruned_loss=0.06873, over 1443039.59 frames. ], batch size: 25, lr: 1.92e-02, grad_scale: 8.0 +2023-03-20 20:28:19,546 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-20 20:28:27,757 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=21595.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:28:32,505 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-20 20:28:34,961 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.818e+02 2.908e+02 3.364e+02 4.556e+02 1.018e+03, threshold=6.727e+02, percent-clipped=4.0 +2023-03-20 20:28:37,121 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=21613.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:28:40,032 INFO [train.py:901] (0/2) Epoch 8, batch 1850, loss[loss=0.2122, simple_loss=0.2765, pruned_loss=0.07396, over 7252.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2666, pruned_loss=0.0691, over 1444101.77 frames. ], batch size: 70, lr: 1.91e-02, grad_scale: 8.0 +2023-03-20 20:28:40,416 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.96 vs. limit=2.0 +2023-03-20 20:28:42,540 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-20 20:28:45,693 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8994, 2.0353, 1.7176, 1.6753, 2.1258, 1.7718, 1.7869, 1.8204], + device='cuda:0'), covar=tensor([0.1224, 0.0677, 0.1124, 0.1031, 0.0702, 0.0754, 0.0879, 0.0968], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0033, 0.0042, 0.0038, 0.0037, 0.0036, 0.0036, 0.0034], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 20:28:50,761 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4163, 3.7406, 3.9800, 3.8624, 3.9639, 4.0558, 4.1618, 3.8102], + device='cuda:0'), covar=tensor([0.0080, 0.0136, 0.0158, 0.0146, 0.0141, 0.0084, 0.0123, 0.0105], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0048, 0.0052, 0.0043, 0.0068, 0.0056, 0.0053, 0.0048], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 20:28:58,849 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-20 20:28:59,463 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=21656.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:29:04,912 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1031, 4.6338, 4.6061, 4.5989, 4.5461, 4.2030, 4.6584, 4.5186], + device='cuda:0'), covar=tensor([0.0480, 0.0375, 0.0422, 0.0437, 0.0309, 0.0329, 0.0364, 0.0496], + device='cuda:0'), in_proj_covar=tensor([0.0138, 0.0145, 0.0110, 0.0109, 0.0091, 0.0135, 0.0119, 0.0096], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 20:29:05,416 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=21668.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:29:05,850 INFO [train.py:901] (0/2) Epoch 8, batch 1900, loss[loss=0.1779, simple_loss=0.2459, pruned_loss=0.05491, over 7326.00 frames. ], tot_loss[loss=0.2032, simple_loss=0.2674, pruned_loss=0.06952, over 1444694.52 frames. ], batch size: 44, lr: 1.91e-02, grad_scale: 8.0 +2023-03-20 20:29:09,319 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=21674.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:29:14,237 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=21684.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:29:23,729 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-20 20:29:26,324 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=21708.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:29:26,680 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.930e+02 2.999e+02 3.585e+02 4.195e+02 1.317e+03, threshold=7.171e+02, percent-clipped=5.0 +2023-03-20 20:29:31,655 INFO [train.py:901] (0/2) Epoch 8, batch 1950, loss[loss=0.1788, simple_loss=0.2502, pruned_loss=0.05368, over 7246.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.2666, pruned_loss=0.0688, over 1446256.84 frames. ], batch size: 47, lr: 1.91e-02, grad_scale: 8.0 +2023-03-20 20:29:35,061 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-20 20:29:39,592 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-20 20:29:40,117 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-20 20:29:50,825 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=21756.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:29:55,104 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=21763.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:29:57,989 INFO [train.py:901] (0/2) Epoch 8, batch 2000, loss[loss=0.2047, simple_loss=0.2725, pruned_loss=0.06851, over 7288.00 frames. ], tot_loss[loss=0.2021, simple_loss=0.2666, pruned_loss=0.06877, over 1446090.00 frames. ], batch size: 68, lr: 1.91e-02, grad_scale: 8.0 +2023-03-20 20:29:58,536 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-20 20:30:00,730 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=21774.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:30:09,233 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-20 20:30:17,675 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-20 20:30:18,619 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.703e+02 2.773e+02 3.531e+02 4.246e+02 7.839e+02, threshold=7.062e+02, percent-clipped=2.0 +2023-03-20 20:30:23,666 INFO [train.py:901] (0/2) Epoch 8, batch 2050, loss[loss=0.1658, simple_loss=0.2362, pruned_loss=0.04771, over 7125.00 frames. ], tot_loss[loss=0.2014, simple_loss=0.2661, pruned_loss=0.06836, over 1446746.26 frames. ], batch size: 41, lr: 1.91e-02, grad_scale: 8.0 +2023-03-20 20:30:25,218 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=21822.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:30:26,314 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=21824.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:30:40,536 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=21850.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:30:50,073 INFO [train.py:901] (0/2) Epoch 8, batch 2100, loss[loss=0.1834, simple_loss=0.2403, pruned_loss=0.06324, over 7206.00 frames. ], tot_loss[loss=0.2018, simple_loss=0.2664, pruned_loss=0.06858, over 1446342.95 frames. ], batch size: 39, lr: 1.90e-02, grad_scale: 8.0 +2023-03-20 20:30:51,598 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-20 20:30:54,155 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-20 20:31:10,068 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.845e+02 2.957e+02 3.693e+02 4.349e+02 9.898e+02, threshold=7.385e+02, percent-clipped=1.0 +2023-03-20 20:31:11,256 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=21911.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:31:14,986 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.51 vs. limit=5.0 +2023-03-20 20:31:15,085 INFO [train.py:901] (0/2) Epoch 8, batch 2150, loss[loss=0.2271, simple_loss=0.2862, pruned_loss=0.08399, over 7367.00 frames. ], tot_loss[loss=0.202, simple_loss=0.2663, pruned_loss=0.06889, over 1446490.30 frames. ], batch size: 51, lr: 1.90e-02, grad_scale: 8.0 +2023-03-20 20:31:21,427 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6990, 3.6158, 3.4624, 3.5714, 3.5907, 3.6551, 3.6709, 3.3788], + device='cuda:0'), covar=tensor([0.0050, 0.0121, 0.0074, 0.0075, 0.0062, 0.0063, 0.0066, 0.0086], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0034, 0.0033, 0.0030, 0.0032, 0.0031, 0.0040, 0.0037], + device='cuda:0'), out_proj_covar=tensor([8.0968e-05, 1.1116e-04, 1.0825e-04, 8.8189e-05, 9.5264e-05, 9.2265e-05, + 1.3054e-04, 1.1200e-04], device='cuda:0') +2023-03-20 20:31:25,807 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.97 vs. limit=2.0 +2023-03-20 20:31:32,738 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=21951.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:31:40,659 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-20 20:31:41,238 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=21968.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:31:41,635 INFO [train.py:901] (0/2) Epoch 8, batch 2200, loss[loss=0.1972, simple_loss=0.2662, pruned_loss=0.0641, over 7333.00 frames. ], tot_loss[loss=0.2016, simple_loss=0.2661, pruned_loss=0.06857, over 1447451.32 frames. ], batch size: 61, lr: 1.90e-02, grad_scale: 16.0 +2023-03-20 20:31:41,730 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=21969.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:31:42,281 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=21970.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:31:49,382 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=21984.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:32:02,857 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.683e+02 2.848e+02 3.645e+02 4.438e+02 1.724e+03, threshold=7.290e+02, percent-clipped=4.0 +2023-03-20 20:32:06,493 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=22016.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:32:07,937 INFO [train.py:901] (0/2) Epoch 8, batch 2250, loss[loss=0.1742, simple_loss=0.2511, pruned_loss=0.04867, over 7341.00 frames. ], tot_loss[loss=0.2007, simple_loss=0.2652, pruned_loss=0.06811, over 1443520.89 frames. ], batch size: 61, lr: 1.90e-02, grad_scale: 16.0 +2023-03-20 20:32:14,703 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22031.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:32:15,115 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=22032.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:32:15,580 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-20 20:32:16,077 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-20 20:32:22,464 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.75 vs. limit=5.0 +2023-03-20 20:32:28,688 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-20 20:32:33,840 INFO [train.py:901] (0/2) Epoch 8, batch 2300, loss[loss=0.2359, simple_loss=0.2975, pruned_loss=0.08714, over 6636.00 frames. ], tot_loss[loss=0.2008, simple_loss=0.2655, pruned_loss=0.06806, over 1445148.34 frames. ], batch size: 106, lr: 1.90e-02, grad_scale: 16.0 +2023-03-20 20:32:51,300 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9365, 2.2003, 2.3004, 3.1630, 1.3960, 2.8623, 1.2980, 3.2102], + device='cuda:0'), covar=tensor([0.0035, 0.0839, 0.1579, 0.0087, 0.4897, 0.0067, 0.1135, 0.0122], + device='cuda:0'), in_proj_covar=tensor([0.0136, 0.0262, 0.0323, 0.0131, 0.0313, 0.0142, 0.0274, 0.0176], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0003, 0.0003, 0.0001, 0.0003, 0.0001, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 20:32:55,475 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.018e+02 3.001e+02 3.476e+02 4.126e+02 8.245e+02, threshold=6.952e+02, percent-clipped=1.0 +2023-03-20 20:33:00,581 INFO [train.py:901] (0/2) Epoch 8, batch 2350, loss[loss=0.2209, simple_loss=0.2849, pruned_loss=0.07844, over 7283.00 frames. ], tot_loss[loss=0.2004, simple_loss=0.265, pruned_loss=0.06785, over 1441576.97 frames. ], batch size: 57, lr: 1.89e-02, grad_scale: 16.0 +2023-03-20 20:33:00,666 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=22119.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:33:06,201 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=22130.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:33:15,782 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-20 20:33:21,846 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-20 20:33:25,879 INFO [train.py:901] (0/2) Epoch 8, batch 2400, loss[loss=0.1659, simple_loss=0.2327, pruned_loss=0.04952, over 7178.00 frames. ], tot_loss[loss=0.2006, simple_loss=0.2649, pruned_loss=0.06814, over 1440916.10 frames. ], batch size: 39, lr: 1.89e-02, grad_scale: 16.0 +2023-03-20 20:33:26,546 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3218, 3.0697, 2.1098, 3.2683, 2.7554, 3.2054, 1.9924, 1.8933], + device='cuda:0'), covar=tensor([0.0071, 0.0298, 0.0945, 0.0148, 0.0119, 0.0119, 0.1187, 0.0912], + device='cuda:0'), in_proj_covar=tensor([0.0155, 0.0186, 0.0293, 0.0175, 0.0191, 0.0181, 0.0268, 0.0275], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 20:33:29,413 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=22176.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:33:32,515 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-20 20:33:35,087 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-20 20:33:35,826 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.31 vs. limit=2.0 +2023-03-20 20:33:38,286 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22191.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:33:46,070 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=22206.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:33:47,353 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.09 vs. limit=2.0 +2023-03-20 20:33:47,540 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.919e+02 2.682e+02 3.425e+02 4.426e+02 1.107e+03, threshold=6.849e+02, percent-clipped=3.0 +2023-03-20 20:33:50,723 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=22215.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:33:52,640 INFO [train.py:901] (0/2) Epoch 8, batch 2450, loss[loss=0.1802, simple_loss=0.2595, pruned_loss=0.05042, over 7307.00 frames. ], tot_loss[loss=0.2, simple_loss=0.2645, pruned_loss=0.0678, over 1439728.08 frames. ], batch size: 86, lr: 1.89e-02, grad_scale: 16.0 +2023-03-20 20:34:01,891 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=22237.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:34:01,905 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22237.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:34:02,272 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-20 20:34:08,824 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=22251.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:34:18,449 INFO [train.py:901] (0/2) Epoch 8, batch 2500, loss[loss=0.2025, simple_loss=0.2694, pruned_loss=0.06786, over 7302.00 frames. ], tot_loss[loss=0.1996, simple_loss=0.2642, pruned_loss=0.06743, over 1438912.50 frames. ], batch size: 68, lr: 1.89e-02, grad_scale: 16.0 +2023-03-20 20:34:18,542 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=22269.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:34:22,673 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22276.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:34:28,397 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-20 20:34:33,406 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22298.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:34:33,841 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=22299.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:34:38,852 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.527e+02 2.963e+02 3.499e+02 4.353e+02 8.938e+02, threshold=6.998e+02, percent-clipped=3.0 +2023-03-20 20:34:42,931 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=22317.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:34:43,842 INFO [train.py:901] (0/2) Epoch 8, batch 2550, loss[loss=0.2295, simple_loss=0.2899, pruned_loss=0.08461, over 7308.00 frames. ], tot_loss[loss=0.1995, simple_loss=0.2642, pruned_loss=0.06738, over 1440887.15 frames. ], batch size: 86, lr: 1.89e-02, grad_scale: 16.0 +2023-03-20 20:34:47,420 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=22326.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:34:56,166 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-20 20:35:08,734 INFO [train.py:901] (0/2) Epoch 8, batch 2600, loss[loss=0.1903, simple_loss=0.257, pruned_loss=0.06184, over 7343.00 frames. ], tot_loss[loss=0.1997, simple_loss=0.2643, pruned_loss=0.06754, over 1441232.08 frames. ], batch size: 63, lr: 1.88e-02, grad_scale: 16.0 +2023-03-20 20:35:17,996 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=22385.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:35:30,554 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.811e+02 2.880e+02 3.559e+02 4.334e+02 1.074e+03, threshold=7.118e+02, percent-clipped=5.0 +2023-03-20 20:35:35,476 INFO [train.py:901] (0/2) Epoch 8, batch 2650, loss[loss=0.2036, simple_loss=0.2631, pruned_loss=0.07205, over 7284.00 frames. ], tot_loss[loss=0.2, simple_loss=0.2644, pruned_loss=0.06781, over 1439712.47 frames. ], batch size: 86, lr: 1.88e-02, grad_scale: 16.0 +2023-03-20 20:35:35,585 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=22419.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:35:44,773 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.87 vs. limit=2.0 +2023-03-20 20:35:48,892 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22446.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:35:59,217 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=22467.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:36:00,168 INFO [train.py:901] (0/2) Epoch 8, batch 2700, loss[loss=0.2112, simple_loss=0.2693, pruned_loss=0.07657, over 7312.00 frames. ], tot_loss[loss=0.2011, simple_loss=0.2655, pruned_loss=0.06833, over 1440283.79 frames. ], batch size: 49, lr: 1.88e-02, grad_scale: 16.0 +2023-03-20 20:36:08,762 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=22486.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:36:18,531 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=22506.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:36:19,865 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.942e+02 2.708e+02 3.424e+02 4.354e+02 9.279e+02, threshold=6.847e+02, percent-clipped=4.0 +2023-03-20 20:36:24,146 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.97 vs. limit=2.0 +2023-03-20 20:36:24,786 INFO [train.py:901] (0/2) Epoch 8, batch 2750, loss[loss=0.2265, simple_loss=0.2822, pruned_loss=0.08543, over 7261.00 frames. ], tot_loss[loss=0.2004, simple_loss=0.265, pruned_loss=0.06794, over 1441040.41 frames. ], batch size: 64, lr: 1.88e-02, grad_scale: 16.0 +2023-03-20 20:36:31,223 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=22532.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:36:38,306 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.88 vs. limit=5.0 +2023-03-20 20:36:41,906 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=22554.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:36:44,769 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.88 vs. limit=2.0 +2023-03-20 20:36:49,288 INFO [train.py:901] (0/2) Epoch 8, batch 2800, loss[loss=0.1944, simple_loss=0.2601, pruned_loss=0.0643, over 7200.00 frames. ], tot_loss[loss=0.1994, simple_loss=0.2638, pruned_loss=0.06748, over 1439473.93 frames. ], batch size: 50, lr: 1.88e-02, grad_scale: 16.0 +2023-03-20 20:36:50,264 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=22571.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:37:01,713 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-8.pt +2023-03-20 20:37:20,198 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-20 20:37:20,253 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-20 20:37:20,282 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-20 20:37:20,298 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-20 20:37:20,807 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-20 20:37:20,822 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-20 20:37:20,834 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-20 20:37:20,942 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-20 20:37:21,022 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-20 20:37:21,038 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-20 20:37:21,056 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-20 20:37:21,238 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-20 20:37:21,245 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-20 20:37:21,253 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-20 20:37:21,423 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-20 20:37:21,426 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-20 20:37:21,489 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-20 20:37:23,763 INFO [train.py:901] (0/2) Epoch 9, batch 0, loss[loss=0.1996, simple_loss=0.269, pruned_loss=0.06508, over 7290.00 frames. ], tot_loss[loss=0.1996, simple_loss=0.269, pruned_loss=0.06508, over 7290.00 frames. ], batch size: 77, lr: 1.80e-02, grad_scale: 16.0 +2023-03-20 20:37:23,765 INFO [train.py:926] (0/2) Computing validation loss +2023-03-20 20:37:48,980 INFO [train.py:935] (0/2) Epoch 9, validation: loss=0.1789, simple_loss=0.2654, pruned_loss=0.04614, over 1622729.00 frames. +2023-03-20 20:37:48,980 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-20 20:37:49,046 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=22593.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:37:55,444 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-20 20:37:58,048 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.041e+02 3.046e+02 3.484e+02 4.247e+02 9.512e+02, threshold=6.968e+02, percent-clipped=1.0 +2023-03-20 20:37:59,243 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1976, 1.3412, 1.2932, 1.1992, 0.9554, 0.8499, 0.9214, 0.7721], + device='cuda:0'), covar=tensor([0.0178, 0.0152, 0.0297, 0.0102, 0.0202, 0.0142, 0.0144, 0.0310], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0017, 0.0016, 0.0017, 0.0020, 0.0018, 0.0018, 0.0021], + device='cuda:0'), out_proj_covar=tensor([2.4015e-05, 2.0190e-05, 2.1133e-05, 1.9416e-05, 2.3126e-05, 2.0851e-05, + 2.1887e-05, 2.8364e-05], device='cuda:0') +2023-03-20 20:38:07,301 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-20 20:38:07,377 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=22626.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:38:13,673 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=22639.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:38:14,529 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-20 20:38:15,512 INFO [train.py:901] (0/2) Epoch 9, batch 50, loss[loss=0.1898, simple_loss=0.2591, pruned_loss=0.06024, over 7317.00 frames. ], tot_loss[loss=0.2046, simple_loss=0.27, pruned_loss=0.06964, over 327376.78 frames. ], batch size: 59, lr: 1.80e-02, grad_scale: 16.0 +2023-03-20 20:38:17,015 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-20 20:38:19,600 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-20 20:38:29,237 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1690, 1.2419, 1.4142, 1.2538, 1.0642, 0.8947, 0.9127, 0.6572], + device='cuda:0'), covar=tensor([0.0301, 0.0157, 0.0138, 0.0113, 0.0164, 0.0150, 0.0249, 0.0519], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0018, 0.0016, 0.0018, 0.0020, 0.0018, 0.0019, 0.0022], + device='cuda:0'), out_proj_covar=tensor([2.4287e-05, 2.0827e-05, 2.1609e-05, 2.0073e-05, 2.3409e-05, 2.1425e-05, + 2.2336e-05, 2.9971e-05], device='cuda:0') +2023-03-20 20:38:31,161 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=22674.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:38:32,245 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1404, 1.4454, 1.7730, 1.1867, 1.0402, 1.2987, 1.2331, 1.3040], + device='cuda:0'), covar=tensor([0.0263, 0.0289, 0.0126, 0.0080, 0.0401, 0.0108, 0.0130, 0.0163], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0019, 0.0019, 0.0019, 0.0020, 0.0016, 0.0018, 0.0019], + device='cuda:0'), out_proj_covar=tensor([4.5583e-05, 4.4506e-05, 4.3115e-05, 3.8908e-05, 4.7113e-05, 3.8814e-05, + 4.0826e-05, 4.6348e-05], device='cuda:0') +2023-03-20 20:38:35,596 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-20 20:38:35,609 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-20 20:38:41,222 INFO [train.py:901] (0/2) Epoch 9, batch 100, loss[loss=0.2078, simple_loss=0.2708, pruned_loss=0.07236, over 7269.00 frames. ], tot_loss[loss=0.2024, simple_loss=0.2671, pruned_loss=0.06882, over 574454.35 frames. ], batch size: 70, lr: 1.80e-02, grad_scale: 16.0 +2023-03-20 20:38:44,926 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22700.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 20:38:48,162 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.01 vs. limit=2.0 +2023-03-20 20:38:49,872 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.903e+02 2.746e+02 3.605e+02 4.493e+02 9.955e+02, threshold=7.209e+02, percent-clipped=4.0 +2023-03-20 20:39:05,902 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=22741.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:39:06,846 INFO [train.py:901] (0/2) Epoch 9, batch 150, loss[loss=0.2028, simple_loss=0.2712, pruned_loss=0.06725, over 7293.00 frames. ], tot_loss[loss=0.1995, simple_loss=0.2649, pruned_loss=0.06701, over 767675.33 frames. ], batch size: 77, lr: 1.79e-02, grad_scale: 16.0 +2023-03-20 20:39:24,468 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2588, 2.0251, 2.0033, 3.4433, 1.5602, 2.8634, 1.1591, 3.2003], + device='cuda:0'), covar=tensor([0.0031, 0.0886, 0.1763, 0.0036, 0.3842, 0.0040, 0.1054, 0.0078], + device='cuda:0'), in_proj_covar=tensor([0.0134, 0.0264, 0.0313, 0.0128, 0.0307, 0.0138, 0.0272, 0.0173], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0003, 0.0003, 0.0001, 0.0003, 0.0001, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 20:39:28,519 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0483, 3.3111, 2.7542, 4.2600, 2.0894, 3.8595, 1.9526, 3.8306], + device='cuda:0'), covar=tensor([0.0055, 0.0545, 0.1483, 0.0036, 0.4124, 0.0054, 0.0930, 0.0088], + device='cuda:0'), in_proj_covar=tensor([0.0135, 0.0265, 0.0315, 0.0129, 0.0309, 0.0139, 0.0272, 0.0173], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0003, 0.0003, 0.0001, 0.0003, 0.0001, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 20:39:28,982 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=22786.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:39:32,391 INFO [train.py:901] (0/2) Epoch 9, batch 200, loss[loss=0.1992, simple_loss=0.2621, pruned_loss=0.06822, over 7260.00 frames. ], tot_loss[loss=0.1979, simple_loss=0.2627, pruned_loss=0.06658, over 914802.23 frames. ], batch size: 47, lr: 1.79e-02, grad_scale: 16.0 +2023-03-20 20:39:33,027 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=22794.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:39:38,436 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-20 20:39:41,344 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.696e+02 2.893e+02 3.607e+02 4.516e+02 8.593e+02, threshold=7.215e+02, percent-clipped=2.0 +2023-03-20 20:39:42,879 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-20 20:39:48,382 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-20 20:39:49,464 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=22825.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:39:49,994 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0104, 1.1956, 1.2022, 1.1908, 1.1985, 0.8956, 0.8547, 0.8912], + device='cuda:0'), covar=tensor([0.0176, 0.0091, 0.0174, 0.0065, 0.0084, 0.0132, 0.0190, 0.0178], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0017, 0.0016, 0.0017, 0.0019, 0.0018, 0.0019, 0.0021], + device='cuda:0'), out_proj_covar=tensor([2.3474e-05, 2.0244e-05, 2.1058e-05, 1.9835e-05, 2.2899e-05, 2.0968e-05, + 2.1925e-05, 2.8564e-05], device='cuda:0') +2023-03-20 20:39:52,933 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=22832.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:39:53,917 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=22834.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:39:58,393 INFO [train.py:901] (0/2) Epoch 9, batch 250, loss[loss=0.1963, simple_loss=0.2641, pruned_loss=0.06425, over 7306.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2635, pruned_loss=0.06701, over 1031125.63 frames. ], batch size: 49, lr: 1.79e-02, grad_scale: 16.0 +2023-03-20 20:40:00,524 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0955, 2.8705, 3.0024, 3.1670, 3.3286, 2.8446, 2.4805, 2.9941], + device='cuda:0'), covar=tensor([0.1080, 0.0566, 0.1050, 0.1488, 0.0916, 0.1078, 0.2487, 0.2015], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0030, 0.0030, 0.0030, 0.0028, 0.0028, 0.0040, 0.0028], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 20:40:01,939 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-20 20:40:04,557 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22855.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:40:10,618 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7238, 2.5064, 2.6622, 2.6384, 2.8015, 2.4543, 2.1947, 2.6979], + device='cuda:0'), covar=tensor([0.1566, 0.0677, 0.1617, 0.3120, 0.1200, 0.1696, 0.3215, 0.1716], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0030, 0.0030, 0.0030, 0.0028, 0.0027, 0.0039, 0.0028], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 20:40:13,185 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=22871.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:40:17,631 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=22880.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:40:21,391 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=22886.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:40:24,274 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-20 20:40:24,762 INFO [train.py:901] (0/2) Epoch 9, batch 300, loss[loss=0.1799, simple_loss=0.24, pruned_loss=0.0599, over 7231.00 frames. ], tot_loss[loss=0.1974, simple_loss=0.2624, pruned_loss=0.06616, over 1123730.66 frames. ], batch size: 45, lr: 1.79e-02, grad_scale: 16.0 +2023-03-20 20:40:24,864 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=22893.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:40:32,773 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.746e+02 2.565e+02 3.158e+02 3.848e+02 7.553e+02, threshold=6.316e+02, percent-clipped=2.0 +2023-03-20 20:40:33,309 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-20 20:40:37,815 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=22919.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:40:48,673 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=22941.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:40:49,617 INFO [train.py:901] (0/2) Epoch 9, batch 350, loss[loss=0.1922, simple_loss=0.2644, pruned_loss=0.05999, over 7291.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2623, pruned_loss=0.0661, over 1191683.68 frames. ], batch size: 70, lr: 1.79e-02, grad_scale: 16.0 +2023-03-20 20:41:07,129 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.06 vs. limit=2.0 +2023-03-20 20:41:07,858 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-20 20:41:15,825 INFO [train.py:901] (0/2) Epoch 9, batch 400, loss[loss=0.1949, simple_loss=0.2564, pruned_loss=0.0667, over 7253.00 frames. ], tot_loss[loss=0.1992, simple_loss=0.264, pruned_loss=0.0672, over 1247173.27 frames. ], batch size: 55, lr: 1.79e-02, grad_scale: 16.0 +2023-03-20 20:41:16,401 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2384, 4.1872, 4.2600, 4.5797, 4.6268, 4.5431, 3.9018, 4.0317], + device='cuda:0'), covar=tensor([0.0857, 0.2230, 0.1618, 0.0867, 0.0478, 0.1172, 0.0708, 0.1087], + device='cuda:0'), in_proj_covar=tensor([0.0159, 0.0258, 0.0212, 0.0196, 0.0154, 0.0261, 0.0145, 0.0175], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0001, 0.0002], + device='cuda:0') +2023-03-20 20:41:16,448 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5521, 3.4643, 3.3480, 3.5913, 3.5313, 3.5695, 3.4308, 3.2600], + device='cuda:0'), covar=tensor([0.0030, 0.0062, 0.0043, 0.0037, 0.0034, 0.0034, 0.0053, 0.0068], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0034, 0.0032, 0.0030, 0.0032, 0.0031, 0.0040, 0.0037], + device='cuda:0'), out_proj_covar=tensor([7.9271e-05, 1.0823e-04, 1.0235e-04, 8.6362e-05, 9.5174e-05, 8.9509e-05, + 1.2859e-04, 1.1018e-04], device='cuda:0') +2023-03-20 20:41:16,924 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=22995.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 20:41:18,890 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6903, 5.2060, 5.2196, 5.1485, 4.8865, 4.7594, 5.2403, 4.9910], + device='cuda:0'), covar=tensor([0.0405, 0.0341, 0.0345, 0.0367, 0.0336, 0.0245, 0.0318, 0.0483], + device='cuda:0'), in_proj_covar=tensor([0.0144, 0.0149, 0.0113, 0.0114, 0.0098, 0.0143, 0.0126, 0.0103], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 20:41:24,106 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.024e+02 2.885e+02 3.630e+02 4.346e+02 7.423e+02, threshold=7.259e+02, percent-clipped=3.0 +2023-03-20 20:41:25,312 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2110, 0.9385, 0.8854, 1.4820, 1.2643, 1.3137, 0.9855, 0.8989], + device='cuda:0'), covar=tensor([0.1024, 0.1403, 0.1238, 0.0542, 0.1015, 0.0843, 0.0602, 0.0941], + device='cuda:0'), in_proj_covar=tensor([0.0030, 0.0037, 0.0032, 0.0031, 0.0029, 0.0030, 0.0037, 0.0034], + device='cuda:0'), out_proj_covar=tensor([6.1476e-05, 8.1017e-05, 6.0636e-05, 6.0643e-05, 6.1580e-05, 6.2853e-05, + 7.4665e-05, 7.2744e-05], device='cuda:0') +2023-03-20 20:41:33,281 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4485, 2.5201, 2.9355, 2.4146, 2.5250, 2.2658, 2.1820, 2.6197], + device='cuda:0'), covar=tensor([0.2003, 0.0623, 0.0622, 0.2495, 0.1377, 0.2150, 0.2428, 0.1370], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0030, 0.0030, 0.0031, 0.0027, 0.0028, 0.0038, 0.0028], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 20:41:40,803 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=23041.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:41:41,723 INFO [train.py:901] (0/2) Epoch 9, batch 450, loss[loss=0.1975, simple_loss=0.259, pruned_loss=0.06805, over 7259.00 frames. ], tot_loss[loss=0.1988, simple_loss=0.2636, pruned_loss=0.06699, over 1289943.29 frames. ], batch size: 55, lr: 1.78e-02, grad_scale: 16.0 +2023-03-20 20:41:49,869 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-20 20:41:50,344 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-20 20:41:55,423 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7212, 3.7088, 3.5439, 3.6848, 3.5791, 3.9355, 3.6042, 3.5008], + device='cuda:0'), covar=tensor([0.0036, 0.0062, 0.0047, 0.0039, 0.0042, 0.0028, 0.0053, 0.0073], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0035, 0.0033, 0.0030, 0.0032, 0.0031, 0.0040, 0.0039], + device='cuda:0'), out_proj_covar=tensor([8.0419e-05, 1.0981e-04, 1.0387e-04, 8.7922e-05, 9.5995e-05, 9.0677e-05, + 1.2873e-04, 1.1410e-04], device='cuda:0') +2023-03-20 20:42:05,444 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=23089.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:42:07,477 INFO [train.py:901] (0/2) Epoch 9, batch 500, loss[loss=0.1956, simple_loss=0.2607, pruned_loss=0.06524, over 7275.00 frames. ], tot_loss[loss=0.1986, simple_loss=0.2637, pruned_loss=0.06675, over 1325942.69 frames. ], batch size: 70, lr: 1.78e-02, grad_scale: 16.0 +2023-03-20 20:42:11,482 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.92 vs. limit=2.0 +2023-03-20 20:42:12,862 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-03-20 20:42:15,276 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=23108.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:42:15,614 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.581e+02 2.594e+02 3.365e+02 3.863e+02 9.016e+02, threshold=6.731e+02, percent-clipped=1.0 +2023-03-20 20:42:22,222 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-20 20:42:23,713 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-20 20:42:24,232 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-20 20:42:24,599 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-03-20 20:42:26,767 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-20 20:42:31,282 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-20 20:42:33,278 INFO [train.py:901] (0/2) Epoch 9, batch 550, loss[loss=0.2171, simple_loss=0.2772, pruned_loss=0.07844, over 7322.00 frames. ], tot_loss[loss=0.1986, simple_loss=0.2638, pruned_loss=0.06675, over 1352298.14 frames. ], batch size: 49, lr: 1.78e-02, grad_scale: 16.0 +2023-03-20 20:42:37,724 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=23150.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:42:43,278 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-20 20:42:44,072 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.87 vs. limit=2.0 +2023-03-20 20:42:47,410 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=23169.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:42:51,216 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-20 20:42:51,820 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0267, 1.7047, 1.9987, 1.4606, 1.0252, 1.1445, 1.2260, 1.0316], + device='cuda:0'), covar=tensor([0.0285, 0.0234, 0.0135, 0.0129, 0.0594, 0.0310, 0.0153, 0.0311], + device='cuda:0'), in_proj_covar=tensor([0.0018, 0.0018, 0.0018, 0.0018, 0.0019, 0.0017, 0.0018, 0.0019], + device='cuda:0'), out_proj_covar=tensor([4.4564e-05, 4.2372e-05, 4.0956e-05, 3.8122e-05, 4.5564e-05, 3.9251e-05, + 4.1558e-05, 4.6159e-05], device='cuda:0') +2023-03-20 20:42:53,224 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=23181.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:42:54,697 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-20 20:42:55,001 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-03-20 20:42:57,036 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.10 vs. limit=2.0 +2023-03-20 20:42:59,276 INFO [train.py:901] (0/2) Epoch 9, batch 600, loss[loss=0.1533, simple_loss=0.2222, pruned_loss=0.04223, over 7339.00 frames. ], tot_loss[loss=0.1985, simple_loss=0.2636, pruned_loss=0.06672, over 1370831.78 frames. ], batch size: 44, lr: 1.78e-02, grad_scale: 16.0 +2023-03-20 20:43:01,816 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-20 20:43:02,440 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0287, 3.7656, 3.5353, 3.1601, 3.5148, 2.2282, 1.5590, 3.8643], + device='cuda:0'), covar=tensor([0.0017, 0.0033, 0.0077, 0.0091, 0.0066, 0.0451, 0.0667, 0.0047], + device='cuda:0'), in_proj_covar=tensor([0.0057, 0.0058, 0.0082, 0.0066, 0.0077, 0.0102, 0.0111, 0.0068], + device='cuda:0'), out_proj_covar=tensor([7.4430e-05, 9.0257e-05, 1.1913e-04, 9.8557e-05, 1.0704e-04, 1.4801e-04, + 1.5917e-04, 9.6810e-05], device='cuda:0') +2023-03-20 20:43:03,509 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.90 vs. limit=2.0 +2023-03-20 20:43:07,641 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.756e+02 2.964e+02 3.775e+02 5.004e+02 1.113e+03, threshold=7.551e+02, percent-clipped=10.0 +2023-03-20 20:43:12,386 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=23217.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:43:15,966 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8360, 3.6090, 3.1806, 3.5642, 2.8818, 2.7294, 3.7443, 3.0310], + device='cuda:0'), covar=tensor([0.0138, 0.0074, 0.0172, 0.0093, 0.0236, 0.0316, 0.0118, 0.0467], + device='cuda:0'), in_proj_covar=tensor([0.0251, 0.0235, 0.0236, 0.0234, 0.0291, 0.0296, 0.0252, 0.0300], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-20 20:43:18,335 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=23229.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:43:19,259 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-20 20:43:25,893 INFO [train.py:901] (0/2) Epoch 9, batch 650, loss[loss=0.2113, simple_loss=0.2775, pruned_loss=0.07261, over 7263.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2627, pruned_loss=0.06622, over 1384877.17 frames. ], batch size: 77, lr: 1.78e-02, grad_scale: 16.0 +2023-03-20 20:43:28,963 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-20 20:43:36,573 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1520, 4.6611, 4.6669, 4.6195, 4.5374, 4.1987, 4.7130, 4.5689], + device='cuda:0'), covar=tensor([0.0474, 0.0380, 0.0432, 0.0430, 0.0334, 0.0358, 0.0309, 0.0492], + device='cuda:0'), in_proj_covar=tensor([0.0149, 0.0156, 0.0119, 0.0117, 0.0101, 0.0149, 0.0131, 0.0104], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 20:43:40,675 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0840, 1.2679, 1.4228, 1.2614, 1.4999, 0.9853, 0.9489, 0.8190], + device='cuda:0'), covar=tensor([0.0200, 0.0177, 0.0258, 0.0179, 0.0156, 0.0189, 0.0119, 0.0378], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0017, 0.0016, 0.0019, 0.0020, 0.0018, 0.0018, 0.0022], + device='cuda:0'), out_proj_covar=tensor([2.4604e-05, 1.9996e-05, 2.1253e-05, 2.1283e-05, 2.3507e-05, 2.0734e-05, + 2.0935e-05, 2.8981e-05], device='cuda:0') +2023-03-20 20:43:43,630 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=23278.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:43:45,948 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-20 20:43:49,547 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=23290.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:43:50,903 INFO [train.py:901] (0/2) Epoch 9, batch 700, loss[loss=0.229, simple_loss=0.2866, pruned_loss=0.08574, over 7324.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2624, pruned_loss=0.06593, over 1398118.07 frames. ], batch size: 75, lr: 1.77e-02, grad_scale: 16.0 +2023-03-20 20:43:51,979 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=23295.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:43:53,906 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-20 20:43:55,720 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.94 vs. limit=2.0 +2023-03-20 20:43:59,403 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.653e+02 2.801e+02 3.274e+02 4.080e+02 8.180e+02, threshold=6.549e+02, percent-clipped=2.0 +2023-03-20 20:44:17,135 INFO [train.py:901] (0/2) Epoch 9, batch 750, loss[loss=0.1823, simple_loss=0.2464, pruned_loss=0.05912, over 7279.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.263, pruned_loss=0.06609, over 1408018.40 frames. ], batch size: 47, lr: 1.77e-02, grad_scale: 16.0 +2023-03-20 20:44:17,186 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=23343.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:44:18,157 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-20 20:44:18,681 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-20 20:44:31,157 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.82 vs. limit=2.0 +2023-03-20 20:44:31,763 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-20 20:44:35,836 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-20 20:44:42,387 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-20 20:44:42,840 INFO [train.py:901] (0/2) Epoch 9, batch 800, loss[loss=0.2039, simple_loss=0.2696, pruned_loss=0.06915, over 7286.00 frames. ], tot_loss[loss=0.1989, simple_loss=0.2644, pruned_loss=0.06672, over 1416481.24 frames. ], batch size: 86, lr: 1.77e-02, grad_scale: 16.0 +2023-03-20 20:44:43,353 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-20 20:44:44,508 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5016, 3.3232, 2.8837, 3.2719, 2.6142, 2.4270, 3.3807, 2.7648], + device='cuda:0'), covar=tensor([0.0108, 0.0119, 0.0206, 0.0098, 0.0244, 0.0386, 0.0280, 0.0389], + device='cuda:0'), in_proj_covar=tensor([0.0248, 0.0233, 0.0230, 0.0231, 0.0282, 0.0290, 0.0249, 0.0288], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-20 20:44:50,176 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.88 vs. limit=2.0 +2023-03-20 20:44:51,712 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.583e+02 2.882e+02 3.186e+02 4.037e+02 8.134e+02, threshold=6.373e+02, percent-clipped=1.0 +2023-03-20 20:44:54,170 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-20 20:45:08,679 INFO [train.py:901] (0/2) Epoch 9, batch 850, loss[loss=0.2183, simple_loss=0.2858, pruned_loss=0.07538, over 7302.00 frames. ], tot_loss[loss=0.1984, simple_loss=0.2637, pruned_loss=0.06656, over 1421519.81 frames. ], batch size: 80, lr: 1.77e-02, grad_scale: 16.0 +2023-03-20 20:45:09,350 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0631, 1.0165, 1.0585, 1.3634, 1.1379, 1.1441, 0.8715, 1.0815], + device='cuda:0'), covar=tensor([0.1270, 0.1438, 0.1051, 0.0502, 0.0964, 0.2097, 0.0534, 0.1019], + device='cuda:0'), in_proj_covar=tensor([0.0029, 0.0037, 0.0030, 0.0030, 0.0027, 0.0030, 0.0038, 0.0034], + device='cuda:0'), out_proj_covar=tensor([6.0691e-05, 8.1401e-05, 5.8466e-05, 5.8429e-05, 5.9574e-05, 6.2296e-05, + 7.5560e-05, 7.2788e-05], device='cuda:0') +2023-03-20 20:45:12,193 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=23450.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:45:12,607 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-20 20:45:12,617 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-20 20:45:17,761 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-20 20:45:19,321 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=23464.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:45:21,327 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-20 20:45:28,451 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=23481.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:45:32,578 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=23488.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:45:34,967 INFO [train.py:901] (0/2) Epoch 9, batch 900, loss[loss=0.1624, simple_loss=0.2305, pruned_loss=0.04714, over 7139.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.263, pruned_loss=0.06563, over 1428166.88 frames. ], batch size: 41, lr: 1.77e-02, grad_scale: 16.0 +2023-03-20 20:45:37,497 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=23498.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:45:39,508 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0324, 4.5310, 4.4817, 4.3941, 4.4539, 4.1246, 4.5227, 4.4219], + device='cuda:0'), covar=tensor([0.0401, 0.0380, 0.0520, 0.0498, 0.0302, 0.0300, 0.0379, 0.0466], + device='cuda:0'), in_proj_covar=tensor([0.0144, 0.0150, 0.0117, 0.0115, 0.0099, 0.0144, 0.0126, 0.0103], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 20:45:42,902 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.727e+02 2.905e+02 3.478e+02 4.319e+02 1.982e+03, threshold=6.955e+02, percent-clipped=5.0 +2023-03-20 20:45:52,944 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=23529.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:45:57,336 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.00 vs. limit=2.0 +2023-03-20 20:45:59,558 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-20 20:46:00,027 INFO [train.py:901] (0/2) Epoch 9, batch 950, loss[loss=0.1949, simple_loss=0.2579, pruned_loss=0.06598, over 7272.00 frames. ], tot_loss[loss=0.1975, simple_loss=0.2632, pruned_loss=0.0659, over 1430322.67 frames. ], batch size: 70, lr: 1.77e-02, grad_scale: 16.0 +2023-03-20 20:46:03,140 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=23549.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 20:46:16,339 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=23573.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:46:22,300 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=23585.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:46:22,917 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4189, 3.3226, 2.5392, 3.2714, 2.4970, 2.2463, 3.4298, 2.6168], + device='cuda:0'), covar=tensor([0.0156, 0.0145, 0.0317, 0.0119, 0.0297, 0.0499, 0.0228, 0.0654], + device='cuda:0'), in_proj_covar=tensor([0.0254, 0.0237, 0.0233, 0.0236, 0.0282, 0.0289, 0.0251, 0.0291], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-20 20:46:23,728 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-20 20:46:26,204 INFO [train.py:901] (0/2) Epoch 9, batch 1000, loss[loss=0.1596, simple_loss=0.2048, pruned_loss=0.05724, over 6110.00 frames. ], tot_loss[loss=0.1971, simple_loss=0.2627, pruned_loss=0.06574, over 1430961.44 frames. ], batch size: 26, lr: 1.76e-02, grad_scale: 16.0 +2023-03-20 20:46:33,737 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.94 vs. limit=2.0 +2023-03-20 20:46:34,371 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.714e+02 2.956e+02 3.607e+02 4.616e+02 9.211e+02, threshold=7.213e+02, percent-clipped=2.0 +2023-03-20 20:46:41,990 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5038, 3.1928, 3.2727, 3.1781, 3.4404, 3.4047, 3.4076, 3.2747], + device='cuda:0'), covar=tensor([0.0027, 0.0080, 0.0054, 0.0053, 0.0045, 0.0041, 0.0058, 0.0064], + device='cuda:0'), in_proj_covar=tensor([0.0029, 0.0036, 0.0034, 0.0032, 0.0034, 0.0033, 0.0042, 0.0040], + device='cuda:0'), out_proj_covar=tensor([8.1427e-05, 1.1432e-04, 1.1092e-04, 8.9954e-05, 1.0116e-04, 9.5429e-05, + 1.3428e-04, 1.1926e-04], device='cuda:0') +2023-03-20 20:46:43,865 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-20 20:46:51,201 INFO [train.py:901] (0/2) Epoch 9, batch 1050, loss[loss=0.1991, simple_loss=0.2636, pruned_loss=0.06737, over 7256.00 frames. ], tot_loss[loss=0.1976, simple_loss=0.2634, pruned_loss=0.06595, over 1434811.76 frames. ], batch size: 64, lr: 1.76e-02, grad_scale: 16.0 +2023-03-20 20:46:53,852 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6982, 2.4800, 2.3058, 2.1364, 2.5847, 1.5737, 2.3644, 2.2503], + device='cuda:0'), covar=tensor([0.0670, 0.1174, 0.0787, 0.1066, 0.0535, 0.1155, 0.0635, 0.1029], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0033, 0.0038, 0.0035, 0.0034, 0.0035, 0.0032, 0.0032], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 20:47:05,630 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-20 20:47:09,645 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-20 20:47:17,726 INFO [train.py:901] (0/2) Epoch 9, batch 1100, loss[loss=0.2083, simple_loss=0.2685, pruned_loss=0.07402, over 7259.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2623, pruned_loss=0.06513, over 1437289.16 frames. ], batch size: 47, lr: 1.76e-02, grad_scale: 16.0 +2023-03-20 20:47:18,031 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-03-20 20:47:25,697 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.755e+02 2.952e+02 3.526e+02 4.290e+02 1.100e+03, threshold=7.053e+02, percent-clipped=3.0 +2023-03-20 20:47:38,354 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-20 20:47:38,851 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 20:47:43,514 INFO [train.py:901] (0/2) Epoch 9, batch 1150, loss[loss=0.2134, simple_loss=0.2783, pruned_loss=0.07426, over 6634.00 frames. ], tot_loss[loss=0.1967, simple_loss=0.2626, pruned_loss=0.06535, over 1437692.31 frames. ], batch size: 107, lr: 1.76e-02, grad_scale: 16.0 +2023-03-20 20:47:52,146 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-20 20:47:52,655 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-20 20:47:54,738 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=23764.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:47:57,168 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2005, 4.1443, 4.1300, 4.5581, 4.5742, 4.5789, 4.1345, 4.0488], + device='cuda:0'), covar=tensor([0.0978, 0.2290, 0.1741, 0.0969, 0.0552, 0.1235, 0.0631, 0.0950], + device='cuda:0'), in_proj_covar=tensor([0.0161, 0.0256, 0.0221, 0.0206, 0.0159, 0.0268, 0.0148, 0.0179], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0001, 0.0002], + device='cuda:0') +2023-03-20 20:48:09,161 INFO [train.py:901] (0/2) Epoch 9, batch 1200, loss[loss=0.2339, simple_loss=0.2913, pruned_loss=0.08832, over 6746.00 frames. ], tot_loss[loss=0.1972, simple_loss=0.2629, pruned_loss=0.06572, over 1438619.81 frames. ], batch size: 107, lr: 1.76e-02, grad_scale: 16.0 +2023-03-20 20:48:17,482 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.614e+02 2.908e+02 3.769e+02 4.878e+02 9.348e+02, threshold=7.537e+02, percent-clipped=4.0 +2023-03-20 20:48:19,036 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=23812.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:48:24,142 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-20 20:48:35,828 INFO [train.py:901] (0/2) Epoch 9, batch 1250, loss[loss=0.195, simple_loss=0.2602, pruned_loss=0.06492, over 7348.00 frames. ], tot_loss[loss=0.1959, simple_loss=0.262, pruned_loss=0.06487, over 1440308.07 frames. ], batch size: 63, lr: 1.75e-02, grad_scale: 16.0 +2023-03-20 20:48:36,406 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=23844.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 20:48:45,402 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6442, 3.4463, 2.8594, 3.4162, 2.6881, 2.4638, 3.4888, 2.8516], + device='cuda:0'), covar=tensor([0.0139, 0.0128, 0.0249, 0.0085, 0.0284, 0.0395, 0.0178, 0.0503], + device='cuda:0'), in_proj_covar=tensor([0.0260, 0.0246, 0.0236, 0.0237, 0.0288, 0.0292, 0.0252, 0.0292], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-20 20:48:48,732 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-20 20:48:50,931 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=23873.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:48:53,301 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-20 20:48:54,159 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.97 vs. limit=2.0 +2023-03-20 20:48:54,833 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-20 20:48:56,961 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=23885.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:49:00,816 INFO [train.py:901] (0/2) Epoch 9, batch 1300, loss[loss=0.2206, simple_loss=0.282, pruned_loss=0.07962, over 7279.00 frames. ], tot_loss[loss=0.1963, simple_loss=0.2624, pruned_loss=0.06511, over 1443861.40 frames. ], batch size: 77, lr: 1.75e-02, grad_scale: 16.0 +2023-03-20 20:49:09,406 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.638e+02 2.579e+02 3.213e+02 4.106e+02 9.260e+02, threshold=6.425e+02, percent-clipped=2.0 +2023-03-20 20:49:14,138 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1686, 1.2274, 1.4550, 1.9471, 1.5456, 1.4626, 1.2884, 1.4211], + device='cuda:0'), covar=tensor([0.1960, 0.2629, 0.1137, 0.0964, 0.2697, 0.2699, 0.1259, 0.4093], + device='cuda:0'), in_proj_covar=tensor([0.0030, 0.0038, 0.0031, 0.0031, 0.0031, 0.0031, 0.0039, 0.0035], + device='cuda:0'), out_proj_covar=tensor([6.3582e-05, 8.4286e-05, 6.1802e-05, 6.1244e-05, 6.6800e-05, 6.6440e-05, + 7.9986e-05, 7.4329e-05], device='cuda:0') +2023-03-20 20:49:15,604 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=23921.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:49:17,534 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-20 20:49:20,741 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-20 20:49:22,275 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=23933.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:49:23,443 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2921, 3.9748, 3.5975, 4.0590, 3.2132, 3.0022, 4.0672, 3.4046], + device='cuda:0'), covar=tensor([0.0102, 0.0126, 0.0174, 0.0088, 0.0247, 0.0316, 0.0189, 0.0397], + device='cuda:0'), in_proj_covar=tensor([0.0259, 0.0249, 0.0237, 0.0241, 0.0291, 0.0296, 0.0254, 0.0298], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-20 20:49:24,252 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-20 20:49:27,210 INFO [train.py:901] (0/2) Epoch 9, batch 1350, loss[loss=0.2279, simple_loss=0.289, pruned_loss=0.08338, over 7319.00 frames. ], tot_loss[loss=0.1957, simple_loss=0.2616, pruned_loss=0.06488, over 1440061.85 frames. ], batch size: 59, lr: 1.75e-02, grad_scale: 32.0 +2023-03-20 20:49:34,686 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-20 20:49:52,683 INFO [train.py:901] (0/2) Epoch 9, batch 1400, loss[loss=0.172, simple_loss=0.2413, pruned_loss=0.05136, over 7312.00 frames. ], tot_loss[loss=0.1943, simple_loss=0.2605, pruned_loss=0.06408, over 1439714.71 frames. ], batch size: 44, lr: 1.75e-02, grad_scale: 32.0 +2023-03-20 20:49:56,574 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-24000.pt +2023-03-20 20:50:04,556 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.640e+02 2.794e+02 3.299e+02 4.063e+02 9.313e+02, threshold=6.598e+02, percent-clipped=3.0 +2023-03-20 20:50:10,224 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-20 20:50:19,386 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=24037.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:50:22,310 INFO [train.py:901] (0/2) Epoch 9, batch 1450, loss[loss=0.2288, simple_loss=0.289, pruned_loss=0.08431, over 7101.00 frames. ], tot_loss[loss=0.1962, simple_loss=0.2625, pruned_loss=0.0649, over 1442049.67 frames. ], batch size: 98, lr: 1.75e-02, grad_scale: 32.0 +2023-03-20 20:50:28,406 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.1840, 4.7940, 4.8814, 5.2208, 5.2493, 5.3045, 4.2652, 4.8822], + device='cuda:0'), covar=tensor([0.0689, 0.2311, 0.1875, 0.0955, 0.0617, 0.1019, 0.0705, 0.0851], + device='cuda:0'), in_proj_covar=tensor([0.0162, 0.0262, 0.0220, 0.0201, 0.0159, 0.0267, 0.0152, 0.0180], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0001, 0.0002], + device='cuda:0') +2023-03-20 20:50:33,907 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-20 20:50:47,921 INFO [train.py:901] (0/2) Epoch 9, batch 1500, loss[loss=0.174, simple_loss=0.2353, pruned_loss=0.05632, over 7209.00 frames. ], tot_loss[loss=0.1962, simple_loss=0.2625, pruned_loss=0.06499, over 1439319.84 frames. ], batch size: 39, lr: 1.75e-02, grad_scale: 32.0 +2023-03-20 20:50:50,404 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-20 20:50:50,526 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=24098.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:50:53,652 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7098, 3.4168, 3.2232, 3.4054, 2.7384, 2.5301, 3.5905, 2.9027], + device='cuda:0'), covar=tensor([0.0127, 0.0130, 0.0152, 0.0097, 0.0218, 0.0345, 0.0193, 0.0376], + device='cuda:0'), in_proj_covar=tensor([0.0250, 0.0246, 0.0229, 0.0236, 0.0286, 0.0288, 0.0253, 0.0288], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-20 20:50:56,464 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.802e+02 2.774e+02 3.566e+02 4.390e+02 1.132e+03, threshold=7.132e+02, percent-clipped=5.0 +2023-03-20 20:51:12,415 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.02 vs. limit=5.0 +2023-03-20 20:51:13,480 INFO [train.py:901] (0/2) Epoch 9, batch 1550, loss[loss=0.156, simple_loss=0.2075, pruned_loss=0.05224, over 6218.00 frames. ], tot_loss[loss=0.1957, simple_loss=0.2621, pruned_loss=0.06467, over 1440849.70 frames. ], batch size: 27, lr: 1.74e-02, grad_scale: 32.0 +2023-03-20 20:51:14,098 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6458, 3.7863, 3.4536, 3.6944, 3.2905, 3.6601, 3.9210, 3.9651], + device='cuda:0'), covar=tensor([0.0202, 0.0137, 0.0241, 0.0191, 0.0516, 0.0219, 0.0209, 0.0177], + device='cuda:0'), in_proj_covar=tensor([0.0093, 0.0093, 0.0088, 0.0102, 0.0097, 0.0075, 0.0076, 0.0077], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 20:51:14,121 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=24144.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 20:51:15,029 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-20 20:51:22,296 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-20 20:51:24,708 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.9452, 0.8653, 0.9516, 1.3293, 1.1846, 1.1449, 1.0762, 1.0166], + device='cuda:0'), covar=tensor([0.1134, 0.1322, 0.1395, 0.0687, 0.1100, 0.1482, 0.0830, 0.1245], + device='cuda:0'), in_proj_covar=tensor([0.0031, 0.0038, 0.0032, 0.0030, 0.0031, 0.0031, 0.0038, 0.0034], + device='cuda:0'), out_proj_covar=tensor([6.4298e-05, 8.4192e-05, 6.2864e-05, 6.0682e-05, 6.6603e-05, 6.6681e-05, + 7.9232e-05, 7.3769e-05], device='cuda:0') +2023-03-20 20:51:39,448 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=24192.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:51:39,895 INFO [train.py:901] (0/2) Epoch 9, batch 1600, loss[loss=0.2139, simple_loss=0.2767, pruned_loss=0.07557, over 7261.00 frames. ], tot_loss[loss=0.1951, simple_loss=0.2615, pruned_loss=0.06435, over 1440486.55 frames. ], batch size: 55, lr: 1.74e-02, grad_scale: 32.0 +2023-03-20 20:51:46,817 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-20 20:51:47,838 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-20 20:51:48,306 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.692e+02 2.929e+02 3.453e+02 4.111e+02 6.914e+02, threshold=6.906e+02, percent-clipped=0.0 +2023-03-20 20:51:50,853 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-20 20:51:59,372 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-20 20:52:03,422 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-20 20:52:05,434 INFO [train.py:901] (0/2) Epoch 9, batch 1650, loss[loss=0.1846, simple_loss=0.2568, pruned_loss=0.05621, over 7247.00 frames. ], tot_loss[loss=0.195, simple_loss=0.2614, pruned_loss=0.0643, over 1441361.98 frames. ], batch size: 89, lr: 1.74e-02, grad_scale: 32.0 +2023-03-20 20:52:08,849 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.05 vs. limit=2.0 +2023-03-20 20:52:12,804 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-20 20:52:13,663 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.90 vs. limit=2.0 +2023-03-20 20:52:13,899 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5721, 5.1251, 5.1127, 4.9900, 4.8480, 4.6703, 5.1352, 4.8599], + device='cuda:0'), covar=tensor([0.0413, 0.0345, 0.0406, 0.0558, 0.0361, 0.0288, 0.0381, 0.0476], + device='cuda:0'), in_proj_covar=tensor([0.0145, 0.0155, 0.0117, 0.0116, 0.0103, 0.0147, 0.0130, 0.0108], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 20:52:15,748 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.88 vs. limit=2.0 +2023-03-20 20:52:21,506 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.9356, 0.8457, 1.0843, 1.3330, 1.2789, 1.3365, 1.0727, 1.1327], + device='cuda:0'), covar=tensor([0.0922, 0.2200, 0.1324, 0.0745, 0.1245, 0.1186, 0.1011, 0.2050], + device='cuda:0'), in_proj_covar=tensor([0.0030, 0.0037, 0.0031, 0.0030, 0.0030, 0.0029, 0.0036, 0.0034], + device='cuda:0'), out_proj_covar=tensor([6.2580e-05, 8.2559e-05, 6.1931e-05, 5.9716e-05, 6.4316e-05, 6.3727e-05, + 7.5727e-05, 7.3535e-05], device='cuda:0') +2023-03-20 20:52:30,082 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 20:52:32,056 INFO [train.py:901] (0/2) Epoch 9, batch 1700, loss[loss=0.1856, simple_loss=0.2597, pruned_loss=0.05571, over 7360.00 frames. ], tot_loss[loss=0.1949, simple_loss=0.2615, pruned_loss=0.0641, over 1444024.55 frames. ], batch size: 63, lr: 1.74e-02, grad_scale: 32.0 +2023-03-20 20:52:34,100 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-20 20:52:34,197 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7425, 3.3516, 3.5250, 3.3213, 3.2727, 3.4055, 3.6173, 3.2657], + device='cuda:0'), covar=tensor([0.0089, 0.0168, 0.0125, 0.0157, 0.0224, 0.0125, 0.0146, 0.0156], + device='cuda:0'), in_proj_covar=tensor([0.0045, 0.0054, 0.0055, 0.0044, 0.0074, 0.0059, 0.0054, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 20:52:40,052 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.595e+02 2.835e+02 3.245e+02 4.182e+02 9.233e+02, threshold=6.490e+02, percent-clipped=3.0 +2023-03-20 20:52:44,312 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.08 vs. limit=2.0 +2023-03-20 20:52:44,519 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-20 20:52:57,653 INFO [train.py:901] (0/2) Epoch 9, batch 1750, loss[loss=0.1915, simple_loss=0.258, pruned_loss=0.06248, over 7284.00 frames. ], tot_loss[loss=0.1948, simple_loss=0.2614, pruned_loss=0.06408, over 1445009.93 frames. ], batch size: 66, lr: 1.74e-02, grad_scale: 32.0 +2023-03-20 20:53:01,339 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6827, 2.4420, 2.0447, 3.2273, 1.4747, 3.3488, 1.1606, 3.1115], + device='cuda:0'), covar=tensor([0.0045, 0.0800, 0.1985, 0.0046, 0.4742, 0.0060, 0.1244, 0.0074], + device='cuda:0'), in_proj_covar=tensor([0.0132, 0.0264, 0.0317, 0.0138, 0.0307, 0.0140, 0.0269, 0.0175], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0003, 0.0003, 0.0001, 0.0003, 0.0001, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 20:53:05,595 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.05 vs. limit=2.0 +2023-03-20 20:53:08,774 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-20 20:53:10,407 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-20 20:53:20,028 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9750, 3.5386, 3.6271, 3.4922, 3.4795, 3.6644, 3.8491, 3.3810], + device='cuda:0'), covar=tensor([0.0081, 0.0145, 0.0154, 0.0143, 0.0201, 0.0100, 0.0132, 0.0154], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0052, 0.0055, 0.0043, 0.0072, 0.0057, 0.0053, 0.0051], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 20:53:23,362 INFO [train.py:901] (0/2) Epoch 9, batch 1800, loss[loss=0.1832, simple_loss=0.2524, pruned_loss=0.05703, over 7350.00 frames. ], tot_loss[loss=0.1948, simple_loss=0.2616, pruned_loss=0.06402, over 1447195.44 frames. ], batch size: 73, lr: 1.74e-02, grad_scale: 32.0 +2023-03-20 20:53:23,427 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=24393.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:53:28,896 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9487, 2.3851, 1.8345, 3.3116, 2.8375, 2.8781, 2.8444, 3.0645], + device='cuda:0'), covar=tensor([0.1521, 0.0591, 0.2062, 0.0337, 0.0049, 0.0082, 0.0073, 0.0055], + device='cuda:0'), in_proj_covar=tensor([0.0290, 0.0219, 0.0273, 0.0236, 0.0120, 0.0121, 0.0126, 0.0141], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 20:53:30,990 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.54 vs. limit=2.0 +2023-03-20 20:53:31,278 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-20 20:53:31,369 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2997, 3.9940, 4.1347, 3.8899, 3.9323, 4.1470, 4.4378, 3.7464], + device='cuda:0'), covar=tensor([0.0135, 0.0107, 0.0125, 0.0109, 0.0159, 0.0076, 0.0103, 0.0148], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0052, 0.0054, 0.0043, 0.0072, 0.0057, 0.0052, 0.0051], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 20:53:31,746 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.774e+02 2.625e+02 3.149e+02 3.948e+02 8.480e+02, threshold=6.298e+02, percent-clipped=3.0 +2023-03-20 20:53:34,417 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4795, 2.3105, 1.9900, 2.2617, 2.5066, 2.1563, 2.2796, 2.5176], + device='cuda:0'), covar=tensor([0.0648, 0.0712, 0.1238, 0.1418, 0.1054, 0.0770, 0.0812, 0.1040], + device='cuda:0'), in_proj_covar=tensor([0.0036, 0.0032, 0.0038, 0.0035, 0.0035, 0.0034, 0.0034, 0.0034], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 20:53:44,876 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-20 20:53:49,369 INFO [train.py:901] (0/2) Epoch 9, batch 1850, loss[loss=0.1798, simple_loss=0.2519, pruned_loss=0.0538, over 7298.00 frames. ], tot_loss[loss=0.1957, simple_loss=0.2622, pruned_loss=0.0646, over 1447160.73 frames. ], batch size: 80, lr: 1.73e-02, grad_scale: 16.0 +2023-03-20 20:53:56,090 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-20 20:54:12,113 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-20 20:54:15,086 INFO [train.py:901] (0/2) Epoch 9, batch 1900, loss[loss=0.1985, simple_loss=0.2695, pruned_loss=0.06377, over 7258.00 frames. ], tot_loss[loss=0.1952, simple_loss=0.2615, pruned_loss=0.06441, over 1442542.97 frames. ], batch size: 70, lr: 1.73e-02, grad_scale: 16.0 +2023-03-20 20:54:23,576 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.695e+02 2.705e+02 3.277e+02 3.704e+02 6.252e+02, threshold=6.553e+02, percent-clipped=0.0 +2023-03-20 20:54:37,808 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-20 20:54:41,450 INFO [train.py:901] (0/2) Epoch 9, batch 1950, loss[loss=0.2019, simple_loss=0.2738, pruned_loss=0.06498, over 7323.00 frames. ], tot_loss[loss=0.1941, simple_loss=0.2606, pruned_loss=0.06386, over 1441491.98 frames. ], batch size: 75, lr: 1.73e-02, grad_scale: 16.0 +2023-03-20 20:54:43,591 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=24547.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:54:49,143 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-20 20:54:53,698 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-20 20:54:54,188 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-20 20:54:54,777 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=24569.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:55:06,814 INFO [train.py:901] (0/2) Epoch 9, batch 2000, loss[loss=0.1785, simple_loss=0.2499, pruned_loss=0.05351, over 7348.00 frames. ], tot_loss[loss=0.1943, simple_loss=0.2608, pruned_loss=0.06387, over 1439444.34 frames. ], batch size: 61, lr: 1.73e-02, grad_scale: 16.0 +2023-03-20 20:55:10,694 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-20 20:55:15,442 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=24608.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:55:16,213 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.744e+02 2.663e+02 3.037e+02 3.690e+02 7.050e+02, threshold=6.074e+02, percent-clipped=1.0 +2023-03-20 20:55:18,502 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.89 vs. limit=2.0 +2023-03-20 20:55:22,190 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-20 20:55:27,053 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=24630.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:55:30,421 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-20 20:55:33,470 INFO [train.py:901] (0/2) Epoch 9, batch 2050, loss[loss=0.187, simple_loss=0.2618, pruned_loss=0.05613, over 7283.00 frames. ], tot_loss[loss=0.1947, simple_loss=0.261, pruned_loss=0.06422, over 1439709.94 frames. ], batch size: 57, lr: 1.73e-02, grad_scale: 16.0 +2023-03-20 20:55:52,699 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.52 vs. limit=5.0 +2023-03-20 20:55:55,209 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.01 vs. limit=5.0 +2023-03-20 20:55:58,547 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6041, 3.4479, 3.5265, 3.7655, 3.6850, 3.6542, 3.6037, 3.4940], + device='cuda:0'), covar=tensor([0.0037, 0.0074, 0.0042, 0.0034, 0.0032, 0.0038, 0.0052, 0.0060], + device='cuda:0'), in_proj_covar=tensor([0.0030, 0.0036, 0.0035, 0.0033, 0.0034, 0.0034, 0.0044, 0.0041], + device='cuda:0'), out_proj_covar=tensor([8.4297e-05, 1.1196e-04, 1.0920e-04, 9.2185e-05, 9.9642e-05, 9.7533e-05, + 1.3783e-04, 1.1933e-04], device='cuda:0') +2023-03-20 20:55:58,944 INFO [train.py:901] (0/2) Epoch 9, batch 2100, loss[loss=0.1827, simple_loss=0.262, pruned_loss=0.05167, over 7361.00 frames. ], tot_loss[loss=0.1937, simple_loss=0.2599, pruned_loss=0.06378, over 1439591.75 frames. ], batch size: 63, lr: 1.73e-02, grad_scale: 16.0 +2023-03-20 20:55:59,053 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=24693.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:56:03,921 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-20 20:56:07,390 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.780e+02 2.739e+02 3.522e+02 4.404e+02 7.179e+02, threshold=7.043e+02, percent-clipped=4.0 +2023-03-20 20:56:07,432 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-20 20:56:10,859 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.02 vs. limit=2.0 +2023-03-20 20:56:23,704 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=24741.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:56:24,612 INFO [train.py:901] (0/2) Epoch 9, batch 2150, loss[loss=0.1913, simple_loss=0.2575, pruned_loss=0.06251, over 7326.00 frames. ], tot_loss[loss=0.1951, simple_loss=0.2615, pruned_loss=0.06431, over 1441361.92 frames. ], batch size: 44, lr: 1.72e-02, grad_scale: 16.0 +2023-03-20 20:56:35,898 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=24765.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:56:38,427 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8344, 2.2639, 2.7570, 2.7223, 3.0331, 2.5827, 2.2021, 2.9217], + device='cuda:0'), covar=tensor([0.0911, 0.0521, 0.1201, 0.1461, 0.0695, 0.1382, 0.2902, 0.1043], + device='cuda:0'), in_proj_covar=tensor([0.0029, 0.0030, 0.0029, 0.0031, 0.0026, 0.0027, 0.0039, 0.0029], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 20:56:50,676 INFO [train.py:901] (0/2) Epoch 9, batch 2200, loss[loss=0.1703, simple_loss=0.2401, pruned_loss=0.05024, over 7212.00 frames. ], tot_loss[loss=0.1945, simple_loss=0.2613, pruned_loss=0.06388, over 1439952.14 frames. ], batch size: 45, lr: 1.72e-02, grad_scale: 16.0 +2023-03-20 20:56:52,714 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-20 20:57:00,151 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.794e+02 2.770e+02 3.163e+02 3.850e+02 5.609e+02, threshold=6.326e+02, percent-clipped=0.0 +2023-03-20 20:57:01,866 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4020, 3.1197, 3.4537, 3.3053, 3.7550, 3.3693, 2.9808, 3.6716], + device='cuda:0'), covar=tensor([0.1739, 0.0526, 0.1250, 0.2396, 0.0621, 0.2076, 0.2329, 0.1063], + device='cuda:0'), in_proj_covar=tensor([0.0029, 0.0030, 0.0029, 0.0031, 0.0026, 0.0027, 0.0038, 0.0029], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 20:57:08,475 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=24826.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:57:13,099 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8977, 3.6903, 3.4811, 3.5879, 2.7291, 2.5361, 3.8811, 2.8664], + device='cuda:0'), covar=tensor([0.0123, 0.0099, 0.0130, 0.0104, 0.0247, 0.0367, 0.0145, 0.0459], + device='cuda:0'), in_proj_covar=tensor([0.0248, 0.0246, 0.0228, 0.0240, 0.0288, 0.0286, 0.0251, 0.0293], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-20 20:57:13,482 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8550, 3.8853, 3.9193, 4.3169, 4.3185, 4.2780, 3.7386, 3.7911], + device='cuda:0'), covar=tensor([0.1030, 0.2430, 0.2016, 0.0751, 0.0597, 0.1248, 0.0773, 0.1030], + device='cuda:0'), in_proj_covar=tensor([0.0161, 0.0260, 0.0218, 0.0194, 0.0157, 0.0264, 0.0148, 0.0177], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0001, 0.0002], + device='cuda:0') +2023-03-20 20:57:14,675 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.03 vs. limit=2.0 +2023-03-20 20:57:16,939 INFO [train.py:901] (0/2) Epoch 9, batch 2250, loss[loss=0.1945, simple_loss=0.2635, pruned_loss=0.06276, over 7325.00 frames. ], tot_loss[loss=0.1943, simple_loss=0.2609, pruned_loss=0.06391, over 1438673.10 frames. ], batch size: 54, lr: 1.72e-02, grad_scale: 16.0 +2023-03-20 20:57:19,713 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=24848.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:57:22,765 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3182, 2.4232, 2.0391, 2.2533, 2.3901, 2.1222, 2.1861, 2.4301], + device='cuda:0'), covar=tensor([0.0385, 0.0300, 0.0653, 0.0410, 0.0625, 0.0510, 0.0700, 0.0431], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0033, 0.0039, 0.0035, 0.0037, 0.0034, 0.0034, 0.0034], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 20:57:27,755 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-20 20:57:27,766 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-20 20:57:37,007 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7465, 3.6572, 3.3116, 3.4993, 2.8277, 2.6306, 3.7954, 2.8479], + device='cuda:0'), covar=tensor([0.0098, 0.0099, 0.0159, 0.0126, 0.0259, 0.0335, 0.0210, 0.0494], + device='cuda:0'), in_proj_covar=tensor([0.0247, 0.0246, 0.0229, 0.0240, 0.0288, 0.0283, 0.0250, 0.0292], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-20 20:57:40,463 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-20 20:57:43,438 INFO [train.py:901] (0/2) Epoch 9, batch 2300, loss[loss=0.2103, simple_loss=0.2659, pruned_loss=0.07732, over 7227.00 frames. ], tot_loss[loss=0.1948, simple_loss=0.2616, pruned_loss=0.06397, over 1440937.97 frames. ], batch size: 45, lr: 1.72e-02, grad_scale: 16.0 +2023-03-20 20:57:48,441 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=24903.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:57:51,531 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=24909.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:57:51,862 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.575e+02 2.909e+02 3.420e+02 4.609e+02 1.423e+03, threshold=6.841e+02, percent-clipped=7.0 +2023-03-20 20:57:56,979 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7058, 5.1138, 5.0986, 5.0697, 4.8846, 4.6495, 5.1526, 4.9168], + device='cuda:0'), covar=tensor([0.0319, 0.0352, 0.0399, 0.0441, 0.0297, 0.0343, 0.0331, 0.0470], + device='cuda:0'), in_proj_covar=tensor([0.0147, 0.0157, 0.0119, 0.0120, 0.0101, 0.0148, 0.0130, 0.0109], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 20:57:59,491 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=24925.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:58:08,490 INFO [train.py:901] (0/2) Epoch 9, batch 2350, loss[loss=0.1872, simple_loss=0.2622, pruned_loss=0.05614, over 7329.00 frames. ], tot_loss[loss=0.1937, simple_loss=0.2607, pruned_loss=0.06337, over 1441863.70 frames. ], batch size: 54, lr: 1.72e-02, grad_scale: 16.0 +2023-03-20 20:58:25,509 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4038, 1.9063, 1.9857, 1.3743, 1.1812, 1.5727, 1.4960, 1.2661], + device='cuda:0'), covar=tensor([0.0286, 0.0152, 0.0193, 0.0132, 0.0279, 0.0209, 0.0168, 0.0202], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0019, 0.0018, 0.0018, 0.0018, 0.0017, 0.0019, 0.0019], + device='cuda:0'), out_proj_covar=tensor([4.4766e-05, 4.4093e-05, 4.2319e-05, 3.7421e-05, 4.4185e-05, 3.9864e-05, + 4.3915e-05, 4.6131e-05], device='cuda:0') +2023-03-20 20:58:27,378 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-20 20:58:33,933 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-20 20:58:34,916 INFO [train.py:901] (0/2) Epoch 9, batch 2400, loss[loss=0.1948, simple_loss=0.2664, pruned_loss=0.06161, over 7255.00 frames. ], tot_loss[loss=0.1935, simple_loss=0.2602, pruned_loss=0.06339, over 1440642.89 frames. ], batch size: 64, lr: 1.72e-02, grad_scale: 16.0 +2023-03-20 20:58:36,021 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=24995.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:58:43,804 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.523e+02 2.636e+02 3.211e+02 4.206e+02 8.943e+02, threshold=6.422e+02, percent-clipped=2.0 +2023-03-20 20:58:44,333 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-20 20:58:46,826 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-20 20:58:56,022 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4326, 1.8864, 2.0676, 1.3910, 1.1430, 1.5503, 1.3901, 1.0905], + device='cuda:0'), covar=tensor([0.0194, 0.0139, 0.0078, 0.0114, 0.0390, 0.0302, 0.0263, 0.0367], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0019, 0.0018, 0.0018, 0.0018, 0.0017, 0.0020, 0.0020], + device='cuda:0'), out_proj_covar=tensor([4.5453e-05, 4.4813e-05, 4.2712e-05, 3.8087e-05, 4.4643e-05, 4.0631e-05, + 4.5680e-05, 4.7681e-05], device='cuda:0') +2023-03-20 20:59:01,026 INFO [train.py:901] (0/2) Epoch 9, batch 2450, loss[loss=0.215, simple_loss=0.2759, pruned_loss=0.077, over 7276.00 frames. ], tot_loss[loss=0.1956, simple_loss=0.2622, pruned_loss=0.06448, over 1442165.69 frames. ], batch size: 52, lr: 1.71e-02, grad_scale: 16.0 +2023-03-20 20:59:07,808 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=25056.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 20:59:13,286 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-20 20:59:27,011 INFO [train.py:901] (0/2) Epoch 9, batch 2500, loss[loss=0.2287, simple_loss=0.2962, pruned_loss=0.08065, over 7338.00 frames. ], tot_loss[loss=0.1954, simple_loss=0.262, pruned_loss=0.06439, over 1443135.80 frames. ], batch size: 54, lr: 1.71e-02, grad_scale: 16.0 +2023-03-20 20:59:35,684 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.881e+02 2.803e+02 3.310e+02 3.977e+02 7.534e+02, threshold=6.620e+02, percent-clipped=2.0 +2023-03-20 20:59:37,697 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-20 20:59:41,875 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=25121.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 20:59:52,972 INFO [train.py:901] (0/2) Epoch 9, batch 2550, loss[loss=0.1935, simple_loss=0.2686, pruned_loss=0.0592, over 7252.00 frames. ], tot_loss[loss=0.1954, simple_loss=0.2624, pruned_loss=0.06422, over 1444931.58 frames. ], batch size: 52, lr: 1.71e-02, grad_scale: 16.0 +2023-03-20 21:00:05,382 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.98 vs. limit=2.0 +2023-03-20 21:00:11,426 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.41 vs. limit=5.0 +2023-03-20 21:00:18,605 INFO [train.py:901] (0/2) Epoch 9, batch 2600, loss[loss=0.2089, simple_loss=0.2794, pruned_loss=0.06919, over 7358.00 frames. ], tot_loss[loss=0.1935, simple_loss=0.2603, pruned_loss=0.06336, over 1441770.08 frames. ], batch size: 63, lr: 1.71e-02, grad_scale: 16.0 +2023-03-20 21:00:23,877 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=25203.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:00:24,291 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=25204.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:00:27,201 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.760e+02 2.724e+02 3.388e+02 4.447e+02 1.109e+03, threshold=6.776e+02, percent-clipped=3.0 +2023-03-20 21:00:34,818 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=25225.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:00:43,380 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.62 vs. limit=2.0 +2023-03-20 21:00:43,538 INFO [train.py:901] (0/2) Epoch 9, batch 2650, loss[loss=0.2074, simple_loss=0.2769, pruned_loss=0.06893, over 7279.00 frames. ], tot_loss[loss=0.1938, simple_loss=0.2603, pruned_loss=0.06365, over 1439611.67 frames. ], batch size: 52, lr: 1.71e-02, grad_scale: 16.0 +2023-03-20 21:00:47,477 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=25251.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:00:50,600 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2252, 4.0390, 3.5457, 3.9894, 3.0464, 2.8570, 4.1284, 3.3186], + device='cuda:0'), covar=tensor([0.0140, 0.0100, 0.0165, 0.0099, 0.0264, 0.0329, 0.0136, 0.0501], + device='cuda:0'), in_proj_covar=tensor([0.0251, 0.0244, 0.0222, 0.0238, 0.0286, 0.0280, 0.0251, 0.0289], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-20 21:00:58,868 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=25273.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:01:08,602 INFO [train.py:901] (0/2) Epoch 9, batch 2700, loss[loss=0.183, simple_loss=0.2502, pruned_loss=0.05787, over 7345.00 frames. ], tot_loss[loss=0.1924, simple_loss=0.2591, pruned_loss=0.06292, over 1439279.68 frames. ], batch size: 54, lr: 1.71e-02, grad_scale: 16.0 +2023-03-20 21:01:17,020 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.798e+02 2.440e+02 3.297e+02 3.906e+02 8.338e+02, threshold=6.594e+02, percent-clipped=1.0 +2023-03-20 21:01:33,769 INFO [train.py:901] (0/2) Epoch 9, batch 2750, loss[loss=0.1731, simple_loss=0.2493, pruned_loss=0.04847, over 7355.00 frames. ], tot_loss[loss=0.1921, simple_loss=0.2591, pruned_loss=0.06255, over 1441867.76 frames. ], batch size: 63, lr: 1.70e-02, grad_scale: 16.0 +2023-03-20 21:01:37,740 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=25351.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 21:01:57,912 INFO [train.py:901] (0/2) Epoch 9, batch 2800, loss[loss=0.2221, simple_loss=0.2776, pruned_loss=0.08326, over 7291.00 frames. ], tot_loss[loss=0.1919, simple_loss=0.2586, pruned_loss=0.06264, over 1443134.72 frames. ], batch size: 57, lr: 1.70e-02, grad_scale: 16.0 +2023-03-20 21:02:04,178 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8396, 3.6076, 3.5176, 3.8222, 3.6749, 3.7468, 3.7317, 3.4855], + device='cuda:0'), covar=tensor([0.0030, 0.0066, 0.0043, 0.0029, 0.0040, 0.0034, 0.0046, 0.0063], + device='cuda:0'), in_proj_covar=tensor([0.0030, 0.0036, 0.0034, 0.0032, 0.0034, 0.0033, 0.0043, 0.0041], + device='cuda:0'), out_proj_covar=tensor([8.1454e-05, 1.1000e-04, 1.0492e-04, 8.8706e-05, 9.6210e-05, 9.6151e-05, + 1.3412e-04, 1.1752e-04], device='cuda:0') +2023-03-20 21:02:06,536 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 2.013e+02 2.834e+02 3.380e+02 4.238e+02 6.274e+02, threshold=6.760e+02, percent-clipped=0.0 +2023-03-20 21:02:10,682 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-9.pt +2023-03-20 21:02:27,583 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-20 21:02:27,639 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-20 21:02:27,663 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-20 21:02:27,682 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-20 21:02:27,856 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-20 21:02:27,871 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-20 21:02:27,883 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-20 21:02:27,982 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-20 21:02:28,057 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-20 21:02:28,072 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-20 21:02:28,089 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-20 21:02:28,251 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-20 21:02:28,257 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-20 21:02:28,620 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-20 21:02:28,760 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-20 21:02:28,764 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-20 21:02:28,822 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-20 21:02:31,140 INFO [train.py:901] (0/2) Epoch 10, batch 0, loss[loss=0.2165, simple_loss=0.2745, pruned_loss=0.07921, over 7344.00 frames. ], tot_loss[loss=0.2165, simple_loss=0.2745, pruned_loss=0.07921, over 7344.00 frames. ], batch size: 63, lr: 1.64e-02, grad_scale: 16.0 +2023-03-20 21:02:31,142 INFO [train.py:926] (0/2) Computing validation loss +2023-03-20 21:02:57,192 INFO [train.py:935] (0/2) Epoch 10, validation: loss=0.1762, simple_loss=0.2634, pruned_loss=0.04454, over 1622729.00 frames. +2023-03-20 21:02:57,193 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-20 21:02:59,251 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=25421.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:03:02,810 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=25428.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 21:03:03,708 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-20 21:03:14,193 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-20 21:03:21,215 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-20 21:03:22,225 INFO [train.py:901] (0/2) Epoch 10, batch 50, loss[loss=0.2173, simple_loss=0.2884, pruned_loss=0.07313, over 7322.00 frames. ], tot_loss[loss=0.19, simple_loss=0.2557, pruned_loss=0.06213, over 322606.18 frames. ], batch size: 59, lr: 1.63e-02, grad_scale: 16.0 +2023-03-20 21:03:23,213 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-20 21:03:23,259 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=25469.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:03:26,366 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-20 21:03:33,956 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=25489.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 21:03:34,166 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.83 vs. limit=2.0 +2023-03-20 21:03:41,259 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=25504.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:03:42,862 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-20 21:03:43,323 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-20 21:03:44,807 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.454e+02 2.645e+02 3.259e+02 3.970e+02 7.214e+02, threshold=6.518e+02, percent-clipped=3.0 +2023-03-20 21:03:48,190 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.10 vs. limit=2.0 +2023-03-20 21:03:48,318 INFO [train.py:901] (0/2) Epoch 10, batch 100, loss[loss=0.22, simple_loss=0.287, pruned_loss=0.07648, over 7218.00 frames. ], tot_loss[loss=0.193, simple_loss=0.2592, pruned_loss=0.06344, over 571594.07 frames. ], batch size: 93, lr: 1.63e-02, grad_scale: 16.0 +2023-03-20 21:03:51,489 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1058, 1.4450, 1.1338, 1.2387, 1.3378, 0.8454, 0.9870, 0.9559], + device='cuda:0'), covar=tensor([0.0075, 0.0088, 0.0191, 0.0078, 0.0099, 0.0043, 0.0232, 0.0134], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0018, 0.0017, 0.0018, 0.0020, 0.0017, 0.0019, 0.0022], + device='cuda:0'), out_proj_covar=tensor([2.5096e-05, 2.1169e-05, 2.1859e-05, 2.0659e-05, 2.4304e-05, 2.0320e-05, + 2.1779e-05, 2.8127e-05], device='cuda:0') +2023-03-20 21:04:05,883 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=25552.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:04:13,854 INFO [train.py:901] (0/2) Epoch 10, batch 150, loss[loss=0.2178, simple_loss=0.2853, pruned_loss=0.07513, over 7215.00 frames. ], tot_loss[loss=0.1925, simple_loss=0.2595, pruned_loss=0.06275, over 765715.25 frames. ], batch size: 93, lr: 1.63e-02, grad_scale: 16.0 +2023-03-20 21:04:20,761 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.02 vs. limit=2.0 +2023-03-20 21:04:36,249 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.747e+02 2.740e+02 3.249e+02 4.303e+02 1.148e+03, threshold=6.498e+02, percent-clipped=2.0 +2023-03-20 21:04:39,806 INFO [train.py:901] (0/2) Epoch 10, batch 200, loss[loss=0.1546, simple_loss=0.2235, pruned_loss=0.04283, over 7160.00 frames. ], tot_loss[loss=0.1918, simple_loss=0.2589, pruned_loss=0.06238, over 914263.20 frames. ], batch size: 39, lr: 1.63e-02, grad_scale: 16.0 +2023-03-20 21:04:41,853 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-20 21:04:44,431 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0876, 1.7412, 2.4176, 1.7488, 1.2382, 1.2799, 2.0011, 1.3807], + device='cuda:0'), covar=tensor([0.0429, 0.0277, 0.0051, 0.0080, 0.0524, 0.0387, 0.0060, 0.0228], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0017, 0.0017, 0.0017, 0.0017, 0.0016, 0.0018, 0.0017], + device='cuda:0'), out_proj_covar=tensor([4.5008e-05, 4.1605e-05, 3.9607e-05, 3.5857e-05, 4.2036e-05, 3.8867e-05, + 4.0978e-05, 4.3316e-05], device='cuda:0') +2023-03-20 21:04:46,788 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-20 21:04:52,874 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-20 21:04:56,397 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6020, 2.3844, 2.7155, 2.5650, 2.6492, 2.4390, 1.9591, 2.5697], + device='cuda:0'), covar=tensor([0.1131, 0.0709, 0.0915, 0.1320, 0.0941, 0.1076, 0.2684, 0.1148], + device='cuda:0'), in_proj_covar=tensor([0.0031, 0.0032, 0.0031, 0.0033, 0.0029, 0.0028, 0.0042, 0.0031], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 21:04:57,357 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=25651.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 21:05:03,002 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6045, 2.3967, 2.8027, 2.7102, 2.6658, 2.4361, 2.0409, 2.6931], + device='cuda:0'), covar=tensor([0.2278, 0.1060, 0.1682, 0.2438, 0.1813, 0.2117, 0.3407, 0.2464], + device='cuda:0'), in_proj_covar=tensor([0.0031, 0.0032, 0.0030, 0.0033, 0.0029, 0.0029, 0.0042, 0.0031], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 21:05:05,340 INFO [train.py:901] (0/2) Epoch 10, batch 250, loss[loss=0.1904, simple_loss=0.2607, pruned_loss=0.06007, over 7264.00 frames. ], tot_loss[loss=0.1926, simple_loss=0.2599, pruned_loss=0.06264, over 1032473.53 frames. ], batch size: 89, lr: 1.63e-02, grad_scale: 16.0 +2023-03-20 21:05:05,348 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-20 21:05:22,193 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=25699.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:05:23,473 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.61 vs. limit=2.0 +2023-03-20 21:05:26,639 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-20 21:05:27,634 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.691e+02 2.598e+02 3.271e+02 3.875e+02 1.164e+03, threshold=6.543e+02, percent-clipped=3.0 +2023-03-20 21:05:31,190 INFO [train.py:901] (0/2) Epoch 10, batch 300, loss[loss=0.1616, simple_loss=0.2285, pruned_loss=0.04736, over 7164.00 frames. ], tot_loss[loss=0.1913, simple_loss=0.2587, pruned_loss=0.06196, over 1122375.65 frames. ], batch size: 41, lr: 1.63e-02, grad_scale: 16.0 +2023-03-20 21:05:35,721 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-20 21:05:40,486 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1525, 2.0161, 2.1139, 3.0202, 1.3216, 2.7772, 1.2659, 3.1702], + device='cuda:0'), covar=tensor([0.0045, 0.0947, 0.1878, 0.0043, 0.4696, 0.0047, 0.1184, 0.0091], + device='cuda:0'), in_proj_covar=tensor([0.0138, 0.0260, 0.0318, 0.0142, 0.0303, 0.0146, 0.0278, 0.0182], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 21:05:57,627 INFO [train.py:901] (0/2) Epoch 10, batch 350, loss[loss=0.1764, simple_loss=0.2497, pruned_loss=0.05157, over 7318.00 frames. ], tot_loss[loss=0.1919, simple_loss=0.2592, pruned_loss=0.06231, over 1194051.21 frames. ], batch size: 83, lr: 1.63e-02, grad_scale: 16.0 +2023-03-20 21:06:06,207 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=25784.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 21:06:11,625 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-20 21:06:19,386 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.387e+02 2.765e+02 3.344e+02 4.046e+02 7.837e+02, threshold=6.687e+02, percent-clipped=1.0 +2023-03-20 21:06:23,526 INFO [train.py:901] (0/2) Epoch 10, batch 400, loss[loss=0.1828, simple_loss=0.2527, pruned_loss=0.05648, over 7300.00 frames. ], tot_loss[loss=0.1914, simple_loss=0.2588, pruned_loss=0.06205, over 1249532.65 frames. ], batch size: 80, lr: 1.62e-02, grad_scale: 16.0 +2023-03-20 21:06:29,743 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1191, 4.2829, 4.0702, 4.1939, 3.9707, 4.2613, 4.5093, 4.5321], + device='cuda:0'), covar=tensor([0.0160, 0.0119, 0.0191, 0.0156, 0.0344, 0.0183, 0.0253, 0.0202], + device='cuda:0'), in_proj_covar=tensor([0.0096, 0.0096, 0.0091, 0.0102, 0.0097, 0.0078, 0.0081, 0.0077], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:06:44,803 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=25858.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 21:06:49,124 INFO [train.py:901] (0/2) Epoch 10, batch 450, loss[loss=0.2026, simple_loss=0.2685, pruned_loss=0.06836, over 7215.00 frames. ], tot_loss[loss=0.1913, simple_loss=0.259, pruned_loss=0.06179, over 1293633.06 frames. ], batch size: 50, lr: 1.62e-02, grad_scale: 16.0 +2023-03-20 21:06:52,714 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-20 21:06:53,238 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-20 21:07:05,682 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.83 vs. limit=2.0 +2023-03-20 21:07:11,261 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.486e+02 2.544e+02 3.144e+02 3.954e+02 7.996e+02, threshold=6.287e+02, percent-clipped=4.0 +2023-03-20 21:07:14,820 INFO [train.py:901] (0/2) Epoch 10, batch 500, loss[loss=0.1928, simple_loss=0.2714, pruned_loss=0.05712, over 7240.00 frames. ], tot_loss[loss=0.1897, simple_loss=0.2577, pruned_loss=0.06083, over 1327179.89 frames. ], batch size: 93, lr: 1.62e-02, grad_scale: 16.0 +2023-03-20 21:07:15,976 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=25919.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 21:07:26,317 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-20 21:07:28,453 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-20 21:07:28,965 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-20 21:07:31,503 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-20 21:07:36,014 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-20 21:07:40,526 INFO [train.py:901] (0/2) Epoch 10, batch 550, loss[loss=0.1799, simple_loss=0.251, pruned_loss=0.0544, over 7320.00 frames. ], tot_loss[loss=0.1897, simple_loss=0.2577, pruned_loss=0.0609, over 1355027.56 frames. ], batch size: 75, lr: 1.62e-02, grad_scale: 16.0 +2023-03-20 21:07:47,124 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-20 21:07:51,765 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2786, 1.7396, 2.3475, 1.7842, 1.4465, 1.5336, 1.7847, 1.4100], + device='cuda:0'), covar=tensor([0.0207, 0.0166, 0.0038, 0.0062, 0.0387, 0.0244, 0.0126, 0.0286], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0019, 0.0018, 0.0017, 0.0019, 0.0017, 0.0019, 0.0019], + device='cuda:0'), out_proj_covar=tensor([4.7740e-05, 4.4969e-05, 4.1846e-05, 3.7031e-05, 4.6040e-05, 4.0549e-05, + 4.4132e-05, 4.7000e-05], device='cuda:0') +2023-03-20 21:07:55,753 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-20 21:07:59,143 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-20 21:08:03,100 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.927e+02 2.798e+02 3.426e+02 4.316e+02 1.105e+03, threshold=6.853e+02, percent-clipped=8.0 +2023-03-20 21:08:05,657 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-20 21:08:06,682 INFO [train.py:901] (0/2) Epoch 10, batch 600, loss[loss=0.2077, simple_loss=0.2802, pruned_loss=0.06757, over 6737.00 frames. ], tot_loss[loss=0.1892, simple_loss=0.2573, pruned_loss=0.06053, over 1375056.19 frames. ], batch size: 106, lr: 1.62e-02, grad_scale: 16.0 +2023-03-20 21:08:22,216 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-20 21:08:30,697 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-20 21:08:32,195 INFO [train.py:901] (0/2) Epoch 10, batch 650, loss[loss=0.1765, simple_loss=0.25, pruned_loss=0.05149, over 7286.00 frames. ], tot_loss[loss=0.1885, simple_loss=0.2564, pruned_loss=0.06033, over 1390279.72 frames. ], batch size: 77, lr: 1.62e-02, grad_scale: 16.0 +2023-03-20 21:08:33,958 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-20 21:08:41,490 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=26084.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 21:08:47,969 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-20 21:08:50,534 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=26102.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:08:54,994 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.614e+02 2.571e+02 3.091e+02 3.840e+02 8.546e+02, threshold=6.182e+02, percent-clipped=2.0 +2023-03-20 21:08:56,051 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-20 21:08:58,593 INFO [train.py:901] (0/2) Epoch 10, batch 700, loss[loss=0.1905, simple_loss=0.2619, pruned_loss=0.05951, over 7278.00 frames. ], tot_loss[loss=0.1896, simple_loss=0.2575, pruned_loss=0.06082, over 1400725.24 frames. ], batch size: 77, lr: 1.61e-02, grad_scale: 16.0 +2023-03-20 21:09:06,131 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=26132.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 21:09:20,054 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-20 21:09:20,568 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-20 21:09:22,258 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=26163.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:09:24,092 INFO [train.py:901] (0/2) Epoch 10, batch 750, loss[loss=0.2541, simple_loss=0.3083, pruned_loss=0.09995, over 6683.00 frames. ], tot_loss[loss=0.1896, simple_loss=0.2575, pruned_loss=0.06086, over 1408373.08 frames. ], batch size: 106, lr: 1.61e-02, grad_scale: 16.0 +2023-03-20 21:09:24,727 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=26168.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:09:31,303 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9274, 2.1355, 1.6593, 3.2526, 3.2199, 2.7231, 2.5451, 2.8360], + device='cuda:0'), covar=tensor([0.1751, 0.0758, 0.2527, 0.0493, 0.0072, 0.0047, 0.0047, 0.0092], + device='cuda:0'), in_proj_covar=tensor([0.0303, 0.0227, 0.0275, 0.0245, 0.0123, 0.0121, 0.0135, 0.0145], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0001, 0.0001, 0.0001, 0.0002], + device='cuda:0') +2023-03-20 21:09:34,162 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-20 21:09:39,301 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-20 21:09:45,696 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-20 21:09:46,712 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.699e+02 2.670e+02 3.310e+02 4.168e+02 9.885e+02, threshold=6.621e+02, percent-clipped=5.0 +2023-03-20 21:09:46,745 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-20 21:09:48,857 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=26214.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 21:09:50,208 INFO [train.py:901] (0/2) Epoch 10, batch 800, loss[loss=0.2434, simple_loss=0.2977, pruned_loss=0.09452, over 7312.00 frames. ], tot_loss[loss=0.1898, simple_loss=0.2576, pruned_loss=0.06104, over 1415004.42 frames. ], batch size: 59, lr: 1.61e-02, grad_scale: 16.0 +2023-03-20 21:09:56,493 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=26229.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:09:57,336 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-20 21:09:59,967 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3451, 2.3006, 2.0036, 2.2829, 2.4202, 2.1718, 2.2887, 2.5243], + device='cuda:0'), covar=tensor([0.0549, 0.0501, 0.0723, 0.1240, 0.0529, 0.1277, 0.0990, 0.0992], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0033, 0.0038, 0.0035, 0.0037, 0.0037, 0.0034, 0.0034], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 21:10:07,667 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=26250.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:10:16,216 INFO [train.py:901] (0/2) Epoch 10, batch 850, loss[loss=0.1792, simple_loss=0.2531, pruned_loss=0.05268, over 7304.00 frames. ], tot_loss[loss=0.1894, simple_loss=0.2574, pruned_loss=0.06069, over 1421271.73 frames. ], batch size: 68, lr: 1.61e-02, grad_scale: 16.0 +2023-03-20 21:10:16,251 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-20 21:10:16,721 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-20 21:10:22,307 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-20 21:10:25,944 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-20 21:10:38,439 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.623e+02 2.746e+02 3.289e+02 3.916e+02 7.980e+02, threshold=6.578e+02, percent-clipped=1.0 +2023-03-20 21:10:39,125 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=26311.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:10:42,004 INFO [train.py:901] (0/2) Epoch 10, batch 900, loss[loss=0.2117, simple_loss=0.2723, pruned_loss=0.07557, over 7275.00 frames. ], tot_loss[loss=0.189, simple_loss=0.2571, pruned_loss=0.06044, over 1423449.26 frames. ], batch size: 70, lr: 1.61e-02, grad_scale: 16.0 +2023-03-20 21:11:05,734 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-20 21:11:07,759 INFO [train.py:901] (0/2) Epoch 10, batch 950, loss[loss=0.203, simple_loss=0.2779, pruned_loss=0.06406, over 7360.00 frames. ], tot_loss[loss=0.1884, simple_loss=0.2564, pruned_loss=0.06016, over 1430076.57 frames. ], batch size: 73, lr: 1.61e-02, grad_scale: 16.0 +2023-03-20 21:11:30,198 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.680e+02 2.551e+02 3.073e+02 3.786e+02 9.128e+02, threshold=6.146e+02, percent-clipped=4.0 +2023-03-20 21:11:30,224 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-20 21:11:32,340 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=26414.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:11:33,726 INFO [train.py:901] (0/2) Epoch 10, batch 1000, loss[loss=0.2049, simple_loss=0.2616, pruned_loss=0.07407, over 7221.00 frames. ], tot_loss[loss=0.1869, simple_loss=0.2551, pruned_loss=0.05937, over 1431427.29 frames. ], batch size: 45, lr: 1.61e-02, grad_scale: 16.0 +2023-03-20 21:11:50,954 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-20 21:11:55,018 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=26458.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:11:59,402 INFO [train.py:901] (0/2) Epoch 10, batch 1050, loss[loss=0.1883, simple_loss=0.2634, pruned_loss=0.05664, over 7342.00 frames. ], tot_loss[loss=0.1865, simple_loss=0.2548, pruned_loss=0.05914, over 1431941.00 frames. ], batch size: 54, lr: 1.60e-02, grad_scale: 32.0 +2023-03-20 21:12:04,207 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=26475.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:12:13,044 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-20 21:12:14,593 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0227, 4.0244, 3.5803, 3.3496, 3.6431, 2.3511, 1.8483, 4.0406], + device='cuda:0'), covar=tensor([0.0018, 0.0024, 0.0058, 0.0059, 0.0062, 0.0361, 0.0514, 0.0038], + device='cuda:0'), in_proj_covar=tensor([0.0060, 0.0058, 0.0082, 0.0068, 0.0081, 0.0102, 0.0108, 0.0072], + device='cuda:0'), out_proj_covar=tensor([7.8431e-05, 8.9109e-05, 1.1614e-04, 1.0026e-04, 1.1145e-04, 1.4580e-04, + 1.5279e-04, 1.0186e-04], device='cuda:0') +2023-03-20 21:12:16,998 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-20 21:12:22,146 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.386e+02 2.736e+02 3.241e+02 4.081e+02 9.142e+02, threshold=6.483e+02, percent-clipped=6.0 +2023-03-20 21:12:22,800 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8978, 3.5008, 3.7082, 3.8167, 3.6952, 3.8917, 3.7407, 3.5552], + device='cuda:0'), covar=tensor([0.0031, 0.0076, 0.0044, 0.0042, 0.0049, 0.0030, 0.0050, 0.0066], + device='cuda:0'), in_proj_covar=tensor([0.0030, 0.0037, 0.0035, 0.0032, 0.0035, 0.0034, 0.0043, 0.0042], + device='cuda:0'), out_proj_covar=tensor([8.1119e-05, 1.1408e-04, 1.0643e-04, 8.6652e-05, 9.6697e-05, 9.4809e-05, + 1.3134e-04, 1.1915e-04], device='cuda:0') +2023-03-20 21:12:24,379 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=26514.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 21:12:25,768 INFO [train.py:901] (0/2) Epoch 10, batch 1100, loss[loss=0.2034, simple_loss=0.2641, pruned_loss=0.07131, over 7320.00 frames. ], tot_loss[loss=0.1873, simple_loss=0.2555, pruned_loss=0.05956, over 1434736.82 frames. ], batch size: 75, lr: 1.60e-02, grad_scale: 32.0 +2023-03-20 21:12:29,307 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=26524.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:12:47,003 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-20 21:12:47,525 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 21:12:48,757 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1099, 2.9446, 2.0563, 3.1043, 2.6909, 3.1711, 1.8757, 1.8789], + device='cuda:0'), covar=tensor([0.0151, 0.0351, 0.1163, 0.0340, 0.0247, 0.0194, 0.1342, 0.1053], + device='cuda:0'), in_proj_covar=tensor([0.0184, 0.0202, 0.0298, 0.0194, 0.0219, 0.0199, 0.0266, 0.0283], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 21:12:49,158 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=26562.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 21:12:51,474 INFO [train.py:901] (0/2) Epoch 10, batch 1150, loss[loss=0.1767, simple_loss=0.2401, pruned_loss=0.05662, over 7341.00 frames. ], tot_loss[loss=0.1873, simple_loss=0.2556, pruned_loss=0.05947, over 1438338.97 frames. ], batch size: 44, lr: 1.60e-02, grad_scale: 32.0 +2023-03-20 21:12:57,056 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1640, 2.0004, 1.9539, 2.1397, 2.3127, 2.0910, 2.3220, 2.2745], + device='cuda:0'), covar=tensor([0.0622, 0.0761, 0.0799, 0.0706, 0.0711, 0.0579, 0.1008, 0.1056], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0032, 0.0038, 0.0034, 0.0036, 0.0035, 0.0034, 0.0034], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 21:12:59,463 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-20 21:12:59,543 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6048, 3.4279, 3.2342, 3.1951, 2.7842, 3.2452, 3.3386, 3.3508], + device='cuda:0'), covar=tensor([0.0182, 0.0193, 0.0222, 0.0248, 0.0421, 0.0193, 0.0325, 0.0176], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0056, 0.0058, 0.0046, 0.0082, 0.0061, 0.0059, 0.0054], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 21:12:59,932 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-20 21:13:07,286 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.14 vs. limit=5.0 +2023-03-20 21:13:11,818 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=26606.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:13:13,759 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.942e+02 2.768e+02 3.288e+02 3.861e+02 9.785e+02, threshold=6.575e+02, percent-clipped=4.0 +2023-03-20 21:13:17,298 INFO [train.py:901] (0/2) Epoch 10, batch 1200, loss[loss=0.2427, simple_loss=0.307, pruned_loss=0.08913, over 6650.00 frames. ], tot_loss[loss=0.1876, simple_loss=0.2558, pruned_loss=0.05967, over 1438907.43 frames. ], batch size: 106, lr: 1.60e-02, grad_scale: 32.0 +2023-03-20 21:13:28,402 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7209, 4.0709, 3.9911, 4.0186, 4.0426, 3.8196, 4.1255, 3.9059], + device='cuda:0'), covar=tensor([0.0954, 0.1133, 0.1118, 0.1172, 0.0672, 0.0754, 0.1064, 0.1087], + device='cuda:0'), in_proj_covar=tensor([0.0148, 0.0160, 0.0119, 0.0122, 0.0105, 0.0149, 0.0136, 0.0109], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:13:32,914 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-20 21:13:43,166 INFO [train.py:901] (0/2) Epoch 10, batch 1250, loss[loss=0.1974, simple_loss=0.2596, pruned_loss=0.06765, over 7284.00 frames. ], tot_loss[loss=0.1881, simple_loss=0.2564, pruned_loss=0.05989, over 1440282.30 frames. ], batch size: 57, lr: 1.60e-02, grad_scale: 16.0 +2023-03-20 21:13:57,734 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-20 21:14:01,792 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-20 21:14:02,801 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-20 21:14:05,805 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.445e+02 2.594e+02 3.216e+02 3.895e+02 8.682e+02, threshold=6.432e+02, percent-clipped=1.0 +2023-03-20 21:14:08,770 INFO [train.py:901] (0/2) Epoch 10, batch 1300, loss[loss=0.231, simple_loss=0.2948, pruned_loss=0.08363, over 7105.00 frames. ], tot_loss[loss=0.1892, simple_loss=0.2574, pruned_loss=0.06053, over 1441956.52 frames. ], batch size: 98, lr: 1.60e-02, grad_scale: 16.0 +2023-03-20 21:14:25,895 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-20 21:14:27,718 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.01 vs. limit=2.0 +2023-03-20 21:14:28,413 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-20 21:14:29,975 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=26758.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:14:31,895 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-20 21:14:34,315 INFO [train.py:901] (0/2) Epoch 10, batch 1350, loss[loss=0.195, simple_loss=0.2668, pruned_loss=0.06166, over 7316.00 frames. ], tot_loss[loss=0.1898, simple_loss=0.2577, pruned_loss=0.06092, over 1441810.08 frames. ], batch size: 75, lr: 1.60e-02, grad_scale: 16.0 +2023-03-20 21:14:35,502 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7092, 4.6238, 4.5738, 5.0480, 5.0362, 5.0497, 4.2520, 4.5831], + device='cuda:0'), covar=tensor([0.0909, 0.1989, 0.1895, 0.0990, 0.0512, 0.1028, 0.0705, 0.0893], + device='cuda:0'), in_proj_covar=tensor([0.0160, 0.0258, 0.0220, 0.0206, 0.0158, 0.0263, 0.0148, 0.0182], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0001, 0.0002], + device='cuda:0') +2023-03-20 21:14:36,525 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=26770.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:14:42,403 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-20 21:14:45,513 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7242, 2.0636, 2.7937, 2.6085, 2.7448, 2.4811, 1.9313, 2.5957], + device='cuda:0'), covar=tensor([0.1101, 0.0409, 0.1112, 0.1865, 0.1078, 0.1626, 0.3487, 0.1602], + device='cuda:0'), in_proj_covar=tensor([0.0030, 0.0031, 0.0029, 0.0032, 0.0030, 0.0029, 0.0042, 0.0031], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 21:14:54,740 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=26806.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:14:57,123 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.502e+02 2.907e+02 3.495e+02 4.130e+02 7.086e+02, threshold=6.990e+02, percent-clipped=4.0 +2023-03-20 21:15:00,105 INFO [train.py:901] (0/2) Epoch 10, batch 1400, loss[loss=0.1729, simple_loss=0.2474, pruned_loss=0.04915, over 7290.00 frames. ], tot_loss[loss=0.1896, simple_loss=0.2576, pruned_loss=0.06078, over 1442064.16 frames. ], batch size: 68, lr: 1.59e-02, grad_scale: 16.0 +2023-03-20 21:15:04,413 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=26824.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:15:15,441 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-20 21:15:26,789 INFO [train.py:901] (0/2) Epoch 10, batch 1450, loss[loss=0.1502, simple_loss=0.2211, pruned_loss=0.03971, over 7176.00 frames. ], tot_loss[loss=0.1886, simple_loss=0.2571, pruned_loss=0.06, over 1441870.07 frames. ], batch size: 41, lr: 1.59e-02, grad_scale: 16.0 +2023-03-20 21:15:29,265 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=26872.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:15:39,858 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-20 21:15:43,673 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-20 21:15:45,357 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=26904.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 21:15:46,293 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=26906.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:15:48,655 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.91 vs. limit=2.0 +2023-03-20 21:15:49,356 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.718e+02 2.577e+02 2.895e+02 3.554e+02 7.159e+02, threshold=5.791e+02, percent-clipped=1.0 +2023-03-20 21:15:52,290 INFO [train.py:901] (0/2) Epoch 10, batch 1500, loss[loss=0.1644, simple_loss=0.2352, pruned_loss=0.04677, over 7353.00 frames. ], tot_loss[loss=0.1897, simple_loss=0.2582, pruned_loss=0.06065, over 1443593.16 frames. ], batch size: 44, lr: 1.59e-02, grad_scale: 16.0 +2023-03-20 21:15:56,453 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1606, 1.6520, 1.9640, 1.4818, 1.4125, 1.7332, 1.5405, 1.4815], + device='cuda:0'), covar=tensor([0.0448, 0.0125, 0.0118, 0.0071, 0.0346, 0.0220, 0.0157, 0.0235], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0019, 0.0018, 0.0017, 0.0019, 0.0016, 0.0019, 0.0019], + device='cuda:0'), out_proj_covar=tensor([4.6492e-05, 4.5858e-05, 4.2377e-05, 3.6128e-05, 4.5393e-05, 3.9950e-05, + 4.4928e-05, 4.7222e-05], device='cuda:0') +2023-03-20 21:15:56,832 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-20 21:16:09,713 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=26950.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:16:11,702 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=26954.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:16:17,361 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.9083, 1.0209, 1.1277, 1.3744, 1.2549, 1.2348, 0.7290, 0.9939], + device='cuda:0'), covar=tensor([0.1557, 0.2137, 0.0809, 0.0569, 0.1070, 0.1533, 0.0507, 0.2318], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0040, 0.0032, 0.0034, 0.0034, 0.0034, 0.0040, 0.0035], + device='cuda:0'), out_proj_covar=tensor([7.5172e-05, 9.1410e-05, 6.8331e-05, 7.1833e-05, 7.3817e-05, 7.5514e-05, + 8.7104e-05, 8.0285e-05], device='cuda:0') +2023-03-20 21:16:17,371 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=26965.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 21:16:18,234 INFO [train.py:901] (0/2) Epoch 10, batch 1550, loss[loss=0.1887, simple_loss=0.2612, pruned_loss=0.05812, over 7325.00 frames. ], tot_loss[loss=0.1888, simple_loss=0.2572, pruned_loss=0.06021, over 1442602.74 frames. ], batch size: 75, lr: 1.59e-02, grad_scale: 16.0 +2023-03-20 21:16:20,755 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-20 21:16:41,107 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.748e+02 2.661e+02 3.295e+02 4.115e+02 7.589e+02, threshold=6.589e+02, percent-clipped=5.0 +2023-03-20 21:16:41,264 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=27011.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:16:44,154 INFO [train.py:901] (0/2) Epoch 10, batch 1600, loss[loss=0.208, simple_loss=0.2702, pruned_loss=0.07289, over 7254.00 frames. ], tot_loss[loss=0.1897, simple_loss=0.2579, pruned_loss=0.06073, over 1441786.90 frames. ], batch size: 55, lr: 1.59e-02, grad_scale: 16.0 +2023-03-20 21:16:51,882 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.12 vs. limit=5.0 +2023-03-20 21:16:53,158 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-20 21:16:53,670 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-20 21:16:56,661 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-20 21:17:05,744 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-20 21:17:06,420 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2941, 2.8116, 2.4494, 3.4454, 1.6471, 3.6375, 1.7868, 3.8111], + device='cuda:0'), covar=tensor([0.0028, 0.0487, 0.1600, 0.0029, 0.4107, 0.0074, 0.0867, 0.0105], + device='cuda:0'), in_proj_covar=tensor([0.0142, 0.0269, 0.0318, 0.0142, 0.0303, 0.0145, 0.0277, 0.0190], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 21:17:09,741 INFO [train.py:901] (0/2) Epoch 10, batch 1650, loss[loss=0.1823, simple_loss=0.2541, pruned_loss=0.05528, over 7325.00 frames. ], tot_loss[loss=0.1886, simple_loss=0.257, pruned_loss=0.06012, over 1440192.96 frames. ], batch size: 61, lr: 1.59e-02, grad_scale: 16.0 +2023-03-20 21:17:09,770 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-20 21:17:11,338 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=27070.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:17:17,734 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-20 21:17:32,622 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.584e+02 2.597e+02 3.202e+02 3.954e+02 1.371e+03, threshold=6.405e+02, percent-clipped=3.0 +2023-03-20 21:17:35,706 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 21:17:36,183 INFO [train.py:901] (0/2) Epoch 10, batch 1700, loss[loss=0.1478, simple_loss=0.2116, pruned_loss=0.04197, over 7003.00 frames. ], tot_loss[loss=0.1878, simple_loss=0.2562, pruned_loss=0.0597, over 1440646.79 frames. ], batch size: 35, lr: 1.59e-02, grad_scale: 16.0 +2023-03-20 21:17:36,757 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=27118.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:17:39,678 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-20 21:17:46,478 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.69 vs. limit=2.0 +2023-03-20 21:17:50,340 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-20 21:17:51,958 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=27148.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:18:01,317 INFO [train.py:901] (0/2) Epoch 10, batch 1750, loss[loss=0.2054, simple_loss=0.2707, pruned_loss=0.07003, over 7309.00 frames. ], tot_loss[loss=0.1881, simple_loss=0.2566, pruned_loss=0.05981, over 1442054.87 frames. ], batch size: 83, lr: 1.58e-02, grad_scale: 16.0 +2023-03-20 21:18:14,982 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-20 21:18:15,994 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-20 21:18:19,719 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0771, 3.7156, 3.5918, 3.8223, 3.1239, 2.8788, 4.0002, 3.2025], + device='cuda:0'), covar=tensor([0.0124, 0.0143, 0.0164, 0.0133, 0.0232, 0.0323, 0.0171, 0.0466], + device='cuda:0'), in_proj_covar=tensor([0.0260, 0.0263, 0.0236, 0.0259, 0.0292, 0.0287, 0.0256, 0.0286], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-20 21:18:20,855 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.02 vs. limit=2.0 +2023-03-20 21:18:24,279 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=27209.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:18:25,131 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.954e+02 2.477e+02 3.182e+02 3.732e+02 1.026e+03, threshold=6.364e+02, percent-clipped=1.0 +2023-03-20 21:18:28,122 INFO [train.py:901] (0/2) Epoch 10, batch 1800, loss[loss=0.1624, simple_loss=0.2401, pruned_loss=0.0423, over 7272.00 frames. ], tot_loss[loss=0.1869, simple_loss=0.2555, pruned_loss=0.05918, over 1441378.86 frames. ], batch size: 70, lr: 1.58e-02, grad_scale: 16.0 +2023-03-20 21:18:38,768 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-20 21:18:43,961 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0825, 1.1339, 1.2412, 1.1759, 1.1668, 1.0205, 1.2738, 0.9409], + device='cuda:0'), covar=tensor([0.0288, 0.0193, 0.0310, 0.0159, 0.0241, 0.0165, 0.0113, 0.0230], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0017, 0.0017, 0.0018, 0.0021, 0.0018, 0.0018, 0.0022], + device='cuda:0'), out_proj_covar=tensor([2.4807e-05, 2.0529e-05, 2.1769e-05, 2.0881e-05, 2.6027e-05, 2.0207e-05, + 2.0995e-05, 2.8404e-05], device='cuda:0') +2023-03-20 21:18:50,509 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=27260.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 21:18:52,399 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-20 21:18:53,902 INFO [train.py:901] (0/2) Epoch 10, batch 1850, loss[loss=0.1604, simple_loss=0.217, pruned_loss=0.05186, over 6156.00 frames. ], tot_loss[loss=0.1869, simple_loss=0.2552, pruned_loss=0.05927, over 1439849.87 frames. ], batch size: 26, lr: 1.58e-02, grad_scale: 16.0 +2023-03-20 21:19:02,437 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-20 21:19:06,148 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2955, 3.7510, 3.9351, 3.8482, 3.6693, 3.8071, 4.1901, 3.7809], + device='cuda:0'), covar=tensor([0.0084, 0.0140, 0.0130, 0.0137, 0.0253, 0.0105, 0.0128, 0.0117], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0059, 0.0059, 0.0047, 0.0083, 0.0062, 0.0061, 0.0056], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 21:19:14,296 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=27306.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:19:16,779 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.365e+02 2.681e+02 3.255e+02 3.906e+02 1.167e+03, threshold=6.510e+02, percent-clipped=5.0 +2023-03-20 21:19:19,307 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-20 21:19:19,784 INFO [train.py:901] (0/2) Epoch 10, batch 1900, loss[loss=0.1976, simple_loss=0.2665, pruned_loss=0.06434, over 7310.00 frames. ], tot_loss[loss=0.1869, simple_loss=0.2555, pruned_loss=0.05916, over 1441611.79 frames. ], batch size: 86, lr: 1.58e-02, grad_scale: 16.0 +2023-03-20 21:19:42,669 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-20 21:19:45,114 INFO [train.py:901] (0/2) Epoch 10, batch 1950, loss[loss=0.1744, simple_loss=0.2564, pruned_loss=0.04622, over 7272.00 frames. ], tot_loss[loss=0.1882, simple_loss=0.2565, pruned_loss=0.05992, over 1443328.69 frames. ], batch size: 70, lr: 1.58e-02, grad_scale: 16.0 +2023-03-20 21:19:53,643 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-20 21:19:53,758 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9715, 4.0123, 3.5440, 3.3134, 3.4864, 2.3134, 1.7235, 3.9510], + device='cuda:0'), covar=tensor([0.0018, 0.0029, 0.0074, 0.0046, 0.0060, 0.0354, 0.0464, 0.0035], + device='cuda:0'), in_proj_covar=tensor([0.0060, 0.0056, 0.0078, 0.0066, 0.0079, 0.0101, 0.0105, 0.0068], + device='cuda:0'), out_proj_covar=tensor([7.8645e-05, 8.4801e-05, 1.0986e-04, 9.5358e-05, 1.0904e-04, 1.4321e-04, + 1.4786e-04, 9.6492e-05], device='cuda:0') +2023-03-20 21:19:57,690 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-20 21:19:58,224 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-20 21:20:01,386 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1902, 1.1331, 1.3570, 1.2207, 1.1742, 0.9770, 0.9903, 0.8946], + device='cuda:0'), covar=tensor([0.0118, 0.0087, 0.0110, 0.0126, 0.0130, 0.0106, 0.0136, 0.0169], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0018, 0.0017, 0.0018, 0.0022, 0.0018, 0.0018, 0.0022], + device='cuda:0'), out_proj_covar=tensor([2.4631e-05, 2.0925e-05, 2.1839e-05, 2.0770e-05, 2.6548e-05, 2.0584e-05, + 2.1489e-05, 2.8433e-05], device='cuda:0') +2023-03-20 21:20:07,179 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1915, 1.1014, 1.3068, 1.5801, 1.5454, 1.4988, 1.0525, 1.3483], + device='cuda:0'), covar=tensor([0.1974, 0.1989, 0.0663, 0.0663, 0.0935, 0.1529, 0.1097, 0.1787], + device='cuda:0'), in_proj_covar=tensor([0.0036, 0.0041, 0.0031, 0.0034, 0.0033, 0.0032, 0.0041, 0.0036], + device='cuda:0'), out_proj_covar=tensor([7.7999e-05, 9.2720e-05, 6.7353e-05, 7.2340e-05, 7.3609e-05, 7.4363e-05, + 8.9643e-05, 8.1942e-05], device='cuda:0') +2023-03-20 21:20:08,032 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.538e+02 2.747e+02 3.327e+02 4.450e+02 8.863e+02, threshold=6.654e+02, percent-clipped=5.0 +2023-03-20 21:20:10,963 INFO [train.py:901] (0/2) Epoch 10, batch 2000, loss[loss=0.2149, simple_loss=0.2764, pruned_loss=0.0767, over 7298.00 frames. ], tot_loss[loss=0.1895, simple_loss=0.2577, pruned_loss=0.06064, over 1443528.18 frames. ], batch size: 86, lr: 1.58e-02, grad_scale: 16.0 +2023-03-20 21:20:11,270 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.76 vs. limit=5.0 +2023-03-20 21:20:14,973 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-20 21:20:22,669 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9891, 2.7525, 2.8014, 2.9396, 3.0667, 2.9550, 2.4011, 2.7866], + device='cuda:0'), covar=tensor([0.1155, 0.0620, 0.2308, 0.2093, 0.1026, 0.1251, 0.2317, 0.1536], + device='cuda:0'), in_proj_covar=tensor([0.0029, 0.0030, 0.0031, 0.0032, 0.0028, 0.0029, 0.0040, 0.0030], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 21:20:26,612 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-20 21:20:34,743 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-20 21:20:37,257 INFO [train.py:901] (0/2) Epoch 10, batch 2050, loss[loss=0.1864, simple_loss=0.2607, pruned_loss=0.05603, over 7300.00 frames. ], tot_loss[loss=0.1886, simple_loss=0.2571, pruned_loss=0.06007, over 1443692.28 frames. ], batch size: 83, lr: 1.58e-02, grad_scale: 16.0 +2023-03-20 21:20:41,661 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.81 vs. limit=2.0 +2023-03-20 21:20:55,677 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=27504.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:20:59,078 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.657e+02 2.559e+02 3.090e+02 3.749e+02 7.460e+02, threshold=6.180e+02, percent-clipped=1.0 +2023-03-20 21:21:02,003 INFO [train.py:901] (0/2) Epoch 10, batch 2100, loss[loss=0.1369, simple_loss=0.1889, pruned_loss=0.04244, over 5951.00 frames. ], tot_loss[loss=0.1894, simple_loss=0.2575, pruned_loss=0.06061, over 1441536.22 frames. ], batch size: 25, lr: 1.57e-02, grad_scale: 16.0 +2023-03-20 21:21:07,794 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-20 21:21:10,855 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-20 21:21:25,165 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=27560.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 21:21:28,516 INFO [train.py:901] (0/2) Epoch 10, batch 2150, loss[loss=0.196, simple_loss=0.262, pruned_loss=0.06499, over 7301.00 frames. ], tot_loss[loss=0.1886, simple_loss=0.2572, pruned_loss=0.06005, over 1441357.29 frames. ], batch size: 83, lr: 1.57e-02, grad_scale: 16.0 +2023-03-20 21:21:40,667 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8856, 3.6129, 3.6613, 3.5610, 3.2379, 3.6554, 3.7109, 3.2748], + device='cuda:0'), covar=tensor([0.0166, 0.0190, 0.0147, 0.0190, 0.0352, 0.0120, 0.0202, 0.0249], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0060, 0.0060, 0.0048, 0.0085, 0.0063, 0.0062, 0.0058], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 21:21:46,538 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1684, 3.7746, 3.9125, 3.7873, 3.5792, 3.8524, 4.0125, 3.4196], + device='cuda:0'), covar=tensor([0.0109, 0.0147, 0.0091, 0.0129, 0.0217, 0.0083, 0.0152, 0.0187], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0060, 0.0060, 0.0048, 0.0085, 0.0063, 0.0061, 0.0058], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 21:21:49,157 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=27606.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:21:50,120 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=27608.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 21:21:51,525 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.534e+02 2.915e+02 3.456e+02 4.146e+02 1.054e+03, threshold=6.911e+02, percent-clipped=5.0 +2023-03-20 21:21:54,473 INFO [train.py:901] (0/2) Epoch 10, batch 2200, loss[loss=0.178, simple_loss=0.2542, pruned_loss=0.05089, over 7250.00 frames. ], tot_loss[loss=0.1883, simple_loss=0.257, pruned_loss=0.05979, over 1444260.63 frames. ], batch size: 47, lr: 1.57e-02, grad_scale: 16.0 +2023-03-20 21:21:56,461 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-20 21:21:58,803 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1290, 1.2090, 1.2493, 1.3004, 1.2014, 0.9833, 1.0099, 0.9827], + device='cuda:0'), covar=tensor([0.0187, 0.0097, 0.0095, 0.0090, 0.0089, 0.0063, 0.0108, 0.0181], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0018, 0.0018, 0.0018, 0.0022, 0.0019, 0.0019, 0.0023], + device='cuda:0'), out_proj_covar=tensor([2.5125e-05, 2.1336e-05, 2.2686e-05, 2.1122e-05, 2.7120e-05, 2.1367e-05, + 2.2092e-05, 2.9680e-05], device='cuda:0') +2023-03-20 21:22:13,769 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=27654.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:22:20,190 INFO [train.py:901] (0/2) Epoch 10, batch 2250, loss[loss=0.1995, simple_loss=0.2593, pruned_loss=0.06991, over 7203.00 frames. ], tot_loss[loss=0.1877, simple_loss=0.2562, pruned_loss=0.05958, over 1442612.58 frames. ], batch size: 50, lr: 1.57e-02, grad_scale: 16.0 +2023-03-20 21:22:29,315 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1927, 3.6701, 3.8959, 3.8016, 3.5806, 3.6551, 4.1100, 3.5870], + device='cuda:0'), covar=tensor([0.0086, 0.0172, 0.0122, 0.0118, 0.0239, 0.0125, 0.0122, 0.0171], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0059, 0.0058, 0.0046, 0.0083, 0.0061, 0.0060, 0.0056], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 21:22:30,698 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-20 21:22:31,197 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-20 21:22:35,349 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1930, 1.0583, 1.1744, 1.5715, 1.4574, 1.3563, 0.9176, 1.1022], + device='cuda:0'), covar=tensor([0.0964, 0.1913, 0.0801, 0.1044, 0.1318, 0.1057, 0.0979, 0.2536], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0041, 0.0031, 0.0035, 0.0034, 0.0034, 0.0042, 0.0035], + device='cuda:0'), out_proj_covar=tensor([8.0345e-05, 9.3623e-05, 6.8307e-05, 7.3993e-05, 7.5331e-05, 7.7677e-05, + 9.1618e-05, 8.1754e-05], device='cuda:0') +2023-03-20 21:22:43,306 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.716e+02 2.679e+02 3.166e+02 3.918e+02 8.017e+02, threshold=6.332e+02, percent-clipped=1.0 +2023-03-20 21:22:43,828 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-20 21:22:44,961 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=27714.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 21:22:46,319 INFO [train.py:901] (0/2) Epoch 10, batch 2300, loss[loss=0.1902, simple_loss=0.253, pruned_loss=0.06373, over 7255.00 frames. ], tot_loss[loss=0.1883, simple_loss=0.2567, pruned_loss=0.05996, over 1443099.94 frames. ], batch size: 55, lr: 1.57e-02, grad_scale: 16.0 +2023-03-20 21:22:52,987 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0583, 2.0068, 2.3232, 1.8828, 1.1395, 1.6571, 1.6193, 1.6617], + device='cuda:0'), covar=tensor([0.0463, 0.0185, 0.0084, 0.0043, 0.0597, 0.0476, 0.0159, 0.0293], + device='cuda:0'), in_proj_covar=tensor([0.0021, 0.0021, 0.0020, 0.0019, 0.0020, 0.0018, 0.0021, 0.0021], + device='cuda:0'), out_proj_covar=tensor([4.9851e-05, 4.9663e-05, 4.6377e-05, 4.0312e-05, 4.9005e-05, 4.3485e-05, + 4.9478e-05, 5.1444e-05], device='cuda:0') +2023-03-20 21:23:11,754 INFO [train.py:901] (0/2) Epoch 10, batch 2350, loss[loss=0.1867, simple_loss=0.2619, pruned_loss=0.05577, over 7324.00 frames. ], tot_loss[loss=0.188, simple_loss=0.2567, pruned_loss=0.05969, over 1444561.82 frames. ], batch size: 75, lr: 1.57e-02, grad_scale: 16.0 +2023-03-20 21:23:15,951 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=27775.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 21:23:31,453 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-20 21:23:32,039 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=27804.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:23:35,330 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.873e+02 2.816e+02 3.299e+02 4.187e+02 9.053e+02, threshold=6.598e+02, percent-clipped=4.0 +2023-03-20 21:23:38,392 INFO [train.py:901] (0/2) Epoch 10, batch 2400, loss[loss=0.2027, simple_loss=0.2675, pruned_loss=0.069, over 7354.00 frames. ], tot_loss[loss=0.1872, simple_loss=0.2559, pruned_loss=0.05929, over 1443466.48 frames. ], batch size: 73, lr: 1.57e-02, grad_scale: 16.0 +2023-03-20 21:23:38,401 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-20 21:23:48,939 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-20 21:23:51,454 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-20 21:23:51,519 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3288, 4.7538, 4.8294, 4.7545, 4.5639, 4.3663, 4.8709, 4.6614], + device='cuda:0'), covar=tensor([0.0399, 0.0385, 0.0299, 0.0360, 0.0323, 0.0320, 0.0268, 0.0393], + device='cuda:0'), in_proj_covar=tensor([0.0155, 0.0169, 0.0124, 0.0127, 0.0107, 0.0157, 0.0137, 0.0108], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:23:55,994 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=27852.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:24:03,558 INFO [train.py:901] (0/2) Epoch 10, batch 2450, loss[loss=0.1518, simple_loss=0.2218, pruned_loss=0.04092, over 7168.00 frames. ], tot_loss[loss=0.1864, simple_loss=0.255, pruned_loss=0.0589, over 1443058.06 frames. ], batch size: 39, lr: 1.56e-02, grad_scale: 16.0 +2023-03-20 21:24:09,055 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.02 vs. limit=2.0 +2023-03-20 21:24:18,758 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-03-20 21:24:19,845 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-20 21:24:26,788 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.336e+02 2.708e+02 3.245e+02 4.079e+02 9.497e+02, threshold=6.490e+02, percent-clipped=3.0 +2023-03-20 21:24:29,836 INFO [train.py:901] (0/2) Epoch 10, batch 2500, loss[loss=0.1618, simple_loss=0.2365, pruned_loss=0.04358, over 7338.00 frames. ], tot_loss[loss=0.1868, simple_loss=0.2553, pruned_loss=0.05919, over 1441420.98 frames. ], batch size: 44, lr: 1.56e-02, grad_scale: 16.0 +2023-03-20 21:24:35,667 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.05 vs. limit=2.0 +2023-03-20 21:24:38,243 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.95 vs. limit=5.0 +2023-03-20 21:24:38,475 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2968, 3.4431, 3.1965, 3.3728, 3.3741, 3.0205, 3.4687, 3.5283], + device='cuda:0'), covar=tensor([0.0366, 0.0213, 0.0295, 0.0276, 0.0362, 0.0611, 0.0379, 0.0285], + device='cuda:0'), in_proj_covar=tensor([0.0096, 0.0096, 0.0090, 0.0097, 0.0093, 0.0078, 0.0076, 0.0077], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:24:43,385 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-20 21:24:49,038 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8455, 4.0064, 3.6883, 4.0676, 3.7802, 4.0206, 4.3622, 4.3604], + device='cuda:0'), covar=tensor([0.0243, 0.0151, 0.0213, 0.0162, 0.0340, 0.0220, 0.0230, 0.0181], + device='cuda:0'), in_proj_covar=tensor([0.0096, 0.0095, 0.0089, 0.0097, 0.0093, 0.0078, 0.0076, 0.0077], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:24:55,327 INFO [train.py:901] (0/2) Epoch 10, batch 2550, loss[loss=0.1709, simple_loss=0.2444, pruned_loss=0.04874, over 7357.00 frames. ], tot_loss[loss=0.188, simple_loss=0.2564, pruned_loss=0.05981, over 1443109.46 frames. ], batch size: 73, lr: 1.56e-02, grad_scale: 16.0 +2023-03-20 21:25:12,768 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-28000.pt +2023-03-20 21:25:21,857 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.650e+02 2.672e+02 3.067e+02 3.861e+02 6.346e+02, threshold=6.134e+02, percent-clipped=0.0 +2023-03-20 21:25:24,833 INFO [train.py:901] (0/2) Epoch 10, batch 2600, loss[loss=0.2411, simple_loss=0.3002, pruned_loss=0.09098, over 6634.00 frames. ], tot_loss[loss=0.188, simple_loss=0.2564, pruned_loss=0.0598, over 1444475.02 frames. ], batch size: 106, lr: 1.56e-02, grad_scale: 16.0 +2023-03-20 21:25:49,407 INFO [train.py:901] (0/2) Epoch 10, batch 2650, loss[loss=0.1792, simple_loss=0.2475, pruned_loss=0.05546, over 7272.00 frames. ], tot_loss[loss=0.1874, simple_loss=0.256, pruned_loss=0.05937, over 1444082.11 frames. ], batch size: 52, lr: 1.56e-02, grad_scale: 16.0 +2023-03-20 21:25:50,954 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=28070.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 21:25:59,979 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.9065, 1.0342, 1.0416, 1.1579, 1.2369, 1.2272, 0.7913, 1.1540], + device='cuda:0'), covar=tensor([0.0800, 0.2303, 0.0699, 0.0571, 0.1045, 0.1071, 0.0589, 0.1284], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0042, 0.0031, 0.0034, 0.0035, 0.0036, 0.0043, 0.0034], + device='cuda:0'), out_proj_covar=tensor([8.1576e-05, 9.5884e-05, 6.9189e-05, 7.4161e-05, 7.8768e-05, 8.1121e-05, + 9.4281e-05, 8.1217e-05], device='cuda:0') +2023-03-20 21:26:04,344 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8372, 3.5804, 3.5817, 3.5552, 3.2976, 3.4798, 3.5857, 3.3543], + device='cuda:0'), covar=tensor([0.0079, 0.0134, 0.0115, 0.0115, 0.0289, 0.0092, 0.0165, 0.0119], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0058, 0.0059, 0.0047, 0.0087, 0.0061, 0.0062, 0.0056], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 21:26:11,631 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.564e+02 2.458e+02 3.041e+02 3.552e+02 7.136e+02, threshold=6.082e+02, percent-clipped=4.0 +2023-03-20 21:26:14,555 INFO [train.py:901] (0/2) Epoch 10, batch 2700, loss[loss=0.1832, simple_loss=0.258, pruned_loss=0.05424, over 7354.00 frames. ], tot_loss[loss=0.1879, simple_loss=0.2566, pruned_loss=0.05961, over 1443267.58 frames. ], batch size: 73, lr: 1.56e-02, grad_scale: 16.0 +2023-03-20 21:26:39,400 INFO [train.py:901] (0/2) Epoch 10, batch 2750, loss[loss=0.1566, simple_loss=0.2189, pruned_loss=0.04711, over 7004.00 frames. ], tot_loss[loss=0.1879, simple_loss=0.2563, pruned_loss=0.05969, over 1444259.53 frames. ], batch size: 35, lr: 1.56e-02, grad_scale: 16.0 +2023-03-20 21:27:01,429 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.637e+02 2.622e+02 3.116e+02 3.809e+02 8.984e+02, threshold=6.231e+02, percent-clipped=2.0 +2023-03-20 21:27:04,493 INFO [train.py:901] (0/2) Epoch 10, batch 2800, loss[loss=0.225, simple_loss=0.2916, pruned_loss=0.07919, over 6671.00 frames. ], tot_loss[loss=0.1876, simple_loss=0.2561, pruned_loss=0.05958, over 1443105.08 frames. ], batch size: 106, lr: 1.56e-02, grad_scale: 16.0 +2023-03-20 21:27:10,501 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=28229.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:27:13,698 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.06 vs. limit=2.0 +2023-03-20 21:27:14,805 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-20 21:27:17,260 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-10.pt +2023-03-20 21:27:35,639 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-20 21:27:35,695 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-20 21:27:35,719 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-20 21:27:35,738 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-20 21:27:35,914 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-20 21:27:35,930 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-20 21:27:35,941 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-20 21:27:36,040 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-20 21:27:36,459 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-20 21:27:36,474 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-20 21:27:36,491 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-20 21:27:36,660 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-20 21:27:36,666 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-20 21:27:36,674 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-20 21:27:36,814 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-20 21:27:36,817 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-20 21:27:36,876 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-20 21:27:39,081 INFO [train.py:901] (0/2) Epoch 11, batch 0, loss[loss=0.2136, simple_loss=0.2754, pruned_loss=0.07591, over 7249.00 frames. ], tot_loss[loss=0.2136, simple_loss=0.2754, pruned_loss=0.07591, over 7249.00 frames. ], batch size: 55, lr: 1.50e-02, grad_scale: 16.0 +2023-03-20 21:27:39,083 INFO [train.py:926] (0/2) Computing validation loss +2023-03-20 21:27:44,671 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1116, 2.0762, 2.0192, 2.9153, 1.4919, 2.7078, 1.2824, 2.6750], + device='cuda:0'), covar=tensor([0.0060, 0.0945, 0.1515, 0.0066, 0.4473, 0.0059, 0.1155, 0.0137], + device='cuda:0'), in_proj_covar=tensor([0.0137, 0.0267, 0.0311, 0.0141, 0.0296, 0.0149, 0.0274, 0.0192], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 21:27:51,941 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4406, 1.5990, 1.1104, 1.2779, 1.0980, 1.0582, 1.2915, 1.0570], + device='cuda:0'), covar=tensor([0.0150, 0.0115, 0.0223, 0.0187, 0.0534, 0.0136, 0.0343, 0.0180], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0018, 0.0018, 0.0019, 0.0021, 0.0019, 0.0018, 0.0023], + device='cuda:0'), out_proj_covar=tensor([2.3543e-05, 2.0855e-05, 2.2578e-05, 2.1583e-05, 2.6241e-05, 2.1273e-05, + 2.1386e-05, 2.9579e-05], device='cuda:0') +2023-03-20 21:28:03,430 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1091, 1.3658, 1.0072, 1.0209, 0.9806, 0.9106, 1.1269, 0.9412], + device='cuda:0'), covar=tensor([0.0193, 0.0110, 0.0289, 0.0269, 0.0419, 0.0116, 0.0177, 0.0299], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0018, 0.0018, 0.0019, 0.0021, 0.0019, 0.0018, 0.0023], + device='cuda:0'), out_proj_covar=tensor([2.3543e-05, 2.0855e-05, 2.2578e-05, 2.1583e-05, 2.6241e-05, 2.1273e-05, + 2.1386e-05, 2.9579e-05], device='cuda:0') +2023-03-20 21:28:04,929 INFO [train.py:935] (0/2) Epoch 11, validation: loss=0.1751, simple_loss=0.2617, pruned_loss=0.04426, over 1622729.00 frames. +2023-03-20 21:28:04,930 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-20 21:28:10,267 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.59 vs. limit=5.0 +2023-03-20 21:28:11,476 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-20 21:28:12,112 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=28255.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 21:28:15,681 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.15 vs. limit=2.0 +2023-03-20 21:28:21,929 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-20 21:28:28,041 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=28287.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:28:28,460 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-20 21:28:29,557 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=28290.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:28:29,923 INFO [train.py:901] (0/2) Epoch 11, batch 50, loss[loss=0.1758, simple_loss=0.2459, pruned_loss=0.05289, over 7231.00 frames. ], tot_loss[loss=0.1867, simple_loss=0.256, pruned_loss=0.05866, over 327128.04 frames. ], batch size: 45, lr: 1.50e-02, grad_scale: 16.0 +2023-03-20 21:28:30,399 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-20 21:28:32,951 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-20 21:28:37,317 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3579, 3.1103, 2.2133, 3.4729, 2.8333, 3.3746, 1.9783, 1.9692], + device='cuda:0'), covar=tensor([0.0126, 0.0290, 0.1271, 0.0222, 0.0216, 0.0245, 0.1631, 0.1178], + device='cuda:0'), in_proj_covar=tensor([0.0183, 0.0209, 0.0302, 0.0193, 0.0226, 0.0208, 0.0268, 0.0281], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 21:28:40,670 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.678e+02 2.663e+02 3.200e+02 3.630e+02 7.149e+02, threshold=6.399e+02, percent-clipped=1.0 +2023-03-20 21:28:43,314 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=28316.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 21:28:47,826 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4668, 2.1661, 2.1941, 3.2235, 1.5040, 2.9202, 1.1687, 2.9005], + device='cuda:0'), covar=tensor([0.0069, 0.0866, 0.1836, 0.0053, 0.4194, 0.0061, 0.1144, 0.0181], + device='cuda:0'), in_proj_covar=tensor([0.0140, 0.0271, 0.0317, 0.0142, 0.0300, 0.0151, 0.0276, 0.0196], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 21:28:48,322 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1335, 1.8291, 1.9138, 1.6354, 1.4395, 1.4470, 1.5474, 1.6835], + device='cuda:0'), covar=tensor([0.0522, 0.0117, 0.0092, 0.0082, 0.0369, 0.0421, 0.0097, 0.0084], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0019, 0.0019, 0.0018, 0.0017, 0.0017, 0.0019, 0.0018], + device='cuda:0'), out_proj_covar=tensor([4.6579e-05, 4.5218e-05, 4.4187e-05, 3.8234e-05, 4.3878e-05, 4.1296e-05, + 4.5301e-05, 4.6207e-05], device='cuda:0') +2023-03-20 21:28:50,700 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-20 21:28:51,893 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-20 21:28:54,311 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.02 vs. limit=2.0 +2023-03-20 21:28:56,451 INFO [train.py:901] (0/2) Epoch 11, batch 100, loss[loss=0.1652, simple_loss=0.2341, pruned_loss=0.04811, over 7171.00 frames. ], tot_loss[loss=0.1864, simple_loss=0.2562, pruned_loss=0.05832, over 574995.08 frames. ], batch size: 41, lr: 1.49e-02, grad_scale: 16.0 +2023-03-20 21:29:00,133 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=28348.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:29:11,329 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=28370.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 21:29:21,751 INFO [train.py:901] (0/2) Epoch 11, batch 150, loss[loss=0.1898, simple_loss=0.2666, pruned_loss=0.05652, over 7312.00 frames. ], tot_loss[loss=0.1848, simple_loss=0.2543, pruned_loss=0.05762, over 768323.98 frames. ], batch size: 80, lr: 1.49e-02, grad_scale: 16.0 +2023-03-20 21:29:29,734 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9797, 4.6571, 4.5570, 5.1000, 5.1657, 5.1834, 4.6864, 4.6957], + device='cuda:0'), covar=tensor([0.0734, 0.2299, 0.2434, 0.1320, 0.0621, 0.1153, 0.0626, 0.0970], + device='cuda:0'), in_proj_covar=tensor([0.0165, 0.0272, 0.0225, 0.0209, 0.0164, 0.0270, 0.0156, 0.0190], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0001, 0.0002], + device='cuda:0') +2023-03-20 21:29:30,289 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2429, 3.5872, 3.8243, 3.8056, 3.6493, 3.9547, 4.0338, 3.4502], + device='cuda:0'), covar=tensor([0.0095, 0.0160, 0.0119, 0.0119, 0.0241, 0.0080, 0.0148, 0.0180], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0061, 0.0062, 0.0050, 0.0091, 0.0064, 0.0064, 0.0059], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 21:29:32,726 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.696e+02 2.668e+02 3.214e+02 3.934e+02 8.081e+02, threshold=6.428e+02, percent-clipped=3.0 +2023-03-20 21:29:34,871 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4826, 3.8475, 4.1801, 4.1061, 3.9637, 4.2508, 4.3453, 3.8409], + device='cuda:0'), covar=tensor([0.0104, 0.0131, 0.0110, 0.0118, 0.0211, 0.0069, 0.0125, 0.0146], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0061, 0.0062, 0.0050, 0.0091, 0.0064, 0.0064, 0.0059], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 21:29:36,375 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=28418.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 21:29:45,608 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9084, 4.7419, 4.5916, 4.0527, 4.6226, 2.9070, 2.6530, 4.8190], + device='cuda:0'), covar=tensor([0.0013, 0.0053, 0.0042, 0.0051, 0.0027, 0.0353, 0.0388, 0.0034], + device='cuda:0'), in_proj_covar=tensor([0.0059, 0.0056, 0.0076, 0.0064, 0.0079, 0.0099, 0.0104, 0.0065], + device='cuda:0'), out_proj_covar=tensor([7.7768e-05, 8.4334e-05, 1.0523e-04, 9.4989e-05, 1.0725e-04, 1.3970e-04, + 1.4574e-04, 9.0740e-05], device='cuda:0') +2023-03-20 21:29:48,524 INFO [train.py:901] (0/2) Epoch 11, batch 200, loss[loss=0.1572, simple_loss=0.2191, pruned_loss=0.04764, over 7179.00 frames. ], tot_loss[loss=0.1824, simple_loss=0.2515, pruned_loss=0.05662, over 914099.26 frames. ], batch size: 39, lr: 1.49e-02, grad_scale: 16.0 +2023-03-20 21:29:54,516 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-20 21:29:59,657 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-20 21:30:05,836 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-20 21:30:12,035 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=28486.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:30:14,383 INFO [train.py:901] (0/2) Epoch 11, batch 250, loss[loss=0.1906, simple_loss=0.2618, pruned_loss=0.05966, over 7282.00 frames. ], tot_loss[loss=0.1829, simple_loss=0.252, pruned_loss=0.05684, over 1032668.30 frames. ], batch size: 77, lr: 1.49e-02, grad_scale: 16.0 +2023-03-20 21:30:15,021 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=28492.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:30:18,987 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-20 21:30:25,036 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.444e+02 2.592e+02 3.063e+02 3.811e+02 9.501e+02, threshold=6.125e+02, percent-clipped=2.0 +2023-03-20 21:30:40,040 INFO [train.py:901] (0/2) Epoch 11, batch 300, loss[loss=0.1809, simple_loss=0.247, pruned_loss=0.05742, over 7268.00 frames. ], tot_loss[loss=0.1822, simple_loss=0.2515, pruned_loss=0.05646, over 1122442.80 frames. ], batch size: 77, lr: 1.49e-02, grad_scale: 16.0 +2023-03-20 21:30:40,054 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-20 21:30:40,903 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.01 vs. limit=2.0 +2023-03-20 21:30:43,221 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=28547.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:30:46,230 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=28553.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:30:48,087 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-20 21:30:50,147 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7271, 3.4754, 3.5764, 3.6555, 3.5929, 3.6904, 3.7290, 3.4595], + device='cuda:0'), covar=tensor([0.0027, 0.0065, 0.0035, 0.0037, 0.0031, 0.0027, 0.0037, 0.0057], + device='cuda:0'), in_proj_covar=tensor([0.0029, 0.0039, 0.0035, 0.0033, 0.0033, 0.0035, 0.0041, 0.0043], + device='cuda:0'), out_proj_covar=tensor([7.5674e-05, 1.1590e-04, 1.0395e-04, 8.7292e-05, 8.7037e-05, 9.6701e-05, + 1.2169e-04, 1.1752e-04], device='cuda:0') +2023-03-20 21:31:01,559 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.88 vs. limit=2.0 +2023-03-20 21:31:02,783 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=28585.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:31:05,750 INFO [train.py:901] (0/2) Epoch 11, batch 350, loss[loss=0.1581, simple_loss=0.2257, pruned_loss=0.04527, over 7226.00 frames. ], tot_loss[loss=0.1831, simple_loss=0.2523, pruned_loss=0.05692, over 1194138.44 frames. ], batch size: 45, lr: 1.49e-02, grad_scale: 16.0 +2023-03-20 21:31:16,853 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.554e+02 2.672e+02 3.200e+02 3.823e+02 8.890e+02, threshold=6.400e+02, percent-clipped=4.0 +2023-03-20 21:31:16,948 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=28611.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 21:31:22,967 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-20 21:31:30,601 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8170, 4.5053, 4.4971, 4.9991, 4.9617, 4.9770, 4.5421, 4.6129], + device='cuda:0'), covar=tensor([0.0869, 0.2554, 0.1987, 0.1083, 0.0659, 0.1125, 0.0749, 0.1034], + device='cuda:0'), in_proj_covar=tensor([0.0171, 0.0281, 0.0228, 0.0220, 0.0171, 0.0277, 0.0163, 0.0196], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:31:32,127 INFO [train.py:901] (0/2) Epoch 11, batch 400, loss[loss=0.1911, simple_loss=0.2632, pruned_loss=0.05949, over 7284.00 frames. ], tot_loss[loss=0.1836, simple_loss=0.2528, pruned_loss=0.0572, over 1250882.12 frames. ], batch size: 77, lr: 1.49e-02, grad_scale: 16.0 +2023-03-20 21:31:33,272 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=28643.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:31:47,984 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8174, 1.9377, 1.8232, 3.0242, 2.4985, 2.8357, 2.6034, 2.3628], + device='cuda:0'), covar=tensor([0.1497, 0.0675, 0.2049, 0.0406, 0.0056, 0.0064, 0.0112, 0.0100], + device='cuda:0'), in_proj_covar=tensor([0.0289, 0.0221, 0.0270, 0.0239, 0.0123, 0.0115, 0.0138, 0.0145], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0002, 0.0001, 0.0001, 0.0001, 0.0002], + device='cuda:0') +2023-03-20 21:31:58,572 INFO [train.py:901] (0/2) Epoch 11, batch 450, loss[loss=0.1678, simple_loss=0.2377, pruned_loss=0.04898, over 7203.00 frames. ], tot_loss[loss=0.182, simple_loss=0.2509, pruned_loss=0.05658, over 1290095.89 frames. ], batch size: 45, lr: 1.49e-02, grad_scale: 32.0 +2023-03-20 21:32:05,655 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-20 21:32:06,173 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-20 21:32:08,669 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.751e+02 2.433e+02 3.005e+02 3.670e+02 6.203e+02, threshold=6.010e+02, percent-clipped=0.0 +2023-03-20 21:32:23,923 INFO [train.py:901] (0/2) Epoch 11, batch 500, loss[loss=0.2037, simple_loss=0.2744, pruned_loss=0.06646, over 7333.00 frames. ], tot_loss[loss=0.1822, simple_loss=0.2514, pruned_loss=0.05653, over 1321837.67 frames. ], batch size: 54, lr: 1.48e-02, grad_scale: 16.0 +2023-03-20 21:32:27,073 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0758, 3.7119, 3.7845, 3.6746, 3.5154, 3.7075, 4.0180, 3.5899], + device='cuda:0'), covar=tensor([0.0089, 0.0122, 0.0116, 0.0124, 0.0317, 0.0093, 0.0125, 0.0127], + device='cuda:0'), in_proj_covar=tensor([0.0052, 0.0060, 0.0061, 0.0051, 0.0090, 0.0065, 0.0064, 0.0061], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:32:27,829 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-03-20 21:32:34,098 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3330, 2.4739, 2.0185, 2.2109, 2.4927, 2.2558, 2.4995, 2.4900], + device='cuda:0'), covar=tensor([0.1675, 0.0566, 0.0748, 0.1311, 0.0702, 0.0589, 0.1342, 0.1001], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0033, 0.0040, 0.0036, 0.0040, 0.0037, 0.0036, 0.0035], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 21:32:39,607 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-20 21:32:41,079 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-20 21:32:42,086 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-20 21:32:44,573 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-20 21:32:49,696 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-20 21:32:50,203 INFO [train.py:901] (0/2) Epoch 11, batch 550, loss[loss=0.1713, simple_loss=0.2485, pruned_loss=0.04708, over 7277.00 frames. ], tot_loss[loss=0.1816, simple_loss=0.2508, pruned_loss=0.05615, over 1348863.31 frames. ], batch size: 70, lr: 1.48e-02, grad_scale: 16.0 +2023-03-20 21:33:00,823 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-20 21:33:01,276 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.735e+02 2.650e+02 3.021e+02 3.738e+02 7.326e+02, threshold=6.042e+02, percent-clipped=1.0 +2023-03-20 21:33:02,893 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0545, 3.7677, 3.7530, 4.2520, 4.2768, 4.2269, 3.8116, 3.8545], + device='cuda:0'), covar=tensor([0.1017, 0.2307, 0.2068, 0.1142, 0.0730, 0.1202, 0.0917, 0.0998], + device='cuda:0'), in_proj_covar=tensor([0.0170, 0.0276, 0.0224, 0.0211, 0.0171, 0.0270, 0.0158, 0.0190], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:33:08,444 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7991, 3.5688, 3.5604, 3.8713, 3.6264, 3.8600, 3.7574, 3.6839], + device='cuda:0'), covar=tensor([0.0029, 0.0074, 0.0041, 0.0033, 0.0032, 0.0030, 0.0038, 0.0052], + device='cuda:0'), in_proj_covar=tensor([0.0030, 0.0039, 0.0036, 0.0034, 0.0033, 0.0035, 0.0042, 0.0043], + device='cuda:0'), out_proj_covar=tensor([7.7244e-05, 1.1808e-04, 1.0589e-04, 8.9486e-05, 8.7272e-05, 9.6121e-05, + 1.2431e-04, 1.1722e-04], device='cuda:0') +2023-03-20 21:33:08,832 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-20 21:33:12,929 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-20 21:33:14,115 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0220, 3.9354, 3.1198, 4.4678, 3.6535, 4.0543, 2.7125, 2.6304], + device='cuda:0'), covar=tensor([0.0114, 0.0189, 0.0905, 0.0154, 0.0174, 0.0333, 0.1177, 0.1136], + device='cuda:0'), in_proj_covar=tensor([0.0190, 0.0210, 0.0298, 0.0196, 0.0228, 0.0215, 0.0269, 0.0285], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 21:33:16,516 INFO [train.py:901] (0/2) Epoch 11, batch 600, loss[loss=0.1792, simple_loss=0.2547, pruned_loss=0.05185, over 7291.00 frames. ], tot_loss[loss=0.1827, simple_loss=0.2518, pruned_loss=0.05679, over 1370768.02 frames. ], batch size: 68, lr: 1.48e-02, grad_scale: 16.0 +2023-03-20 21:33:17,097 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=28842.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:33:19,525 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-20 21:33:20,081 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=28848.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:33:27,386 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=28861.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:33:33,876 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8861, 3.4123, 3.7749, 4.0386, 3.8127, 3.9384, 3.6507, 3.7810], + device='cuda:0'), covar=tensor([0.0032, 0.0082, 0.0033, 0.0029, 0.0030, 0.0029, 0.0047, 0.0045], + device='cuda:0'), in_proj_covar=tensor([0.0030, 0.0040, 0.0035, 0.0033, 0.0033, 0.0035, 0.0042, 0.0043], + device='cuda:0'), out_proj_covar=tensor([7.7457e-05, 1.1903e-04, 1.0443e-04, 8.8246e-05, 8.5988e-05, 9.6074e-05, + 1.2532e-04, 1.1668e-04], device='cuda:0') +2023-03-20 21:33:36,261 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-20 21:33:39,288 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=28885.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:33:42,189 INFO [train.py:901] (0/2) Epoch 11, batch 650, loss[loss=0.1378, simple_loss=0.2009, pruned_loss=0.03739, over 7022.00 frames. ], tot_loss[loss=0.1823, simple_loss=0.2515, pruned_loss=0.05654, over 1386803.22 frames. ], batch size: 35, lr: 1.48e-02, grad_scale: 16.0 +2023-03-20 21:33:45,162 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-20 21:33:52,306 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=28911.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 21:33:52,675 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.509e+02 2.625e+02 3.059e+02 3.685e+02 8.000e+02, threshold=6.119e+02, percent-clipped=1.0 +2023-03-20 21:33:58,550 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=28922.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:34:01,943 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-20 21:34:03,913 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=28933.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:34:07,919 INFO [train.py:901] (0/2) Epoch 11, batch 700, loss[loss=0.2145, simple_loss=0.2808, pruned_loss=0.07411, over 7135.00 frames. ], tot_loss[loss=0.183, simple_loss=0.252, pruned_loss=0.05699, over 1398658.74 frames. ], batch size: 98, lr: 1.48e-02, grad_scale: 16.0 +2023-03-20 21:34:09,739 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=28943.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:34:11,129 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-20 21:34:17,764 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=28959.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 21:34:33,881 INFO [train.py:901] (0/2) Epoch 11, batch 750, loss[loss=0.1849, simple_loss=0.2579, pruned_loss=0.05593, over 7361.00 frames. ], tot_loss[loss=0.1832, simple_loss=0.2522, pruned_loss=0.0571, over 1409436.25 frames. ], batch size: 63, lr: 1.48e-02, grad_scale: 16.0 +2023-03-20 21:34:33,940 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=28991.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:34:34,427 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-20 21:34:34,950 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-20 21:34:35,638 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7684, 3.4957, 3.3356, 3.5614, 2.8262, 2.5782, 3.7345, 2.9128], + device='cuda:0'), covar=tensor([0.0115, 0.0127, 0.0198, 0.0168, 0.0253, 0.0364, 0.0172, 0.0583], + device='cuda:0'), in_proj_covar=tensor([0.0264, 0.0273, 0.0235, 0.0269, 0.0294, 0.0291, 0.0270, 0.0291], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-20 21:34:36,087 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.9977, 0.9131, 1.0793, 1.2152, 1.1769, 1.4029, 0.8841, 1.1654], + device='cuda:0'), covar=tensor([0.1125, 0.1803, 0.0618, 0.0698, 0.0970, 0.0698, 0.0523, 0.1190], + device='cuda:0'), in_proj_covar=tensor([0.0036, 0.0040, 0.0030, 0.0033, 0.0034, 0.0032, 0.0041, 0.0035], + device='cuda:0'), out_proj_covar=tensor([8.0412e-05, 9.2676e-05, 6.8498e-05, 7.3063e-05, 7.7357e-05, 7.7109e-05, + 9.0857e-05, 8.1365e-05], device='cuda:0') +2023-03-20 21:34:40,585 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=29002.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:34:45,595 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.603e+02 2.567e+02 3.046e+02 3.773e+02 7.395e+02, threshold=6.092e+02, percent-clipped=4.0 +2023-03-20 21:34:50,168 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-20 21:34:55,275 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-20 21:35:00,678 INFO [train.py:901] (0/2) Epoch 11, batch 800, loss[loss=0.2132, simple_loss=0.2807, pruned_loss=0.07289, over 7274.00 frames. ], tot_loss[loss=0.1843, simple_loss=0.2534, pruned_loss=0.05762, over 1416561.14 frames. ], batch size: 66, lr: 1.48e-02, grad_scale: 16.0 +2023-03-20 21:35:00,687 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-20 21:35:02,180 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-20 21:35:10,341 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8280, 4.6415, 4.4621, 5.0100, 4.9139, 5.0378, 4.5630, 4.6862], + device='cuda:0'), covar=tensor([0.0825, 0.1869, 0.1737, 0.0911, 0.0631, 0.0919, 0.0475, 0.0742], + device='cuda:0'), in_proj_covar=tensor([0.0174, 0.0284, 0.0226, 0.0216, 0.0173, 0.0277, 0.0158, 0.0193], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:35:12,004 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=29063.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:35:12,362 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-20 21:35:20,222 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.86 vs. limit=2.0 +2023-03-20 21:35:26,630 INFO [train.py:901] (0/2) Epoch 11, batch 850, loss[loss=0.1675, simple_loss=0.239, pruned_loss=0.04794, over 7302.00 frames. ], tot_loss[loss=0.1836, simple_loss=0.2529, pruned_loss=0.05709, over 1423525.92 frames. ], batch size: 66, lr: 1.48e-02, grad_scale: 16.0 +2023-03-20 21:35:31,623 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.93 vs. limit=2.0 +2023-03-20 21:35:31,870 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-20 21:35:32,383 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-20 21:35:34,998 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7108, 4.4612, 4.4409, 4.9422, 4.9390, 4.9636, 4.3902, 4.5218], + device='cuda:0'), covar=tensor([0.0845, 0.2292, 0.1970, 0.1023, 0.0559, 0.1131, 0.0836, 0.0939], + device='cuda:0'), in_proj_covar=tensor([0.0174, 0.0284, 0.0227, 0.0215, 0.0172, 0.0278, 0.0158, 0.0195], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:35:37,484 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.651e+02 2.474e+02 2.954e+02 3.658e+02 5.777e+02, threshold=5.907e+02, percent-clipped=0.0 +2023-03-20 21:35:37,511 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-20 21:35:41,686 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-20 21:35:52,629 INFO [train.py:901] (0/2) Epoch 11, batch 900, loss[loss=0.2036, simple_loss=0.262, pruned_loss=0.07263, over 7230.00 frames. ], tot_loss[loss=0.1844, simple_loss=0.2536, pruned_loss=0.05764, over 1429745.77 frames. ], batch size: 45, lr: 1.47e-02, grad_scale: 16.0 +2023-03-20 21:35:53,222 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=29142.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:35:56,292 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=29148.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:36:17,962 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=29190.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:36:18,429 INFO [train.py:901] (0/2) Epoch 11, batch 950, loss[loss=0.1938, simple_loss=0.2574, pruned_loss=0.06509, over 7359.00 frames. ], tot_loss[loss=0.1834, simple_loss=0.2526, pruned_loss=0.05711, over 1432328.94 frames. ], batch size: 73, lr: 1.47e-02, grad_scale: 16.0 +2023-03-20 21:36:19,441 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-20 21:36:21,020 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=29196.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:36:30,012 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.761e+02 2.655e+02 3.153e+02 4.188e+02 9.756e+02, threshold=6.306e+02, percent-clipped=6.0 +2023-03-20 21:36:32,615 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=29217.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:36:42,189 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=29236.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:36:44,574 INFO [train.py:901] (0/2) Epoch 11, batch 1000, loss[loss=0.1709, simple_loss=0.2422, pruned_loss=0.04982, over 7171.00 frames. ], tot_loss[loss=0.1827, simple_loss=0.2516, pruned_loss=0.05693, over 1429565.42 frames. ], batch size: 41, lr: 1.47e-02, grad_scale: 16.0 +2023-03-20 21:36:44,624 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-20 21:37:01,729 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.04 vs. limit=5.0 +2023-03-20 21:37:05,462 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-20 21:37:11,095 INFO [train.py:901] (0/2) Epoch 11, batch 1050, loss[loss=0.1717, simple_loss=0.2512, pruned_loss=0.04616, over 7320.00 frames. ], tot_loss[loss=0.182, simple_loss=0.2511, pruned_loss=0.05644, over 1431747.22 frames. ], batch size: 80, lr: 1.47e-02, grad_scale: 16.0 +2023-03-20 21:37:12,406 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.28 vs. limit=5.0 +2023-03-20 21:37:14,287 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=29297.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:37:21,637 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.563e+02 2.500e+02 3.009e+02 3.529e+02 7.221e+02, threshold=6.018e+02, percent-clipped=4.0 +2023-03-20 21:37:27,678 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-20 21:37:31,228 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-20 21:37:36,255 INFO [train.py:901] (0/2) Epoch 11, batch 1100, loss[loss=0.1836, simple_loss=0.257, pruned_loss=0.05505, over 7275.00 frames. ], tot_loss[loss=0.1816, simple_loss=0.2508, pruned_loss=0.05619, over 1433225.39 frames. ], batch size: 66, lr: 1.47e-02, grad_scale: 16.0 +2023-03-20 21:37:45,429 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=29358.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:38:00,610 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-20 21:38:01,098 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 21:38:01,743 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8250, 2.0393, 1.9760, 3.0339, 1.4043, 2.6424, 1.1938, 2.6926], + device='cuda:0'), covar=tensor([0.0049, 0.0845, 0.1594, 0.0053, 0.3838, 0.0033, 0.1037, 0.0123], + device='cuda:0'), in_proj_covar=tensor([0.0138, 0.0264, 0.0308, 0.0142, 0.0297, 0.0145, 0.0268, 0.0193], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 21:38:02,627 INFO [train.py:901] (0/2) Epoch 11, batch 1150, loss[loss=0.2023, simple_loss=0.2739, pruned_loss=0.06533, over 7312.00 frames. ], tot_loss[loss=0.1817, simple_loss=0.2515, pruned_loss=0.05599, over 1437436.97 frames. ], batch size: 80, lr: 1.47e-02, grad_scale: 16.0 +2023-03-20 21:38:13,507 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.646e+02 2.488e+02 2.965e+02 3.681e+02 7.032e+02, threshold=5.930e+02, percent-clipped=2.0 +2023-03-20 21:38:13,555 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-20 21:38:14,542 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-20 21:38:28,823 INFO [train.py:901] (0/2) Epoch 11, batch 1200, loss[loss=0.188, simple_loss=0.2634, pruned_loss=0.05628, over 7296.00 frames. ], tot_loss[loss=0.1819, simple_loss=0.2514, pruned_loss=0.05616, over 1438664.47 frames. ], batch size: 77, lr: 1.47e-02, grad_scale: 16.0 +2023-03-20 21:38:34,769 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.80 vs. limit=5.0 +2023-03-20 21:38:47,630 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-20 21:38:54,678 INFO [train.py:901] (0/2) Epoch 11, batch 1250, loss[loss=0.1918, simple_loss=0.2604, pruned_loss=0.06163, over 7317.00 frames. ], tot_loss[loss=0.1825, simple_loss=0.2518, pruned_loss=0.05662, over 1438072.98 frames. ], batch size: 75, lr: 1.47e-02, grad_scale: 16.0 +2023-03-20 21:39:05,277 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.616e+02 2.504e+02 3.020e+02 3.915e+02 7.336e+02, threshold=6.040e+02, percent-clipped=4.0 +2023-03-20 21:39:07,921 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=29517.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:39:12,861 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-20 21:39:17,028 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-20 21:39:18,022 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-20 21:39:20,483 INFO [train.py:901] (0/2) Epoch 11, batch 1300, loss[loss=0.1844, simple_loss=0.2557, pruned_loss=0.0566, over 7277.00 frames. ], tot_loss[loss=0.1826, simple_loss=0.2519, pruned_loss=0.05662, over 1439248.05 frames. ], batch size: 70, lr: 1.47e-02, grad_scale: 8.0 +2023-03-20 21:39:26,245 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=29551.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:39:33,263 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=29565.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:39:41,742 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-20 21:39:43,795 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-20 21:39:45,415 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8615, 3.6930, 3.6022, 3.5226, 3.0542, 3.5780, 3.6241, 3.6381], + device='cuda:0'), covar=tensor([0.0157, 0.0192, 0.0176, 0.0219, 0.0431, 0.0172, 0.0257, 0.0154], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0062, 0.0064, 0.0052, 0.0093, 0.0066, 0.0065, 0.0062], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:39:46,286 INFO [train.py:901] (0/2) Epoch 11, batch 1350, loss[loss=0.1817, simple_loss=0.2437, pruned_loss=0.05988, over 7232.00 frames. ], tot_loss[loss=0.1827, simple_loss=0.2518, pruned_loss=0.0568, over 1438552.71 frames. ], batch size: 45, lr: 1.46e-02, grad_scale: 8.0 +2023-03-20 21:39:46,865 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=29592.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:39:47,322 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-20 21:39:57,299 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=29612.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:39:57,657 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.715e+02 2.822e+02 3.499e+02 4.170e+02 8.091e+02, threshold=6.998e+02, percent-clipped=6.0 +2023-03-20 21:39:58,849 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-20 21:39:59,996 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7114, 2.1433, 1.9010, 3.0985, 1.3618, 2.7516, 1.1558, 2.7769], + device='cuda:0'), covar=tensor([0.0056, 0.0877, 0.1660, 0.0055, 0.3705, 0.0052, 0.1143, 0.0193], + device='cuda:0'), in_proj_covar=tensor([0.0140, 0.0268, 0.0310, 0.0143, 0.0299, 0.0147, 0.0271, 0.0193], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 21:40:04,764 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.89 vs. limit=2.0 +2023-03-20 21:40:12,599 INFO [train.py:901] (0/2) Epoch 11, batch 1400, loss[loss=0.1855, simple_loss=0.2571, pruned_loss=0.05699, over 7312.00 frames. ], tot_loss[loss=0.1827, simple_loss=0.2519, pruned_loss=0.05677, over 1440024.89 frames. ], batch size: 83, lr: 1.46e-02, grad_scale: 8.0 +2023-03-20 21:40:21,811 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=29658.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:40:23,816 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=29662.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:40:32,187 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8319, 3.5764, 3.6290, 3.6492, 3.6319, 3.7696, 3.7945, 3.7359], + device='cuda:0'), covar=tensor([0.0032, 0.0073, 0.0037, 0.0038, 0.0037, 0.0028, 0.0040, 0.0047], + device='cuda:0'), in_proj_covar=tensor([0.0032, 0.0042, 0.0037, 0.0036, 0.0036, 0.0037, 0.0044, 0.0045], + device='cuda:0'), out_proj_covar=tensor([8.1175e-05, 1.2370e-04, 1.0846e-04, 9.3559e-05, 9.3923e-05, 9.9597e-05, + 1.2858e-04, 1.2138e-04], device='cuda:0') +2023-03-20 21:40:33,105 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-20 21:40:38,175 INFO [train.py:901] (0/2) Epoch 11, batch 1450, loss[loss=0.1787, simple_loss=0.2509, pruned_loss=0.05329, over 7345.00 frames. ], tot_loss[loss=0.1819, simple_loss=0.2515, pruned_loss=0.05618, over 1439705.10 frames. ], batch size: 54, lr: 1.46e-02, grad_scale: 8.0 +2023-03-20 21:40:40,652 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.84 vs. limit=2.0 +2023-03-20 21:40:46,586 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=29706.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:40:49,930 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.574e+02 2.541e+02 3.078e+02 3.922e+02 1.429e+03, threshold=6.155e+02, percent-clipped=1.0 +2023-03-20 21:40:55,018 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=29723.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 21:40:56,686 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.95 vs. limit=2.0 +2023-03-20 21:40:57,409 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-20 21:41:04,513 INFO [train.py:901] (0/2) Epoch 11, batch 1500, loss[loss=0.2026, simple_loss=0.2719, pruned_loss=0.06664, over 6705.00 frames. ], tot_loss[loss=0.1831, simple_loss=0.2526, pruned_loss=0.05679, over 1440902.00 frames. ], batch size: 106, lr: 1.46e-02, grad_scale: 8.0 +2023-03-20 21:41:13,023 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-20 21:41:15,590 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2291, 2.9864, 2.1529, 2.4998, 2.8783, 2.0794, 2.7626, 2.7088], + device='cuda:0'), covar=tensor([0.1378, 0.0524, 0.1385, 0.0938, 0.1883, 0.1185, 0.1189, 0.1170], + device='cuda:0'), in_proj_covar=tensor([0.0036, 0.0035, 0.0041, 0.0036, 0.0038, 0.0038, 0.0035, 0.0036], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 21:41:20,652 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9451, 3.5194, 3.6249, 3.5155, 3.3267, 3.5362, 3.7620, 3.3373], + device='cuda:0'), covar=tensor([0.0130, 0.0166, 0.0137, 0.0159, 0.0345, 0.0111, 0.0172, 0.0144], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0062, 0.0065, 0.0052, 0.0095, 0.0067, 0.0064, 0.0062], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:41:30,362 INFO [train.py:901] (0/2) Epoch 11, batch 1550, loss[loss=0.2023, simple_loss=0.2711, pruned_loss=0.06672, over 7287.00 frames. ], tot_loss[loss=0.1817, simple_loss=0.2514, pruned_loss=0.05605, over 1441257.31 frames. ], batch size: 57, lr: 1.46e-02, grad_scale: 8.0 +2023-03-20 21:41:36,517 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=29802.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:41:37,408 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-20 21:41:41,949 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.731e+02 2.521e+02 3.160e+02 3.849e+02 9.328e+02, threshold=6.319e+02, percent-clipped=4.0 +2023-03-20 21:41:52,013 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6194, 3.3851, 3.4178, 3.1266, 2.8645, 3.3314, 3.3455, 3.2519], + device='cuda:0'), covar=tensor([0.0156, 0.0232, 0.0162, 0.0308, 0.0420, 0.0165, 0.0248, 0.0160], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0061, 0.0064, 0.0052, 0.0094, 0.0066, 0.0064, 0.0062], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:41:56,882 INFO [train.py:901] (0/2) Epoch 11, batch 1600, loss[loss=0.1488, simple_loss=0.2232, pruned_loss=0.03716, over 7261.00 frames. ], tot_loss[loss=0.1819, simple_loss=0.2514, pruned_loss=0.05617, over 1438932.41 frames. ], batch size: 47, lr: 1.46e-02, grad_scale: 8.0 +2023-03-20 21:41:59,530 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1206, 1.2566, 0.8925, 1.2563, 1.0820, 0.8130, 0.8550, 0.8043], + device='cuda:0'), covar=tensor([0.0148, 0.0080, 0.0253, 0.0069, 0.0091, 0.0137, 0.0110, 0.0147], + device='cuda:0'), in_proj_covar=tensor([0.0019, 0.0019, 0.0019, 0.0018, 0.0021, 0.0019, 0.0020, 0.0025], + device='cuda:0'), out_proj_covar=tensor([2.3691e-05, 2.2191e-05, 2.3622e-05, 2.0941e-05, 2.6299e-05, 2.2284e-05, + 2.3806e-05, 3.0537e-05], device='cuda:0') +2023-03-20 21:42:08,206 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-20 21:42:08,316 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=29863.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:42:09,209 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-20 21:42:11,800 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-20 21:42:21,951 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-20 21:42:22,963 INFO [train.py:901] (0/2) Epoch 11, batch 1650, loss[loss=0.196, simple_loss=0.2614, pruned_loss=0.06533, over 7245.00 frames. ], tot_loss[loss=0.1822, simple_loss=0.2519, pruned_loss=0.05624, over 1441486.73 frames. ], batch size: 89, lr: 1.46e-02, grad_scale: 8.0 +2023-03-20 21:42:23,570 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=29892.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:42:24,552 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5195, 3.7707, 3.5150, 3.7644, 3.4702, 3.6580, 3.9579, 4.0076], + device='cuda:0'), covar=tensor([0.0244, 0.0138, 0.0200, 0.0170, 0.0298, 0.0309, 0.0246, 0.0194], + device='cuda:0'), in_proj_covar=tensor([0.0098, 0.0099, 0.0093, 0.0103, 0.0096, 0.0081, 0.0079, 0.0082], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:42:25,982 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-20 21:42:31,782 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=29907.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:42:34,722 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.522e+02 2.457e+02 3.036e+02 3.652e+02 9.046e+02, threshold=6.071e+02, percent-clipped=4.0 +2023-03-20 21:42:34,749 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-20 21:42:34,847 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5339, 3.8227, 4.0592, 3.9746, 3.9391, 4.1216, 4.3168, 3.7913], + device='cuda:0'), covar=tensor([0.0089, 0.0147, 0.0135, 0.0138, 0.0250, 0.0088, 0.0123, 0.0143], + device='cuda:0'), in_proj_covar=tensor([0.0055, 0.0064, 0.0066, 0.0054, 0.0098, 0.0068, 0.0066, 0.0064], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:42:48,513 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=29940.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:42:49,002 INFO [train.py:901] (0/2) Epoch 11, batch 1700, loss[loss=0.1391, simple_loss=0.1911, pruned_loss=0.04353, over 6031.00 frames. ], tot_loss[loss=0.1815, simple_loss=0.2512, pruned_loss=0.05592, over 1439703.65 frames. ], batch size: 26, lr: 1.46e-02, grad_scale: 8.0 +2023-03-20 21:42:51,549 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 21:42:55,600 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-20 21:43:06,917 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-20 21:43:15,700 INFO [train.py:901] (0/2) Epoch 11, batch 1750, loss[loss=0.2121, simple_loss=0.2838, pruned_loss=0.07019, over 7330.00 frames. ], tot_loss[loss=0.1809, simple_loss=0.2507, pruned_loss=0.05554, over 1440739.18 frames. ], batch size: 75, lr: 1.45e-02, grad_scale: 8.0 +2023-03-20 21:43:27,040 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.718e+02 2.468e+02 2.815e+02 3.490e+02 6.653e+02, threshold=5.630e+02, percent-clipped=1.0 +2023-03-20 21:43:29,596 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=30018.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 21:43:31,998 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-20 21:43:32,985 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-20 21:43:41,056 INFO [train.py:901] (0/2) Epoch 11, batch 1800, loss[loss=0.1881, simple_loss=0.2591, pruned_loss=0.05853, over 7273.00 frames. ], tot_loss[loss=0.1802, simple_loss=0.2501, pruned_loss=0.05516, over 1439599.85 frames. ], batch size: 77, lr: 1.45e-02, grad_scale: 8.0 +2023-03-20 21:43:48,429 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5381, 3.4554, 3.5161, 3.0946, 3.6925, 3.3650, 2.6515, 3.6086], + device='cuda:0'), covar=tensor([0.1101, 0.0394, 0.1130, 0.3092, 0.0980, 0.1008, 0.4260, 0.1449], + device='cuda:0'), in_proj_covar=tensor([0.0031, 0.0032, 0.0031, 0.0033, 0.0030, 0.0028, 0.0042, 0.0032], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 21:43:54,766 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-20 21:44:07,479 INFO [train.py:901] (0/2) Epoch 11, batch 1850, loss[loss=0.1528, simple_loss=0.2194, pruned_loss=0.04309, over 7144.00 frames. ], tot_loss[loss=0.1812, simple_loss=0.2509, pruned_loss=0.05577, over 1438758.25 frames. ], batch size: 39, lr: 1.45e-02, grad_scale: 8.0 +2023-03-20 21:44:08,527 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-20 21:44:14,581 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0696, 2.0909, 2.0657, 2.8739, 1.4038, 3.3698, 1.3121, 2.7296], + device='cuda:0'), covar=tensor([0.0062, 0.0939, 0.1571, 0.0051, 0.4429, 0.0058, 0.1051, 0.0135], + device='cuda:0'), in_proj_covar=tensor([0.0146, 0.0274, 0.0318, 0.0149, 0.0304, 0.0151, 0.0276, 0.0199], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 21:44:17,459 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-20 21:44:18,417 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.751e+02 2.849e+02 3.186e+02 3.684e+02 5.512e+02, threshold=6.372e+02, percent-clipped=0.0 +2023-03-20 21:44:19,122 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2628, 2.3294, 2.1380, 2.9705, 1.3774, 3.4456, 1.2744, 2.7273], + device='cuda:0'), covar=tensor([0.0055, 0.0851, 0.1725, 0.0049, 0.4687, 0.0047, 0.1072, 0.0079], + device='cuda:0'), in_proj_covar=tensor([0.0145, 0.0272, 0.0317, 0.0149, 0.0303, 0.0150, 0.0275, 0.0198], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 21:44:24,700 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=30125.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:44:33,228 INFO [train.py:901] (0/2) Epoch 11, batch 1900, loss[loss=0.1836, simple_loss=0.2619, pruned_loss=0.0527, over 7283.00 frames. ], tot_loss[loss=0.1811, simple_loss=0.251, pruned_loss=0.05557, over 1442247.94 frames. ], batch size: 66, lr: 1.45e-02, grad_scale: 8.0 +2023-03-20 21:44:34,253 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-20 21:44:41,823 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=30158.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:44:56,566 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=30186.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:44:58,938 INFO [train.py:901] (0/2) Epoch 11, batch 1950, loss[loss=0.1947, simple_loss=0.2568, pruned_loss=0.06632, over 7222.00 frames. ], tot_loss[loss=0.182, simple_loss=0.2518, pruned_loss=0.05611, over 1443914.40 frames. ], batch size: 45, lr: 1.45e-02, grad_scale: 8.0 +2023-03-20 21:44:59,963 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-20 21:45:07,579 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=30207.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:45:08,161 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9084, 2.2034, 1.8145, 2.5874, 2.7031, 2.2752, 2.0329, 2.0556], + device='cuda:0'), covar=tensor([0.1769, 0.0905, 0.2642, 0.0601, 0.0106, 0.0046, 0.0123, 0.0120], + device='cuda:0'), in_proj_covar=tensor([0.0299, 0.0227, 0.0280, 0.0253, 0.0129, 0.0120, 0.0145, 0.0161], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0001, 0.0001, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:45:08,652 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=30209.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:45:10,567 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.824e+02 2.624e+02 3.249e+02 3.958e+02 8.410e+02, threshold=6.499e+02, percent-clipped=2.0 +2023-03-20 21:45:11,130 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-20 21:45:16,917 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-20 21:45:17,418 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-20 21:45:25,564 INFO [train.py:901] (0/2) Epoch 11, batch 2000, loss[loss=0.1669, simple_loss=0.2432, pruned_loss=0.04535, over 7213.00 frames. ], tot_loss[loss=0.1811, simple_loss=0.2512, pruned_loss=0.05549, over 1444359.72 frames. ], batch size: 93, lr: 1.45e-02, grad_scale: 8.0 +2023-03-20 21:45:31,286 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5509, 3.2073, 3.4952, 3.5631, 3.4772, 3.4769, 3.3406, 3.4785], + device='cuda:0'), covar=tensor([0.0027, 0.0072, 0.0033, 0.0032, 0.0031, 0.0036, 0.0059, 0.0042], + device='cuda:0'), in_proj_covar=tensor([0.0032, 0.0041, 0.0038, 0.0036, 0.0035, 0.0037, 0.0043, 0.0044], + device='cuda:0'), out_proj_covar=tensor([7.9855e-05, 1.2248e-04, 1.1001e-04, 9.3109e-05, 8.9823e-05, 9.7556e-05, + 1.2526e-04, 1.1814e-04], device='cuda:0') +2023-03-20 21:45:33,274 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=30255.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:45:33,752 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-20 21:45:40,873 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=30270.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:45:44,341 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-20 21:45:51,250 INFO [train.py:901] (0/2) Epoch 11, batch 2050, loss[loss=0.17, simple_loss=0.2506, pruned_loss=0.04472, over 7299.00 frames. ], tot_loss[loss=0.1809, simple_loss=0.2512, pruned_loss=0.05533, over 1446125.83 frames. ], batch size: 59, lr: 1.45e-02, grad_scale: 8.0 +2023-03-20 21:45:52,753 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-20 21:45:52,853 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8560, 3.4841, 3.7203, 3.6578, 3.5602, 3.7456, 3.7200, 3.6289], + device='cuda:0'), covar=tensor([0.0028, 0.0078, 0.0033, 0.0037, 0.0037, 0.0032, 0.0040, 0.0052], + device='cuda:0'), in_proj_covar=tensor([0.0031, 0.0042, 0.0037, 0.0036, 0.0035, 0.0037, 0.0042, 0.0044], + device='cuda:0'), out_proj_covar=tensor([7.9572e-05, 1.2340e-04, 1.0871e-04, 9.2908e-05, 9.0002e-05, 9.6980e-05, + 1.2391e-04, 1.1924e-04], device='cuda:0') +2023-03-20 21:46:03,016 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.718e+02 2.450e+02 2.835e+02 3.331e+02 9.536e+02, threshold=5.670e+02, percent-clipped=1.0 +2023-03-20 21:46:05,758 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=30318.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 21:46:06,531 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.95 vs. limit=2.0 +2023-03-20 21:46:07,803 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6352, 2.4819, 2.9229, 2.8505, 2.6891, 2.6123, 2.1127, 2.7517], + device='cuda:0'), covar=tensor([0.1932, 0.0493, 0.0882, 0.1299, 0.1108, 0.1113, 0.2691, 0.1456], + device='cuda:0'), in_proj_covar=tensor([0.0032, 0.0032, 0.0031, 0.0033, 0.0029, 0.0029, 0.0041, 0.0032], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0002, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 21:46:17,832 INFO [train.py:901] (0/2) Epoch 11, batch 2100, loss[loss=0.1728, simple_loss=0.255, pruned_loss=0.04534, over 7263.00 frames. ], tot_loss[loss=0.1812, simple_loss=0.2514, pruned_loss=0.05554, over 1446524.87 frames. ], batch size: 52, lr: 1.45e-02, grad_scale: 8.0 +2023-03-20 21:46:26,972 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-20 21:46:27,611 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=30360.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:46:30,474 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-20 21:46:30,532 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=30366.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:46:41,345 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.99 vs. limit=2.0 +2023-03-20 21:46:42,977 INFO [train.py:901] (0/2) Epoch 11, batch 2150, loss[loss=0.2118, simple_loss=0.268, pruned_loss=0.07784, over 7281.00 frames. ], tot_loss[loss=0.1821, simple_loss=0.2522, pruned_loss=0.056, over 1446642.45 frames. ], batch size: 66, lr: 1.45e-02, grad_scale: 8.0 +2023-03-20 21:46:45,557 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.79 vs. limit=2.0 +2023-03-20 21:46:54,913 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.318e+02 2.491e+02 3.235e+02 3.688e+02 7.373e+02, threshold=6.470e+02, percent-clipped=3.0 +2023-03-20 21:46:59,063 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=30421.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:47:09,684 INFO [train.py:901] (0/2) Epoch 11, batch 2200, loss[loss=0.1852, simple_loss=0.2597, pruned_loss=0.05534, over 7326.00 frames. ], tot_loss[loss=0.1812, simple_loss=0.2513, pruned_loss=0.05557, over 1444517.48 frames. ], batch size: 59, lr: 1.44e-02, grad_scale: 8.0 +2023-03-20 21:47:16,883 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-20 21:47:18,454 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=30458.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:47:30,607 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=30481.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:47:35,576 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.87 vs. limit=2.0 +2023-03-20 21:47:35,663 INFO [train.py:901] (0/2) Epoch 11, batch 2250, loss[loss=0.1418, simple_loss=0.2042, pruned_loss=0.03968, over 6958.00 frames. ], tot_loss[loss=0.1811, simple_loss=0.2511, pruned_loss=0.05553, over 1444884.05 frames. ], batch size: 35, lr: 1.44e-02, grad_scale: 8.0 +2023-03-20 21:47:43,311 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=30506.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:47:47,340 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.268e+02 2.464e+02 3.038e+02 3.870e+02 9.111e+02, threshold=6.075e+02, percent-clipped=2.0 +2023-03-20 21:47:48,179 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.01 vs. limit=2.0 +2023-03-20 21:47:50,862 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-20 21:47:50,874 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-20 21:48:00,682 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-20 21:48:01,422 INFO [train.py:901] (0/2) Epoch 11, batch 2300, loss[loss=0.18, simple_loss=0.2566, pruned_loss=0.0517, over 7197.00 frames. ], tot_loss[loss=0.1822, simple_loss=0.2521, pruned_loss=0.05614, over 1443135.18 frames. ], batch size: 50, lr: 1.44e-02, grad_scale: 8.0 +2023-03-20 21:48:02,930 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-20 21:48:14,268 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=30565.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:48:27,184 INFO [train.py:901] (0/2) Epoch 11, batch 2350, loss[loss=0.1801, simple_loss=0.2493, pruned_loss=0.0555, over 7342.00 frames. ], tot_loss[loss=0.1816, simple_loss=0.2516, pruned_loss=0.05574, over 1443308.02 frames. ], batch size: 63, lr: 1.44e-02, grad_scale: 8.0 +2023-03-20 21:48:39,217 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.593e+02 2.571e+02 3.025e+02 3.938e+02 7.428e+02, threshold=6.050e+02, percent-clipped=3.0 +2023-03-20 21:48:50,869 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-20 21:48:53,297 INFO [train.py:901] (0/2) Epoch 11, batch 2400, loss[loss=0.181, simple_loss=0.2588, pruned_loss=0.05166, over 7226.00 frames. ], tot_loss[loss=0.181, simple_loss=0.2512, pruned_loss=0.05541, over 1444072.72 frames. ], batch size: 93, lr: 1.44e-02, grad_scale: 8.0 +2023-03-20 21:48:57,974 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-20 21:49:01,171 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=30655.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:49:08,226 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-20 21:49:11,274 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-20 21:49:19,894 INFO [train.py:901] (0/2) Epoch 11, batch 2450, loss[loss=0.1929, simple_loss=0.2605, pruned_loss=0.06263, over 7298.00 frames. ], tot_loss[loss=0.1811, simple_loss=0.2514, pruned_loss=0.05536, over 1445623.44 frames. ], batch size: 49, lr: 1.44e-02, grad_scale: 8.0 +2023-03-20 21:49:31,054 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.590e+02 2.414e+02 2.901e+02 3.572e+02 6.150e+02, threshold=5.802e+02, percent-clipped=2.0 +2023-03-20 21:49:32,674 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=30716.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:49:32,731 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=30716.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 21:49:38,121 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-20 21:49:45,818 INFO [train.py:901] (0/2) Epoch 11, batch 2500, loss[loss=0.1721, simple_loss=0.235, pruned_loss=0.05465, over 7283.00 frames. ], tot_loss[loss=0.18, simple_loss=0.2506, pruned_loss=0.05471, over 1446206.15 frames. ], batch size: 47, lr: 1.44e-02, grad_scale: 8.0 +2023-03-20 21:50:03,742 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-20 21:50:06,830 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=30781.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:50:11,663 INFO [train.py:901] (0/2) Epoch 11, batch 2550, loss[loss=0.1577, simple_loss=0.2168, pruned_loss=0.04934, over 7004.00 frames. ], tot_loss[loss=0.1806, simple_loss=0.2511, pruned_loss=0.05501, over 1447483.89 frames. ], batch size: 35, lr: 1.44e-02, grad_scale: 8.0 +2023-03-20 21:50:22,944 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.481e+02 2.521e+02 2.930e+02 3.805e+02 7.129e+02, threshold=5.860e+02, percent-clipped=4.0 +2023-03-20 21:50:28,750 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4681, 3.7393, 4.1019, 4.0806, 3.8117, 4.0844, 4.2100, 3.5843], + device='cuda:0'), covar=tensor([0.0073, 0.0156, 0.0104, 0.0112, 0.0276, 0.0078, 0.0136, 0.0215], + device='cuda:0'), in_proj_covar=tensor([0.0056, 0.0065, 0.0065, 0.0053, 0.0101, 0.0070, 0.0068, 0.0066], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0001, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:50:31,311 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4681, 3.7058, 3.3937, 3.6114, 3.3715, 3.4253, 3.7146, 3.7075], + device='cuda:0'), covar=tensor([0.0340, 0.0303, 0.0336, 0.0399, 0.0529, 0.0462, 0.0425, 0.0447], + device='cuda:0'), in_proj_covar=tensor([0.0092, 0.0094, 0.0086, 0.0098, 0.0092, 0.0078, 0.0073, 0.0077], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:50:31,739 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=30829.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:50:37,736 INFO [train.py:901] (0/2) Epoch 11, batch 2600, loss[loss=0.1826, simple_loss=0.2515, pruned_loss=0.05683, over 7345.00 frames. ], tot_loss[loss=0.1802, simple_loss=0.2506, pruned_loss=0.05485, over 1447112.28 frames. ], batch size: 61, lr: 1.43e-02, grad_scale: 8.0 +2023-03-20 21:50:49,933 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=30865.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:50:59,009 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0633, 3.0692, 2.2110, 3.7301, 2.5095, 3.1392, 1.7928, 1.8154], + device='cuda:0'), covar=tensor([0.0180, 0.0312, 0.1493, 0.0236, 0.0274, 0.0356, 0.2126, 0.1283], + device='cuda:0'), in_proj_covar=tensor([0.0192, 0.0216, 0.0296, 0.0207, 0.0226, 0.0216, 0.0267, 0.0284], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 21:51:03,257 INFO [train.py:901] (0/2) Epoch 11, batch 2650, loss[loss=0.1773, simple_loss=0.2525, pruned_loss=0.05103, over 7287.00 frames. ], tot_loss[loss=0.1791, simple_loss=0.2497, pruned_loss=0.05422, over 1446277.05 frames. ], batch size: 86, lr: 1.43e-02, grad_scale: 8.0 +2023-03-20 21:51:09,379 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=30903.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:51:14,262 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.479e+02 2.453e+02 2.842e+02 3.553e+02 6.219e+02, threshold=5.685e+02, percent-clipped=2.0 +2023-03-20 21:51:14,318 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=30913.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:51:15,346 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6137, 2.5626, 2.9967, 2.8584, 2.9270, 2.4624, 2.2721, 2.6475], + device='cuda:0'), covar=tensor([0.2360, 0.0346, 0.1456, 0.1468, 0.1119, 0.1236, 0.2495, 0.2084], + device='cuda:0'), in_proj_covar=tensor([0.0032, 0.0032, 0.0031, 0.0032, 0.0029, 0.0028, 0.0041, 0.0030], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0001, 0.0001, 0.0002, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 21:51:20,302 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=30925.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:51:24,843 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=30934.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:51:28,162 INFO [train.py:901] (0/2) Epoch 11, batch 2700, loss[loss=0.2012, simple_loss=0.2723, pruned_loss=0.06508, over 7266.00 frames. ], tot_loss[loss=0.1789, simple_loss=0.2496, pruned_loss=0.05415, over 1446546.79 frames. ], batch size: 68, lr: 1.43e-02, grad_scale: 8.0 +2023-03-20 21:51:39,726 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=30964.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:51:50,704 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=30986.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:51:53,061 INFO [train.py:901] (0/2) Epoch 11, batch 2750, loss[loss=0.1667, simple_loss=0.2361, pruned_loss=0.04869, over 7358.00 frames. ], tot_loss[loss=0.1797, simple_loss=0.2498, pruned_loss=0.05478, over 1443422.94 frames. ], batch size: 44, lr: 1.43e-02, grad_scale: 8.0 +2023-03-20 21:51:55,147 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=30995.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:52:03,412 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=31011.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 21:52:03,481 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=31011.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 21:52:04,280 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.554e+02 2.469e+02 3.015e+02 3.883e+02 6.735e+02, threshold=6.029e+02, percent-clipped=4.0 +2023-03-20 21:52:05,803 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=31016.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:52:18,066 INFO [train.py:901] (0/2) Epoch 11, batch 2800, loss[loss=0.1943, simple_loss=0.2653, pruned_loss=0.06164, over 7345.00 frames. ], tot_loss[loss=0.1808, simple_loss=0.2508, pruned_loss=0.05535, over 1443596.34 frames. ], batch size: 73, lr: 1.43e-02, grad_scale: 8.0 +2023-03-20 21:52:31,332 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-11.pt +2023-03-20 21:52:50,127 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-20 21:52:50,180 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-20 21:52:50,206 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-20 21:52:50,229 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-20 21:52:50,403 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-20 21:52:50,419 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-20 21:52:50,430 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-20 21:52:50,530 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-20 21:52:50,605 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-20 21:52:50,621 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-20 21:52:50,637 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-20 21:52:51,176 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-20 21:52:51,181 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-20 21:52:51,188 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-20 21:52:51,328 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-20 21:52:51,331 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-20 21:52:51,390 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-20 21:52:53,166 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=31064.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:52:53,616 INFO [train.py:901] (0/2) Epoch 12, batch 0, loss[loss=0.1942, simple_loss=0.263, pruned_loss=0.06267, over 7355.00 frames. ], tot_loss[loss=0.1942, simple_loss=0.263, pruned_loss=0.06267, over 7355.00 frames. ], batch size: 63, lr: 1.38e-02, grad_scale: 8.0 +2023-03-20 21:52:53,618 INFO [train.py:926] (0/2) Computing validation loss +2023-03-20 21:52:58,021 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6978, 4.3598, 4.2499, 3.9566, 4.3014, 3.1826, 2.2837, 4.5380], + device='cuda:0'), covar=tensor([0.0018, 0.0061, 0.0032, 0.0045, 0.0025, 0.0296, 0.0476, 0.0030], + device='cuda:0'), in_proj_covar=tensor([0.0062, 0.0060, 0.0078, 0.0070, 0.0083, 0.0104, 0.0108, 0.0071], + device='cuda:0'), out_proj_covar=tensor([8.1462e-05, 8.8969e-05, 1.0513e-04, 1.0114e-04, 1.1331e-04, 1.4381e-04, + 1.4938e-04, 9.8415e-05], device='cuda:0') +2023-03-20 21:52:59,698 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3665, 4.6490, 4.5649, 4.6236, 4.4810, 4.2629, 4.6396, 4.4330], + device='cuda:0'), covar=tensor([0.0410, 0.0393, 0.0497, 0.0479, 0.0318, 0.0330, 0.0359, 0.0512], + device='cuda:0'), in_proj_covar=tensor([0.0157, 0.0174, 0.0127, 0.0127, 0.0112, 0.0162, 0.0141, 0.0107], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:53:20,416 INFO [train.py:935] (0/2) Epoch 12, validation: loss=0.1724, simple_loss=0.26, pruned_loss=0.04238, over 1622729.00 frames. +2023-03-20 21:53:20,416 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-20 21:53:24,066 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=31072.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 21:53:27,952 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-20 21:53:37,909 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-20 21:53:44,611 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-20 21:53:45,085 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.395e+02 2.570e+02 3.180e+02 3.808e+02 7.988e+02, threshold=6.360e+02, percent-clipped=6.0 +2023-03-20 21:53:46,115 INFO [train.py:901] (0/2) Epoch 12, batch 50, loss[loss=0.1712, simple_loss=0.2438, pruned_loss=0.04935, over 7338.00 frames. ], tot_loss[loss=0.1803, simple_loss=0.2517, pruned_loss=0.05445, over 325079.87 frames. ], batch size: 61, lr: 1.38e-02, grad_scale: 8.0 +2023-03-20 21:53:47,128 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-20 21:53:49,629 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-20 21:53:55,949 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.71 vs. limit=2.0 +2023-03-20 21:54:05,363 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-20 21:54:05,878 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-20 21:54:11,853 INFO [train.py:901] (0/2) Epoch 12, batch 100, loss[loss=0.1891, simple_loss=0.2635, pruned_loss=0.05737, over 7276.00 frames. ], tot_loss[loss=0.1819, simple_loss=0.2526, pruned_loss=0.05558, over 570356.68 frames. ], batch size: 52, lr: 1.38e-02, grad_scale: 8.0 +2023-03-20 21:54:20,960 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1723, 2.2368, 2.1119, 2.9536, 1.3207, 3.5548, 1.1892, 2.8264], + device='cuda:0'), covar=tensor([0.0059, 0.0835, 0.1402, 0.0059, 0.4084, 0.0049, 0.1012, 0.0141], + device='cuda:0'), in_proj_covar=tensor([0.0141, 0.0267, 0.0305, 0.0147, 0.0294, 0.0142, 0.0270, 0.0196], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 21:54:23,442 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9010, 3.4465, 3.6724, 3.6287, 3.4381, 3.5048, 3.7013, 3.5068], + device='cuda:0'), covar=tensor([0.0100, 0.0189, 0.0118, 0.0146, 0.0329, 0.0127, 0.0147, 0.0138], + device='cuda:0'), in_proj_covar=tensor([0.0055, 0.0062, 0.0063, 0.0051, 0.0097, 0.0068, 0.0066, 0.0064], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:54:36,937 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.513e+02 2.411e+02 2.828e+02 3.722e+02 7.099e+02, threshold=5.657e+02, percent-clipped=3.0 +2023-03-20 21:54:37,969 INFO [train.py:901] (0/2) Epoch 12, batch 150, loss[loss=0.1473, simple_loss=0.203, pruned_loss=0.0458, over 5929.00 frames. ], tot_loss[loss=0.1813, simple_loss=0.2523, pruned_loss=0.05508, over 764651.88 frames. ], batch size: 26, lr: 1.38e-02, grad_scale: 8.0 +2023-03-20 21:54:54,715 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0361, 2.4409, 1.9302, 2.4936, 2.3493, 2.0484, 2.4557, 2.2494], + device='cuda:0'), covar=tensor([0.1219, 0.0403, 0.0818, 0.0866, 0.1152, 0.1215, 0.0719, 0.0979], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0037, 0.0044, 0.0039, 0.0040, 0.0038, 0.0038, 0.0035], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 21:54:56,257 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1831, 3.0442, 2.0478, 3.6276, 2.3819, 2.9965, 1.8660, 1.8064], + device='cuda:0'), covar=tensor([0.0181, 0.0385, 0.1491, 0.0312, 0.0232, 0.0223, 0.1849, 0.1373], + device='cuda:0'), in_proj_covar=tensor([0.0194, 0.0217, 0.0300, 0.0209, 0.0229, 0.0214, 0.0267, 0.0287], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004], + device='cuda:0') +2023-03-20 21:55:00,546 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=31259.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:55:03,408 INFO [train.py:901] (0/2) Epoch 12, batch 200, loss[loss=0.1887, simple_loss=0.2601, pruned_loss=0.05868, over 7315.00 frames. ], tot_loss[loss=0.1805, simple_loss=0.2513, pruned_loss=0.05485, over 914678.29 frames. ], batch size: 83, lr: 1.37e-02, grad_scale: 8.0 +2023-03-20 21:55:04,975 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-20 21:55:09,951 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-20 21:55:12,046 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=31281.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:55:16,377 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-20 21:55:16,437 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=31290.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:55:21,629 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5093, 3.1247, 3.2005, 3.2285, 2.5684, 2.3963, 3.4892, 2.4950], + device='cuda:0'), covar=tensor([0.0213, 0.0258, 0.0197, 0.0225, 0.0297, 0.0421, 0.0299, 0.0683], + device='cuda:0'), in_proj_covar=tensor([0.0272, 0.0276, 0.0242, 0.0292, 0.0302, 0.0297, 0.0275, 0.0293], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-20 21:55:24,251 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.98 vs. limit=2.0 +2023-03-20 21:55:26,933 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=31311.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:55:27,236 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.78 vs. limit=2.0 +2023-03-20 21:55:27,809 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.519e+02 2.425e+02 2.946e+02 3.483e+02 6.440e+02, threshold=5.892e+02, percent-clipped=1.0 +2023-03-20 21:55:28,882 INFO [train.py:901] (0/2) Epoch 12, batch 250, loss[loss=0.1819, simple_loss=0.2505, pruned_loss=0.05667, over 7272.00 frames. ], tot_loss[loss=0.1807, simple_loss=0.2509, pruned_loss=0.05523, over 1033182.82 frames. ], batch size: 47, lr: 1.37e-02, grad_scale: 8.0 +2023-03-20 21:55:30,001 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-20 21:55:50,536 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-20 21:55:51,520 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=31359.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:55:54,416 INFO [train.py:901] (0/2) Epoch 12, batch 300, loss[loss=0.1573, simple_loss=0.2289, pruned_loss=0.0428, over 7197.00 frames. ], tot_loss[loss=0.1798, simple_loss=0.2504, pruned_loss=0.05458, over 1125392.79 frames. ], batch size: 39, lr: 1.37e-02, grad_scale: 8.0 +2023-03-20 21:55:55,452 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=31367.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 21:56:00,032 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-20 21:56:02,940 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.88 vs. limit=5.0 +2023-03-20 21:56:18,388 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=31409.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:56:20,276 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.667e+02 2.464e+02 2.963e+02 3.732e+02 1.202e+03, threshold=5.926e+02, percent-clipped=2.0 +2023-03-20 21:56:21,319 INFO [train.py:901] (0/2) Epoch 12, batch 350, loss[loss=0.179, simple_loss=0.2477, pruned_loss=0.05515, over 7303.00 frames. ], tot_loss[loss=0.1791, simple_loss=0.2498, pruned_loss=0.05413, over 1196558.94 frames. ], batch size: 80, lr: 1.37e-02, grad_scale: 8.0 +2023-03-20 21:56:30,648 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3016, 4.8264, 4.8872, 4.7955, 4.6704, 4.3882, 4.9015, 4.7533], + device='cuda:0'), covar=tensor([0.0427, 0.0422, 0.0456, 0.0515, 0.0383, 0.0321, 0.0390, 0.0518], + device='cuda:0'), in_proj_covar=tensor([0.0159, 0.0173, 0.0126, 0.0132, 0.0113, 0.0163, 0.0139, 0.0109], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:56:35,618 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-20 21:56:39,340 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.43 vs. limit=5.0 +2023-03-20 21:56:47,120 INFO [train.py:901] (0/2) Epoch 12, batch 400, loss[loss=0.1739, simple_loss=0.2538, pruned_loss=0.04694, over 7234.00 frames. ], tot_loss[loss=0.1783, simple_loss=0.249, pruned_loss=0.05376, over 1250700.41 frames. ], batch size: 89, lr: 1.37e-02, grad_scale: 8.0 +2023-03-20 21:56:49,770 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=31470.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:57:00,297 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=31491.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 21:57:11,816 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.329e+02 2.122e+02 2.465e+02 3.114e+02 1.072e+03, threshold=4.930e+02, percent-clipped=4.0 +2023-03-20 21:57:12,853 INFO [train.py:901] (0/2) Epoch 12, batch 450, loss[loss=0.1515, simple_loss=0.217, pruned_loss=0.04294, over 6994.00 frames. ], tot_loss[loss=0.1775, simple_loss=0.2487, pruned_loss=0.0532, over 1292223.22 frames. ], batch size: 35, lr: 1.37e-02, grad_scale: 8.0 +2023-03-20 21:57:16,376 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-20 21:57:16,495 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9067, 2.8119, 2.9502, 2.8404, 3.2005, 2.9786, 2.4569, 2.8324], + device='cuda:0'), covar=tensor([0.2013, 0.0875, 0.2197, 0.3466, 0.1211, 0.1371, 0.2493, 0.1946], + device='cuda:0'), in_proj_covar=tensor([0.0036, 0.0038, 0.0035, 0.0037, 0.0033, 0.0032, 0.0047, 0.0034], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:57:16,896 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-20 21:57:32,048 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=31552.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 21:57:35,477 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=31559.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:57:38,347 INFO [train.py:901] (0/2) Epoch 12, batch 500, loss[loss=0.1906, simple_loss=0.2647, pruned_loss=0.05822, over 7326.00 frames. ], tot_loss[loss=0.1777, simple_loss=0.2488, pruned_loss=0.05329, over 1326128.57 frames. ], batch size: 75, lr: 1.37e-02, grad_scale: 16.0 +2023-03-20 21:57:47,064 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=31581.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:57:49,999 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-20 21:57:51,092 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5085, 5.0138, 5.0351, 4.9868, 4.8736, 4.6256, 5.0579, 4.8732], + device='cuda:0'), covar=tensor([0.0395, 0.0390, 0.0459, 0.0436, 0.0305, 0.0275, 0.0345, 0.0471], + device='cuda:0'), in_proj_covar=tensor([0.0158, 0.0173, 0.0127, 0.0132, 0.0111, 0.0163, 0.0138, 0.0110], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:57:51,519 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-20 21:57:51,599 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=31590.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:57:52,535 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-20 21:57:54,523 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-20 21:57:58,839 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-20 21:58:00,362 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=31607.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:58:03,281 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.442e+02 2.477e+02 2.904e+02 3.674e+02 7.030e+02, threshold=5.807e+02, percent-clipped=4.0 +2023-03-20 21:58:04,294 INFO [train.py:901] (0/2) Epoch 12, batch 550, loss[loss=0.1753, simple_loss=0.2442, pruned_loss=0.05321, over 7341.00 frames. ], tot_loss[loss=0.1773, simple_loss=0.2481, pruned_loss=0.05326, over 1352243.23 frames. ], batch size: 61, lr: 1.37e-02, grad_scale: 16.0 +2023-03-20 21:58:10,815 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-20 21:58:11,863 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=31629.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:58:16,373 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=31638.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:58:18,818 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-20 21:58:22,343 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-20 21:58:23,698 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.81 vs. limit=2.0 +2023-03-20 21:58:29,237 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-20 21:58:29,726 INFO [train.py:901] (0/2) Epoch 12, batch 600, loss[loss=0.1941, simple_loss=0.2633, pruned_loss=0.06248, over 7293.00 frames. ], tot_loss[loss=0.1781, simple_loss=0.2484, pruned_loss=0.05389, over 1370845.01 frames. ], batch size: 77, lr: 1.37e-02, grad_scale: 16.0 +2023-03-20 21:58:31,427 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=31667.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 21:58:45,237 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.92 vs. limit=2.0 +2023-03-20 21:58:45,362 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-20 21:58:49,968 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=31704.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:58:54,861 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.764e+02 2.406e+02 2.959e+02 3.734e+02 7.911e+02, threshold=5.917e+02, percent-clipped=4.0 +2023-03-20 21:58:55,410 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-20 21:58:55,890 INFO [train.py:901] (0/2) Epoch 12, batch 650, loss[loss=0.1796, simple_loss=0.2527, pruned_loss=0.05324, over 7265.00 frames. ], tot_loss[loss=0.1775, simple_loss=0.2481, pruned_loss=0.05343, over 1386748.29 frames. ], batch size: 89, lr: 1.37e-02, grad_scale: 16.0 +2023-03-20 21:58:55,943 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=31715.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 21:58:59,571 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8563, 3.8367, 3.1562, 3.1062, 3.1839, 2.1680, 1.4511, 3.7801], + device='cuda:0'), covar=tensor([0.0037, 0.0061, 0.0105, 0.0090, 0.0099, 0.0515, 0.0624, 0.0066], + device='cuda:0'), in_proj_covar=tensor([0.0063, 0.0060, 0.0078, 0.0070, 0.0084, 0.0104, 0.0107, 0.0073], + device='cuda:0'), out_proj_covar=tensor([8.2505e-05, 8.8932e-05, 1.0465e-04, 1.0106e-04, 1.1379e-04, 1.4358e-04, + 1.4759e-04, 1.0015e-04], device='cuda:0') +2023-03-20 21:59:12,169 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-20 21:59:22,008 INFO [train.py:901] (0/2) Epoch 12, batch 700, loss[loss=0.1501, simple_loss=0.2299, pruned_loss=0.0351, over 7321.00 frames. ], tot_loss[loss=0.1773, simple_loss=0.2482, pruned_loss=0.05322, over 1399382.81 frames. ], batch size: 75, lr: 1.36e-02, grad_scale: 16.0 +2023-03-20 21:59:22,072 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=31765.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:59:22,136 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=31765.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:59:22,501 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-20 21:59:27,235 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2391, 3.9348, 4.0372, 4.1804, 4.1433, 4.1576, 4.1104, 3.8127], + device='cuda:0'), covar=tensor([0.0024, 0.0057, 0.0029, 0.0029, 0.0030, 0.0029, 0.0032, 0.0052], + device='cuda:0'), in_proj_covar=tensor([0.0032, 0.0041, 0.0038, 0.0036, 0.0036, 0.0037, 0.0044, 0.0045], + device='cuda:0'), out_proj_covar=tensor([8.1091e-05, 1.1989e-04, 1.0501e-04, 9.2904e-05, 9.1946e-05, 9.8820e-05, + 1.2530e-04, 1.1850e-04], device='cuda:0') +2023-03-20 21:59:33,743 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=31788.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 21:59:42,658 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8784, 2.6455, 3.0886, 2.8780, 2.9362, 2.4713, 2.2233, 2.7155], + device='cuda:0'), covar=tensor([0.1554, 0.0500, 0.1006, 0.2055, 0.0991, 0.1523, 0.2510, 0.1809], + device='cuda:0'), in_proj_covar=tensor([0.0034, 0.0037, 0.0033, 0.0034, 0.0032, 0.0031, 0.0044, 0.0032], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 21:59:46,995 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.502e+02 2.229e+02 2.839e+02 3.592e+02 6.099e+02, threshold=5.677e+02, percent-clipped=2.0 +2023-03-20 21:59:47,030 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-20 21:59:47,537 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-20 21:59:48,025 INFO [train.py:901] (0/2) Epoch 12, batch 750, loss[loss=0.1504, simple_loss=0.2201, pruned_loss=0.0404, over 6276.00 frames. ], tot_loss[loss=0.1766, simple_loss=0.2473, pruned_loss=0.05299, over 1406565.23 frames. ], batch size: 27, lr: 1.36e-02, grad_scale: 16.0 +2023-03-20 21:59:52,162 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=31823.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:00:02,260 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-20 22:00:04,928 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=31847.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 22:00:05,996 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=31849.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:00:06,861 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-20 22:00:12,464 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0832, 4.0921, 3.7298, 3.3717, 3.5485, 2.5067, 1.7372, 4.1080], + device='cuda:0'), covar=tensor([0.0019, 0.0026, 0.0044, 0.0056, 0.0051, 0.0324, 0.0505, 0.0033], + device='cuda:0'), in_proj_covar=tensor([0.0063, 0.0059, 0.0078, 0.0070, 0.0083, 0.0103, 0.0107, 0.0072], + device='cuda:0'), out_proj_covar=tensor([8.1849e-05, 8.8076e-05, 1.0516e-04, 1.0052e-04, 1.1267e-04, 1.4226e-04, + 1.4744e-04, 9.8715e-05], device='cuda:0') +2023-03-20 22:00:12,872 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-20 22:00:13,839 INFO [train.py:901] (0/2) Epoch 12, batch 800, loss[loss=0.1716, simple_loss=0.2387, pruned_loss=0.05227, over 7336.00 frames. ], tot_loss[loss=0.1762, simple_loss=0.2469, pruned_loss=0.0528, over 1414304.02 frames. ], batch size: 49, lr: 1.36e-02, grad_scale: 16.0 +2023-03-20 22:00:14,363 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-20 22:00:20,505 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=31878.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:00:23,597 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=31884.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:00:25,500 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-20 22:00:38,421 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.738e+02 2.334e+02 2.815e+02 3.523e+02 9.191e+02, threshold=5.630e+02, percent-clipped=3.0 +2023-03-20 22:00:39,488 INFO [train.py:901] (0/2) Epoch 12, batch 850, loss[loss=0.1377, simple_loss=0.2106, pruned_loss=0.03238, over 7331.00 frames. ], tot_loss[loss=0.1765, simple_loss=0.2473, pruned_loss=0.05281, over 1421285.64 frames. ], batch size: 44, lr: 1.36e-02, grad_scale: 16.0 +2023-03-20 22:00:39,640 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=31915.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:00:43,514 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-20 22:00:43,525 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-20 22:00:48,945 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-20 22:00:52,281 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=31939.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:00:53,222 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-20 22:00:57,764 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=31950.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:01:05,069 INFO [train.py:901] (0/2) Epoch 12, batch 900, loss[loss=0.1742, simple_loss=0.2489, pruned_loss=0.04971, over 7280.00 frames. ], tot_loss[loss=0.1773, simple_loss=0.2481, pruned_loss=0.05326, over 1427464.08 frames. ], batch size: 52, lr: 1.36e-02, grad_scale: 16.0 +2023-03-20 22:01:05,396 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-03-20 22:01:11,333 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=31976.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:01:23,464 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-32000.pt +2023-03-20 22:01:28,435 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7093, 2.9415, 1.8637, 3.3610, 2.2686, 2.8107, 1.5202, 1.7908], + device='cuda:0'), covar=tensor([0.0149, 0.0542, 0.1564, 0.0307, 0.0251, 0.0199, 0.2025, 0.1247], + device='cuda:0'), in_proj_covar=tensor([0.0196, 0.0214, 0.0300, 0.0207, 0.0229, 0.0214, 0.0266, 0.0287], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 22:01:32,931 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=32011.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:01:33,763 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.557e+02 2.399e+02 2.836e+02 3.405e+02 6.362e+02, threshold=5.672e+02, percent-clipped=3.0 +2023-03-20 22:01:33,941 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9668, 3.4611, 2.8700, 3.0922, 3.4167, 2.4094, 3.0672, 2.7680], + device='cuda:0'), covar=tensor([0.0691, 0.0578, 0.0773, 0.1282, 0.0576, 0.0957, 0.0722, 0.1092], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0037, 0.0042, 0.0038, 0.0039, 0.0038, 0.0040, 0.0036], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 22:01:34,368 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-20 22:01:34,803 INFO [train.py:901] (0/2) Epoch 12, batch 950, loss[loss=0.1808, simple_loss=0.2568, pruned_loss=0.05237, over 7235.00 frames. ], tot_loss[loss=0.1766, simple_loss=0.2475, pruned_loss=0.05288, over 1429802.65 frames. ], batch size: 89, lr: 1.36e-02, grad_scale: 16.0 +2023-03-20 22:01:58,841 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=32060.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:01:59,273 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-20 22:02:01,271 INFO [train.py:901] (0/2) Epoch 12, batch 1000, loss[loss=0.1553, simple_loss=0.2292, pruned_loss=0.04068, over 7131.00 frames. ], tot_loss[loss=0.1765, simple_loss=0.2474, pruned_loss=0.05281, over 1432188.87 frames. ], batch size: 41, lr: 1.36e-02, grad_scale: 16.0 +2023-03-20 22:02:01,367 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=32065.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:02:19,919 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-20 22:02:25,236 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.696e+02 2.567e+02 3.123e+02 3.891e+02 1.111e+03, threshold=6.246e+02, percent-clipped=6.0 +2023-03-20 22:02:25,323 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=32113.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:02:26,944 INFO [train.py:901] (0/2) Epoch 12, batch 1050, loss[loss=0.1869, simple_loss=0.2578, pruned_loss=0.05799, over 7332.00 frames. ], tot_loss[loss=0.1764, simple_loss=0.2475, pruned_loss=0.05271, over 1435984.39 frames. ], batch size: 61, lr: 1.36e-02, grad_scale: 16.0 +2023-03-20 22:02:41,148 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-20 22:02:41,725 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=32144.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:02:43,823 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=32147.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 22:02:46,709 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-20 22:02:52,643 INFO [train.py:901] (0/2) Epoch 12, batch 1100, loss[loss=0.1473, simple_loss=0.2248, pruned_loss=0.0349, over 7149.00 frames. ], tot_loss[loss=0.1765, simple_loss=0.2475, pruned_loss=0.05273, over 1435925.09 frames. ], batch size: 41, lr: 1.36e-02, grad_scale: 16.0 +2023-03-20 22:02:59,735 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=32179.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:03:07,684 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=32195.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 22:03:08,404 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-03-20 22:03:14,770 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-20 22:03:15,258 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 22:03:17,755 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.242e+02 2.391e+02 2.835e+02 3.446e+02 6.409e+02, threshold=5.669e+02, percent-clipped=2.0 +2023-03-20 22:03:18,791 INFO [train.py:901] (0/2) Epoch 12, batch 1150, loss[loss=0.1759, simple_loss=0.2423, pruned_loss=0.05476, over 7208.00 frames. ], tot_loss[loss=0.1765, simple_loss=0.2475, pruned_loss=0.05272, over 1436669.94 frames. ], batch size: 50, lr: 1.36e-02, grad_scale: 16.0 +2023-03-20 22:03:21,019 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1692, 3.7585, 3.7090, 3.7300, 3.1139, 3.1038, 4.3206, 3.3294], + device='cuda:0'), covar=tensor([0.0207, 0.0259, 0.0163, 0.0164, 0.0363, 0.0398, 0.0313, 0.0487], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0278, 0.0240, 0.0288, 0.0299, 0.0297, 0.0275, 0.0288], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-20 22:03:28,498 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-20 22:03:28,989 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-20 22:03:29,049 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=32234.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:03:44,689 INFO [train.py:901] (0/2) Epoch 12, batch 1200, loss[loss=0.1885, simple_loss=0.2577, pruned_loss=0.05959, over 7228.00 frames. ], tot_loss[loss=0.1761, simple_loss=0.247, pruned_loss=0.05262, over 1435363.96 frames. ], batch size: 93, lr: 1.35e-02, grad_scale: 16.0 +2023-03-20 22:03:47,774 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=32271.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:04:00,856 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-20 22:04:04,919 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0073, 4.0215, 3.4044, 3.2698, 3.4042, 2.0979, 1.6167, 3.9615], + device='cuda:0'), covar=tensor([0.0030, 0.0031, 0.0098, 0.0095, 0.0107, 0.0511, 0.0646, 0.0067], + device='cuda:0'), in_proj_covar=tensor([0.0066, 0.0061, 0.0082, 0.0072, 0.0087, 0.0108, 0.0112, 0.0076], + device='cuda:0'), out_proj_covar=tensor([8.5782e-05, 9.0320e-05, 1.1056e-04, 1.0414e-04, 1.1784e-04, 1.4829e-04, + 1.5448e-04, 1.0428e-04], device='cuda:0') +2023-03-20 22:04:05,842 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=32306.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:04:09,312 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.537e+02 2.225e+02 2.639e+02 3.097e+02 5.486e+02, threshold=5.278e+02, percent-clipped=0.0 +2023-03-20 22:04:10,303 INFO [train.py:901] (0/2) Epoch 12, batch 1250, loss[loss=0.1992, simple_loss=0.2577, pruned_loss=0.07035, over 7296.00 frames. ], tot_loss[loss=0.1767, simple_loss=0.2477, pruned_loss=0.0529, over 1438411.81 frames. ], batch size: 86, lr: 1.35e-02, grad_scale: 16.0 +2023-03-20 22:04:14,447 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2899, 3.5103, 3.3772, 3.5066, 3.1862, 3.4639, 3.7663, 3.7513], + device='cuda:0'), covar=tensor([0.0307, 0.0210, 0.0246, 0.0216, 0.0461, 0.0328, 0.0270, 0.0239], + device='cuda:0'), in_proj_covar=tensor([0.0098, 0.0099, 0.0093, 0.0106, 0.0098, 0.0085, 0.0076, 0.0082], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:04:25,359 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-20 22:04:29,310 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-20 22:04:30,791 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-20 22:04:33,436 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=32360.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:04:35,922 INFO [train.py:901] (0/2) Epoch 12, batch 1300, loss[loss=0.1318, simple_loss=0.2009, pruned_loss=0.03135, over 7012.00 frames. ], tot_loss[loss=0.1772, simple_loss=0.2479, pruned_loss=0.05328, over 1440450.99 frames. ], batch size: 35, lr: 1.35e-02, grad_scale: 16.0 +2023-03-20 22:04:53,794 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-20 22:04:56,155 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-20 22:04:59,339 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=32408.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:04:59,835 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-20 22:05:01,800 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.535e+02 2.570e+02 3.128e+02 3.930e+02 7.869e+02, threshold=6.255e+02, percent-clipped=7.0 +2023-03-20 22:05:02,825 INFO [train.py:901] (0/2) Epoch 12, batch 1350, loss[loss=0.2135, simple_loss=0.279, pruned_loss=0.07397, over 6738.00 frames. ], tot_loss[loss=0.1784, simple_loss=0.2493, pruned_loss=0.05374, over 1440579.06 frames. ], batch size: 106, lr: 1.35e-02, grad_scale: 16.0 +2023-03-20 22:05:10,299 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-20 22:05:17,361 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=32444.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:05:25,272 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.03 vs. limit=2.0 +2023-03-20 22:05:27,174 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4689, 1.3985, 1.4828, 1.3406, 1.8848, 1.4128, 1.4001, 1.4481], + device='cuda:0'), covar=tensor([0.0224, 0.0218, 0.0060, 0.0057, 0.0189, 0.0350, 0.0135, 0.0186], + device='cuda:0'), in_proj_covar=tensor([0.0021, 0.0019, 0.0020, 0.0018, 0.0019, 0.0019, 0.0020, 0.0020], + device='cuda:0'), out_proj_covar=tensor([5.1560e-05, 4.7678e-05, 4.6464e-05, 4.0651e-05, 4.7935e-05, 4.5374e-05, + 4.7551e-05, 5.1525e-05], device='cuda:0') +2023-03-20 22:05:28,596 INFO [train.py:901] (0/2) Epoch 12, batch 1400, loss[loss=0.1834, simple_loss=0.2548, pruned_loss=0.05602, over 7280.00 frames. ], tot_loss[loss=0.1784, simple_loss=0.2493, pruned_loss=0.05378, over 1439766.11 frames. ], batch size: 70, lr: 1.35e-02, grad_scale: 16.0 +2023-03-20 22:05:35,670 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=32479.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:05:42,688 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-20 22:05:42,733 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=32492.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:05:52,745 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4542, 4.9273, 4.9290, 4.8158, 4.6329, 4.4986, 5.0314, 4.5939], + device='cuda:0'), covar=tensor([0.0849, 0.0814, 0.0599, 0.1024, 0.0759, 0.0686, 0.0621, 0.1145], + device='cuda:0'), in_proj_covar=tensor([0.0162, 0.0177, 0.0130, 0.0135, 0.0113, 0.0169, 0.0139, 0.0113], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:05:53,143 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.809e+02 2.399e+02 2.899e+02 3.458e+02 5.987e+02, threshold=5.799e+02, percent-clipped=0.0 +2023-03-20 22:05:54,228 INFO [train.py:901] (0/2) Epoch 12, batch 1450, loss[loss=0.1956, simple_loss=0.263, pruned_loss=0.06411, over 7310.00 frames. ], tot_loss[loss=0.1783, simple_loss=0.2489, pruned_loss=0.05378, over 1439902.04 frames. ], batch size: 80, lr: 1.35e-02, grad_scale: 16.0 +2023-03-20 22:06:00,333 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=32527.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:06:00,405 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7538, 4.6793, 4.3675, 3.9395, 4.4591, 3.1498, 2.2646, 4.5840], + device='cuda:0'), covar=tensor([0.0011, 0.0024, 0.0045, 0.0045, 0.0025, 0.0293, 0.0394, 0.0030], + device='cuda:0'), in_proj_covar=tensor([0.0067, 0.0061, 0.0081, 0.0071, 0.0086, 0.0108, 0.0110, 0.0075], + device='cuda:0'), out_proj_covar=tensor([8.6513e-05, 9.0528e-05, 1.0956e-04, 1.0253e-04, 1.1741e-04, 1.4785e-04, + 1.5182e-04, 1.0278e-04], device='cuda:0') +2023-03-20 22:06:03,938 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=32534.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:06:05,662 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5878, 3.2611, 3.1354, 3.2064, 2.6953, 2.5111, 3.6211, 2.4803], + device='cuda:0'), covar=tensor([0.0196, 0.0151, 0.0217, 0.0212, 0.0311, 0.0420, 0.0201, 0.0725], + device='cuda:0'), in_proj_covar=tensor([0.0272, 0.0275, 0.0237, 0.0289, 0.0297, 0.0294, 0.0271, 0.0287], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-20 22:06:06,497 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-20 22:06:20,135 INFO [train.py:901] (0/2) Epoch 12, batch 1500, loss[loss=0.1643, simple_loss=0.2363, pruned_loss=0.04621, over 7365.00 frames. ], tot_loss[loss=0.1773, simple_loss=0.248, pruned_loss=0.05334, over 1438542.18 frames. ], batch size: 51, lr: 1.35e-02, grad_scale: 16.0 +2023-03-20 22:06:22,719 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-20 22:06:23,760 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=32571.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:06:29,201 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=32582.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:06:41,439 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=32606.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:06:44,759 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.637e+02 2.462e+02 2.853e+02 3.546e+02 6.447e+02, threshold=5.706e+02, percent-clipped=3.0 +2023-03-20 22:06:45,774 INFO [train.py:901] (0/2) Epoch 12, batch 1550, loss[loss=0.199, simple_loss=0.2743, pruned_loss=0.06182, over 7258.00 frames. ], tot_loss[loss=0.1778, simple_loss=0.2485, pruned_loss=0.05357, over 1441309.36 frames. ], batch size: 89, lr: 1.35e-02, grad_scale: 16.0 +2023-03-20 22:06:46,794 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-20 22:06:47,781 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=32619.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:06:48,067 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.99 vs. limit=2.0 +2023-03-20 22:07:06,067 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=32654.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:07:12,099 INFO [train.py:901] (0/2) Epoch 12, batch 1600, loss[loss=0.1846, simple_loss=0.2549, pruned_loss=0.05715, over 7288.00 frames. ], tot_loss[loss=0.1777, simple_loss=0.2482, pruned_loss=0.05363, over 1439375.52 frames. ], batch size: 66, lr: 1.35e-02, grad_scale: 16.0 +2023-03-20 22:07:13,787 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1134, 2.3748, 2.1649, 3.2086, 1.3370, 2.8943, 1.1873, 3.0180], + device='cuda:0'), covar=tensor([0.0072, 0.0797, 0.1634, 0.0057, 0.4660, 0.0066, 0.1085, 0.0281], + device='cuda:0'), in_proj_covar=tensor([0.0146, 0.0261, 0.0310, 0.0146, 0.0295, 0.0149, 0.0267, 0.0205], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 22:07:18,147 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-20 22:07:19,151 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-20 22:07:22,238 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-20 22:07:24,867 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7702, 3.3255, 3.3987, 3.5085, 2.5593, 2.5556, 3.7576, 2.6476], + device='cuda:0'), covar=tensor([0.0197, 0.0200, 0.0244, 0.0213, 0.0426, 0.0527, 0.0264, 0.0725], + device='cuda:0'), in_proj_covar=tensor([0.0277, 0.0280, 0.0239, 0.0293, 0.0299, 0.0298, 0.0275, 0.0290], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-20 22:07:25,884 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1964, 1.6220, 1.7117, 1.6794, 1.8486, 1.6858, 1.5327, 1.3082], + device='cuda:0'), covar=tensor([0.0870, 0.0197, 0.0085, 0.0119, 0.0323, 0.0473, 0.0209, 0.0296], + device='cuda:0'), in_proj_covar=tensor([0.0021, 0.0019, 0.0020, 0.0018, 0.0019, 0.0019, 0.0020, 0.0020], + device='cuda:0'), out_proj_covar=tensor([5.1564e-05, 4.7127e-05, 4.6825e-05, 4.0175e-05, 4.7614e-05, 4.5023e-05, + 4.7783e-05, 5.1671e-05], device='cuda:0') +2023-03-20 22:07:31,720 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-20 22:07:36,879 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.457e+02 2.362e+02 2.862e+02 3.457e+02 8.164e+02, threshold=5.723e+02, percent-clipped=2.0 +2023-03-20 22:07:36,923 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-20 22:07:37,915 INFO [train.py:901] (0/2) Epoch 12, batch 1650, loss[loss=0.1774, simple_loss=0.2513, pruned_loss=0.05171, over 7255.00 frames. ], tot_loss[loss=0.1774, simple_loss=0.2481, pruned_loss=0.05339, over 1438296.05 frames. ], batch size: 55, lr: 1.34e-02, grad_scale: 16.0 +2023-03-20 22:07:44,919 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-20 22:07:59,918 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.90 vs. limit=2.0 +2023-03-20 22:08:02,080 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 22:08:03,200 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0415, 3.9805, 3.6080, 3.3452, 3.3557, 2.2983, 1.7083, 3.9653], + device='cuda:0'), covar=tensor([0.0019, 0.0022, 0.0065, 0.0055, 0.0083, 0.0380, 0.0477, 0.0041], + device='cuda:0'), in_proj_covar=tensor([0.0065, 0.0061, 0.0081, 0.0070, 0.0088, 0.0107, 0.0109, 0.0075], + device='cuda:0'), out_proj_covar=tensor([8.4359e-05, 9.0606e-05, 1.0875e-04, 1.0137e-04, 1.1982e-04, 1.4694e-04, + 1.4970e-04, 1.0274e-04], device='cuda:0') +2023-03-20 22:08:03,574 INFO [train.py:901] (0/2) Epoch 12, batch 1700, loss[loss=0.1723, simple_loss=0.2453, pruned_loss=0.04965, over 7350.00 frames. ], tot_loss[loss=0.1777, simple_loss=0.2486, pruned_loss=0.05342, over 1441061.65 frames. ], batch size: 73, lr: 1.34e-02, grad_scale: 16.0 +2023-03-20 22:08:05,596 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-20 22:08:17,089 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-20 22:08:28,546 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.594e+02 2.417e+02 2.972e+02 3.822e+02 7.800e+02, threshold=5.944e+02, percent-clipped=3.0 +2023-03-20 22:08:29,531 INFO [train.py:901] (0/2) Epoch 12, batch 1750, loss[loss=0.1614, simple_loss=0.2381, pruned_loss=0.04237, over 7144.00 frames. ], tot_loss[loss=0.1772, simple_loss=0.2484, pruned_loss=0.05307, over 1441256.96 frames. ], batch size: 41, lr: 1.34e-02, grad_scale: 16.0 +2023-03-20 22:08:33,873 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-03-20 22:08:43,147 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-20 22:08:44,157 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-20 22:08:55,112 INFO [train.py:901] (0/2) Epoch 12, batch 1800, loss[loss=0.1618, simple_loss=0.2371, pruned_loss=0.04328, over 7333.00 frames. ], tot_loss[loss=0.1764, simple_loss=0.2475, pruned_loss=0.05259, over 1443037.17 frames. ], batch size: 54, lr: 1.34e-02, grad_scale: 16.0 +2023-03-20 22:09:05,600 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-20 22:09:13,876 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3799, 2.5022, 2.1182, 2.3987, 2.3576, 2.2190, 2.5367, 2.3302], + device='cuda:0'), covar=tensor([0.0385, 0.0536, 0.1507, 0.0761, 0.0928, 0.0549, 0.1465, 0.1192], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0040, 0.0046, 0.0040, 0.0042, 0.0040, 0.0043, 0.0038], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0001, 0.0002, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 22:09:17,427 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3399, 1.7107, 1.6549, 1.7037, 1.8417, 1.7713, 1.6125, 1.3582], + device='cuda:0'), covar=tensor([0.0347, 0.0208, 0.0179, 0.0057, 0.0323, 0.0238, 0.0140, 0.0194], + device='cuda:0'), in_proj_covar=tensor([0.0021, 0.0020, 0.0021, 0.0019, 0.0020, 0.0018, 0.0020, 0.0021], + device='cuda:0'), out_proj_covar=tensor([5.2965e-05, 4.9513e-05, 4.8262e-05, 4.1590e-05, 4.9420e-05, 4.4489e-05, + 4.8019e-05, 5.2920e-05], device='cuda:0') +2023-03-20 22:09:19,265 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-20 22:09:19,726 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.412e+02 2.518e+02 3.025e+02 3.704e+02 1.165e+03, threshold=6.050e+02, percent-clipped=2.0 +2023-03-20 22:09:20,761 INFO [train.py:901] (0/2) Epoch 12, batch 1850, loss[loss=0.1334, simple_loss=0.208, pruned_loss=0.02944, over 7338.00 frames. ], tot_loss[loss=0.1758, simple_loss=0.2472, pruned_loss=0.0522, over 1440221.47 frames. ], batch size: 44, lr: 1.34e-02, grad_scale: 16.0 +2023-03-20 22:09:29,548 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2896, 2.5557, 2.0279, 3.3910, 1.4412, 3.0309, 1.3492, 2.7602], + device='cuda:0'), covar=tensor([0.0063, 0.0877, 0.1843, 0.0063, 0.4153, 0.0065, 0.1088, 0.0134], + device='cuda:0'), in_proj_covar=tensor([0.0145, 0.0263, 0.0303, 0.0148, 0.0293, 0.0148, 0.0269, 0.0204], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 22:09:29,952 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-20 22:09:46,115 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-20 22:09:46,605 INFO [train.py:901] (0/2) Epoch 12, batch 1900, loss[loss=0.1855, simple_loss=0.2579, pruned_loss=0.05655, over 7322.00 frames. ], tot_loss[loss=0.1757, simple_loss=0.247, pruned_loss=0.05219, over 1438997.91 frames. ], batch size: 59, lr: 1.34e-02, grad_scale: 16.0 +2023-03-20 22:09:47,231 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=32966.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:10:10,655 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-20 22:10:12,139 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.404e+02 2.470e+02 2.955e+02 3.740e+02 5.660e+02, threshold=5.911e+02, percent-clipped=0.0 +2023-03-20 22:10:13,160 INFO [train.py:901] (0/2) Epoch 12, batch 1950, loss[loss=0.169, simple_loss=0.2426, pruned_loss=0.04772, over 7281.00 frames. ], tot_loss[loss=0.1766, simple_loss=0.248, pruned_loss=0.05264, over 1441177.75 frames. ], batch size: 68, lr: 1.34e-02, grad_scale: 16.0 +2023-03-20 22:10:19,218 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=33027.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:10:21,569 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-20 22:10:26,048 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-20 22:10:26,554 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-20 22:10:30,134 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6597, 5.1127, 5.1574, 5.1388, 4.8592, 4.6636, 5.1927, 4.9993], + device='cuda:0'), covar=tensor([0.0306, 0.0319, 0.0344, 0.0349, 0.0296, 0.0291, 0.0304, 0.0428], + device='cuda:0'), in_proj_covar=tensor([0.0160, 0.0178, 0.0130, 0.0133, 0.0111, 0.0168, 0.0144, 0.0114], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:10:38,030 INFO [train.py:901] (0/2) Epoch 12, batch 2000, loss[loss=0.1929, simple_loss=0.2539, pruned_loss=0.06599, over 7291.00 frames. ], tot_loss[loss=0.1771, simple_loss=0.2485, pruned_loss=0.05279, over 1440943.50 frames. ], batch size: 49, lr: 1.34e-02, grad_scale: 16.0 +2023-03-20 22:10:40,721 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=33069.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:10:43,053 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-20 22:10:46,257 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4436, 2.9729, 2.3103, 3.6881, 1.4703, 3.6019, 1.4391, 3.1825], + device='cuda:0'), covar=tensor([0.0057, 0.0753, 0.1616, 0.0054, 0.4362, 0.0069, 0.1250, 0.0149], + device='cuda:0'), in_proj_covar=tensor([0.0143, 0.0257, 0.0297, 0.0146, 0.0287, 0.0147, 0.0264, 0.0201], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 22:10:54,588 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-20 22:11:02,271 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-20 22:11:03,252 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.631e+02 2.429e+02 3.002e+02 3.468e+02 9.132e+02, threshold=6.004e+02, percent-clipped=3.0 +2023-03-20 22:11:04,288 INFO [train.py:901] (0/2) Epoch 12, batch 2050, loss[loss=0.1704, simple_loss=0.2478, pruned_loss=0.04648, over 7288.00 frames. ], tot_loss[loss=0.1769, simple_loss=0.2481, pruned_loss=0.05286, over 1439942.82 frames. ], batch size: 68, lr: 1.34e-02, grad_scale: 16.0 +2023-03-20 22:11:11,891 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=33130.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:11:17,645 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.73 vs. limit=2.0 +2023-03-20 22:11:24,385 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=33155.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:11:29,952 INFO [train.py:901] (0/2) Epoch 12, batch 2100, loss[loss=0.1485, simple_loss=0.2276, pruned_loss=0.03466, over 7322.00 frames. ], tot_loss[loss=0.1762, simple_loss=0.2474, pruned_loss=0.05253, over 1439821.97 frames. ], batch size: 44, lr: 1.34e-02, grad_scale: 16.0 +2023-03-20 22:11:33,334 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.86 vs. limit=2.0 +2023-03-20 22:11:35,998 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-20 22:11:39,119 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-20 22:11:42,233 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7726, 3.4935, 3.3946, 3.4706, 2.9555, 3.2741, 3.4900, 3.1740], + device='cuda:0'), covar=tensor([0.0186, 0.0196, 0.0246, 0.0200, 0.0543, 0.0197, 0.0286, 0.0227], + device='cuda:0'), in_proj_covar=tensor([0.0055, 0.0063, 0.0064, 0.0052, 0.0102, 0.0068, 0.0067, 0.0064], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0001, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:11:50,684 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.86 vs. limit=2.0 +2023-03-20 22:11:54,914 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.733e+02 2.641e+02 3.038e+02 3.805e+02 9.449e+02, threshold=6.075e+02, percent-clipped=6.0 +2023-03-20 22:11:55,930 INFO [train.py:901] (0/2) Epoch 12, batch 2150, loss[loss=0.195, simple_loss=0.2642, pruned_loss=0.06291, over 7241.00 frames. ], tot_loss[loss=0.1768, simple_loss=0.2479, pruned_loss=0.05285, over 1441673.49 frames. ], batch size: 55, lr: 1.33e-02, grad_scale: 16.0 +2023-03-20 22:11:56,581 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=33216.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:12:13,114 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7024, 4.3795, 4.3234, 4.8065, 4.8137, 4.8336, 4.0894, 4.3343], + device='cuda:0'), covar=tensor([0.0704, 0.2292, 0.2554, 0.1031, 0.0756, 0.1131, 0.0731, 0.0922], + device='cuda:0'), in_proj_covar=tensor([0.0173, 0.0281, 0.0233, 0.0222, 0.0165, 0.0291, 0.0157, 0.0198], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0001, 0.0002], + device='cuda:0') +2023-03-20 22:12:21,550 INFO [train.py:901] (0/2) Epoch 12, batch 2200, loss[loss=0.1674, simple_loss=0.2304, pruned_loss=0.05219, over 7125.00 frames. ], tot_loss[loss=0.1763, simple_loss=0.2474, pruned_loss=0.05262, over 1441266.05 frames. ], batch size: 41, lr: 1.33e-02, grad_scale: 16.0 +2023-03-20 22:12:26,184 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-20 22:12:46,183 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.353e+02 2.489e+02 2.908e+02 3.489e+02 9.898e+02, threshold=5.816e+02, percent-clipped=3.0 +2023-03-20 22:12:47,225 INFO [train.py:901] (0/2) Epoch 12, batch 2250, loss[loss=0.1581, simple_loss=0.229, pruned_loss=0.04362, over 7203.00 frames. ], tot_loss[loss=0.1764, simple_loss=0.2474, pruned_loss=0.05267, over 1439817.34 frames. ], batch size: 45, lr: 1.33e-02, grad_scale: 16.0 +2023-03-20 22:12:50,790 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=33322.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:12:59,941 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-20 22:13:00,427 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-20 22:13:13,455 INFO [train.py:901] (0/2) Epoch 12, batch 2300, loss[loss=0.1831, simple_loss=0.2539, pruned_loss=0.05612, over 7327.00 frames. ], tot_loss[loss=0.1776, simple_loss=0.2487, pruned_loss=0.05322, over 1439398.23 frames. ], batch size: 61, lr: 1.33e-02, grad_scale: 16.0 +2023-03-20 22:13:13,475 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-20 22:13:37,839 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.704e+02 2.453e+02 2.925e+02 3.664e+02 6.582e+02, threshold=5.850e+02, percent-clipped=1.0 +2023-03-20 22:13:38,888 INFO [train.py:901] (0/2) Epoch 12, batch 2350, loss[loss=0.1889, simple_loss=0.2661, pruned_loss=0.05586, over 7267.00 frames. ], tot_loss[loss=0.177, simple_loss=0.2483, pruned_loss=0.0529, over 1439225.29 frames. ], batch size: 64, lr: 1.33e-02, grad_scale: 16.0 +2023-03-20 22:13:44,677 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=33425.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:13:59,852 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-20 22:14:05,404 INFO [train.py:901] (0/2) Epoch 12, batch 2400, loss[loss=0.1624, simple_loss=0.2452, pruned_loss=0.03979, over 7260.00 frames. ], tot_loss[loss=0.1767, simple_loss=0.2483, pruned_loss=0.0526, over 1441523.39 frames. ], batch size: 52, lr: 1.33e-02, grad_scale: 16.0 +2023-03-20 22:14:05,457 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-20 22:14:09,034 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=33472.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:14:10,454 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4092, 4.9762, 5.0681, 4.9510, 4.8607, 4.4737, 5.0679, 4.8626], + device='cuda:0'), covar=tensor([0.0455, 0.0357, 0.0278, 0.0403, 0.0277, 0.0357, 0.0297, 0.0465], + device='cuda:0'), in_proj_covar=tensor([0.0161, 0.0181, 0.0131, 0.0133, 0.0112, 0.0167, 0.0141, 0.0114], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:14:15,926 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-20 22:14:18,481 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-20 22:14:24,031 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9493, 4.3618, 4.4035, 4.2998, 4.3426, 4.0321, 4.4076, 4.2735], + device='cuda:0'), covar=tensor([0.0433, 0.0370, 0.0298, 0.0465, 0.0293, 0.0323, 0.0320, 0.0440], + device='cuda:0'), in_proj_covar=tensor([0.0162, 0.0181, 0.0131, 0.0134, 0.0113, 0.0168, 0.0142, 0.0113], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:14:29,175 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=33511.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:14:30,082 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.498e+02 2.266e+02 2.777e+02 3.531e+02 6.731e+02, threshold=5.555e+02, percent-clipped=2.0 +2023-03-20 22:14:31,093 INFO [train.py:901] (0/2) Epoch 12, batch 2450, loss[loss=0.1663, simple_loss=0.2455, pruned_loss=0.04357, over 7253.00 frames. ], tot_loss[loss=0.1761, simple_loss=0.2477, pruned_loss=0.05227, over 1443790.97 frames. ], batch size: 55, lr: 1.33e-02, grad_scale: 16.0 +2023-03-20 22:14:40,719 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=33533.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:14:42,619 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4494, 3.5980, 3.3812, 3.5768, 3.4362, 3.5913, 3.8575, 3.9148], + device='cuda:0'), covar=tensor([0.0202, 0.0147, 0.0222, 0.0171, 0.0256, 0.0213, 0.0212, 0.0148], + device='cuda:0'), in_proj_covar=tensor([0.0097, 0.0097, 0.0093, 0.0103, 0.0096, 0.0081, 0.0076, 0.0079], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:14:44,572 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-20 22:14:45,731 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2450, 2.2815, 2.0654, 3.0917, 1.3216, 2.7048, 1.2694, 2.7636], + device='cuda:0'), covar=tensor([0.0084, 0.0926, 0.1642, 0.0080, 0.4621, 0.0082, 0.1087, 0.0221], + device='cuda:0'), in_proj_covar=tensor([0.0146, 0.0264, 0.0304, 0.0150, 0.0296, 0.0152, 0.0269, 0.0211], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 22:14:49,515 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.79 vs. limit=5.0 +2023-03-20 22:14:56,693 INFO [train.py:901] (0/2) Epoch 12, batch 2500, loss[loss=0.1655, simple_loss=0.2356, pruned_loss=0.0477, over 7317.00 frames. ], tot_loss[loss=0.1759, simple_loss=0.2477, pruned_loss=0.05201, over 1444002.01 frames. ], batch size: 80, lr: 1.33e-02, grad_scale: 32.0 +2023-03-20 22:15:10,487 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1393, 2.7630, 3.1004, 2.8935, 3.1351, 2.7232, 2.3689, 2.8753], + device='cuda:0'), covar=tensor([0.1091, 0.0376, 0.1359, 0.1964, 0.0936, 0.1202, 0.2700, 0.1714], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0035, 0.0034, 0.0035, 0.0032, 0.0031, 0.0046, 0.0034], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:15:10,894 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-20 22:15:22,708 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.564e+02 2.335e+02 2.794e+02 3.571e+02 7.375e+02, threshold=5.588e+02, percent-clipped=5.0 +2023-03-20 22:15:23,265 INFO [train.py:901] (0/2) Epoch 12, batch 2550, loss[loss=0.1807, simple_loss=0.2536, pruned_loss=0.05392, over 7269.00 frames. ], tot_loss[loss=0.1762, simple_loss=0.2479, pruned_loss=0.05223, over 1442853.92 frames. ], batch size: 77, lr: 1.33e-02, grad_scale: 16.0 +2023-03-20 22:15:26,857 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=33622.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:15:44,650 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.02 vs. limit=5.0 +2023-03-20 22:15:48,221 INFO [train.py:901] (0/2) Epoch 12, batch 2600, loss[loss=0.1497, simple_loss=0.2162, pruned_loss=0.04166, over 7166.00 frames. ], tot_loss[loss=0.1756, simple_loss=0.2474, pruned_loss=0.05189, over 1441396.38 frames. ], batch size: 39, lr: 1.33e-02, grad_scale: 16.0 +2023-03-20 22:15:50,752 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=33670.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:15:54,340 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=33677.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:16:03,556 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1165, 1.3020, 1.1257, 1.2736, 1.1730, 0.8366, 0.9367, 0.7232], + device='cuda:0'), covar=tensor([0.0134, 0.0082, 0.0277, 0.0098, 0.0114, 0.0191, 0.0114, 0.0136], + device='cuda:0'), in_proj_covar=tensor([0.0021, 0.0019, 0.0019, 0.0019, 0.0020, 0.0018, 0.0020, 0.0025], + device='cuda:0'), out_proj_covar=tensor([2.4845e-05, 2.1831e-05, 2.4039e-05, 2.1017e-05, 2.4883e-05, 2.0626e-05, + 2.3367e-05, 3.1175e-05], device='cuda:0') +2023-03-20 22:16:10,455 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3826, 3.3038, 3.3316, 3.3536, 2.5208, 2.6108, 3.7456, 2.6956], + device='cuda:0'), covar=tensor([0.0122, 0.0158, 0.0234, 0.0225, 0.0340, 0.0513, 0.0285, 0.0758], + device='cuda:0'), in_proj_covar=tensor([0.0276, 0.0281, 0.0239, 0.0291, 0.0294, 0.0295, 0.0281, 0.0292], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 22:16:12,796 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.411e+02 2.314e+02 2.858e+02 3.397e+02 6.641e+02, threshold=5.717e+02, percent-clipped=1.0 +2023-03-20 22:16:13,315 INFO [train.py:901] (0/2) Epoch 12, batch 2650, loss[loss=0.1578, simple_loss=0.2313, pruned_loss=0.04216, over 7212.00 frames. ], tot_loss[loss=0.1761, simple_loss=0.2481, pruned_loss=0.05207, over 1443459.46 frames. ], batch size: 45, lr: 1.33e-02, grad_scale: 16.0 +2023-03-20 22:16:18,292 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=33725.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:16:24,654 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=33738.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 22:16:37,752 INFO [train.py:901] (0/2) Epoch 12, batch 2700, loss[loss=0.1929, simple_loss=0.2543, pruned_loss=0.06577, over 7350.00 frames. ], tot_loss[loss=0.176, simple_loss=0.2476, pruned_loss=0.05222, over 1443191.77 frames. ], batch size: 73, lr: 1.32e-02, grad_scale: 16.0 +2023-03-20 22:16:41,731 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=33773.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:16:45,875 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7150, 3.6236, 3.5840, 3.6933, 2.8565, 2.8331, 3.8309, 2.9289], + device='cuda:0'), covar=tensor([0.0175, 0.0189, 0.0223, 0.0212, 0.0342, 0.0509, 0.0235, 0.0773], + device='cuda:0'), in_proj_covar=tensor([0.0277, 0.0279, 0.0238, 0.0289, 0.0292, 0.0291, 0.0281, 0.0289], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 22:17:01,352 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=33811.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:17:02,646 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.586e+02 2.411e+02 2.993e+02 3.645e+02 8.242e+02, threshold=5.986e+02, percent-clipped=5.0 +2023-03-20 22:17:03,131 INFO [train.py:901] (0/2) Epoch 12, batch 2750, loss[loss=0.2004, simple_loss=0.2719, pruned_loss=0.06449, over 7250.00 frames. ], tot_loss[loss=0.176, simple_loss=0.2477, pruned_loss=0.05216, over 1445168.53 frames. ], batch size: 55, lr: 1.32e-02, grad_scale: 16.0 +2023-03-20 22:17:09,565 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=33828.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:17:14,095 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.87 vs. limit=2.0 +2023-03-20 22:17:20,798 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7971, 3.3209, 3.5972, 3.8264, 3.7964, 3.7466, 3.8461, 3.7003], + device='cuda:0'), covar=tensor([0.0025, 0.0075, 0.0042, 0.0028, 0.0031, 0.0028, 0.0036, 0.0043], + device='cuda:0'), in_proj_covar=tensor([0.0033, 0.0043, 0.0039, 0.0036, 0.0038, 0.0038, 0.0044, 0.0046], + device='cuda:0'), out_proj_covar=tensor([7.9362e-05, 1.1943e-04, 1.0851e-04, 9.1438e-05, 9.6025e-05, 9.7157e-05, + 1.2206e-04, 1.2060e-04], device='cuda:0') +2023-03-20 22:17:24,659 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=33859.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:17:27,450 INFO [train.py:901] (0/2) Epoch 12, batch 2800, loss[loss=0.1513, simple_loss=0.2114, pruned_loss=0.04565, over 6982.00 frames. ], tot_loss[loss=0.1764, simple_loss=0.2479, pruned_loss=0.05248, over 1440963.22 frames. ], batch size: 35, lr: 1.32e-02, grad_scale: 16.0 +2023-03-20 22:17:31,474 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2044, 3.1328, 2.7315, 3.0048, 2.4279, 2.1618, 3.2619, 2.1769], + device='cuda:0'), covar=tensor([0.0223, 0.0283, 0.0208, 0.0240, 0.0447, 0.0513, 0.0298, 0.0805], + device='cuda:0'), in_proj_covar=tensor([0.0279, 0.0279, 0.0238, 0.0289, 0.0292, 0.0291, 0.0279, 0.0289], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 22:17:40,062 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-12.pt +2023-03-20 22:17:58,374 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-20 22:17:58,426 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-20 22:17:58,452 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-20 22:17:58,469 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-20 22:17:58,645 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-20 22:17:58,661 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-20 22:17:58,672 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-20 22:17:59,106 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-20 22:17:59,183 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-20 22:17:59,199 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-20 22:17:59,215 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-20 22:17:59,381 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-20 22:17:59,387 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-20 22:17:59,394 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-20 22:17:59,535 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-20 22:17:59,538 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-20 22:17:59,597 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-20 22:18:01,882 INFO [train.py:901] (0/2) Epoch 13, batch 0, loss[loss=0.1947, simple_loss=0.2578, pruned_loss=0.06574, over 7336.00 frames. ], tot_loss[loss=0.1947, simple_loss=0.2578, pruned_loss=0.06574, over 7336.00 frames. ], batch size: 61, lr: 1.28e-02, grad_scale: 16.0 +2023-03-20 22:18:01,883 INFO [train.py:926] (0/2) Computing validation loss +2023-03-20 22:18:25,038 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6615, 4.9239, 4.9368, 4.8838, 4.5971, 4.4630, 4.9608, 4.5710], + device='cuda:0'), covar=tensor([0.0388, 0.0422, 0.0401, 0.0498, 0.0465, 0.0357, 0.0336, 0.0727], + device='cuda:0'), in_proj_covar=tensor([0.0162, 0.0178, 0.0132, 0.0132, 0.0111, 0.0167, 0.0147, 0.0115], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:18:27,578 INFO [train.py:935] (0/2) Epoch 13, validation: loss=0.1711, simple_loss=0.259, pruned_loss=0.04159, over 1622729.00 frames. +2023-03-20 22:18:27,579 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-20 22:18:34,124 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-20 22:18:40,142 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.660e+02 2.486e+02 3.069e+02 3.797e+02 7.499e+02, threshold=6.139e+02, percent-clipped=2.0 +2023-03-20 22:18:44,657 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-20 22:18:52,415 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-20 22:18:53,140 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.02 vs. limit=2.0 +2023-03-20 22:18:53,340 INFO [train.py:901] (0/2) Epoch 13, batch 50, loss[loss=0.1768, simple_loss=0.2477, pruned_loss=0.05292, over 7279.00 frames. ], tot_loss[loss=0.1734, simple_loss=0.2466, pruned_loss=0.0501, over 325625.77 frames. ], batch size: 77, lr: 1.28e-02, grad_scale: 16.0 +2023-03-20 22:18:53,490 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=33939.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:18:54,376 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-20 22:18:56,890 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-20 22:19:01,809 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.88 vs. limit=2.0 +2023-03-20 22:19:13,403 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-20 22:19:13,889 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-20 22:19:18,816 INFO [train.py:901] (0/2) Epoch 13, batch 100, loss[loss=0.1677, simple_loss=0.2381, pruned_loss=0.04864, over 7276.00 frames. ], tot_loss[loss=0.1752, simple_loss=0.2483, pruned_loss=0.05108, over 574212.63 frames. ], batch size: 52, lr: 1.28e-02, grad_scale: 16.0 +2023-03-20 22:19:24,768 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=34000.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:19:32,285 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.611e+02 2.332e+02 2.851e+02 3.491e+02 5.832e+02, threshold=5.703e+02, percent-clipped=0.0 +2023-03-20 22:19:41,881 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=34033.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 22:19:44,833 INFO [train.py:901] (0/2) Epoch 13, batch 150, loss[loss=0.171, simple_loss=0.2444, pruned_loss=0.04882, over 7259.00 frames. ], tot_loss[loss=0.176, simple_loss=0.2483, pruned_loss=0.05188, over 766081.97 frames. ], batch size: 64, lr: 1.27e-02, grad_scale: 16.0 +2023-03-20 22:19:47,479 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=34044.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:20:10,703 INFO [train.py:901] (0/2) Epoch 13, batch 200, loss[loss=0.1801, simple_loss=0.2559, pruned_loss=0.05214, over 7351.00 frames. ], tot_loss[loss=0.1749, simple_loss=0.2474, pruned_loss=0.05121, over 916935.85 frames. ], batch size: 73, lr: 1.27e-02, grad_scale: 16.0 +2023-03-20 22:20:14,695 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-20 22:20:19,315 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=34105.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:20:20,172 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-20 22:20:23,661 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.678e+02 2.373e+02 2.823e+02 3.333e+02 8.127e+02, threshold=5.645e+02, percent-clipped=4.0 +2023-03-20 22:20:26,717 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-20 22:20:30,814 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=34128.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:20:36,199 INFO [train.py:901] (0/2) Epoch 13, batch 250, loss[loss=0.1617, simple_loss=0.2354, pruned_loss=0.04396, over 7272.00 frames. ], tot_loss[loss=0.1731, simple_loss=0.2455, pruned_loss=0.05037, over 1031486.17 frames. ], batch size: 57, lr: 1.27e-02, grad_scale: 16.0 +2023-03-20 22:20:39,263 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-20 22:20:39,872 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=34145.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:20:55,415 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=34176.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:20:59,389 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-20 22:21:02,454 INFO [train.py:901] (0/2) Epoch 13, batch 300, loss[loss=0.1737, simple_loss=0.2468, pruned_loss=0.05031, over 7334.00 frames. ], tot_loss[loss=0.1742, simple_loss=0.2464, pruned_loss=0.05096, over 1122303.44 frames. ], batch size: 75, lr: 1.27e-02, grad_scale: 16.0 +2023-03-20 22:21:09,362 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-20 22:21:11,444 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=34206.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:21:15,224 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.470e+02 2.521e+02 2.911e+02 3.518e+02 6.846e+02, threshold=5.822e+02, percent-clipped=2.0 +2023-03-20 22:21:19,516 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.50 vs. limit=2.0 +2023-03-20 22:21:28,194 INFO [train.py:901] (0/2) Epoch 13, batch 350, loss[loss=0.1663, simple_loss=0.2385, pruned_loss=0.04707, over 7298.00 frames. ], tot_loss[loss=0.174, simple_loss=0.2462, pruned_loss=0.0509, over 1196069.49 frames. ], batch size: 59, lr: 1.27e-02, grad_scale: 16.0 +2023-03-20 22:21:41,076 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.09 vs. limit=5.0 +2023-03-20 22:21:41,455 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3691, 3.0195, 2.9009, 3.1172, 2.4991, 2.3205, 3.1700, 2.2169], + device='cuda:0'), covar=tensor([0.0221, 0.0218, 0.0254, 0.0282, 0.0337, 0.0390, 0.0331, 0.0846], + device='cuda:0'), in_proj_covar=tensor([0.0279, 0.0283, 0.0241, 0.0297, 0.0298, 0.0294, 0.0285, 0.0291], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 22:21:42,618 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.05 vs. limit=2.0 +2023-03-20 22:21:43,329 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-20 22:21:52,091 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5488, 2.7608, 2.4419, 3.3636, 2.6671, 3.0442, 2.7465, 2.8294], + device='cuda:0'), covar=tensor([0.1381, 0.0434, 0.2506, 0.0370, 0.0066, 0.0036, 0.0086, 0.0121], + device='cuda:0'), in_proj_covar=tensor([0.0287, 0.0226, 0.0272, 0.0252, 0.0124, 0.0120, 0.0144, 0.0161], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0001, 0.0001, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:21:53,954 INFO [train.py:901] (0/2) Epoch 13, batch 400, loss[loss=0.1611, simple_loss=0.2291, pruned_loss=0.04658, over 7148.00 frames. ], tot_loss[loss=0.1736, simple_loss=0.2455, pruned_loss=0.0508, over 1250111.67 frames. ], batch size: 41, lr: 1.27e-02, grad_scale: 16.0 +2023-03-20 22:21:57,097 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=34295.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:22:02,699 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8051, 2.3653, 2.7770, 2.6004, 2.8281, 2.8745, 2.2807, 2.7608], + device='cuda:0'), covar=tensor([0.1944, 0.0515, 0.2615, 0.2074, 0.1109, 0.1103, 0.3395, 0.1434], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0038, 0.0035, 0.0037, 0.0033, 0.0033, 0.0047, 0.0035], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:22:06,468 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.678e+02 2.410e+02 2.808e+02 3.305e+02 7.616e+02, threshold=5.616e+02, percent-clipped=2.0 +2023-03-20 22:22:16,670 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=34333.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:22:19,531 INFO [train.py:901] (0/2) Epoch 13, batch 450, loss[loss=0.1755, simple_loss=0.2459, pruned_loss=0.05252, over 7298.00 frames. ], tot_loss[loss=0.173, simple_loss=0.2451, pruned_loss=0.05048, over 1290648.45 frames. ], batch size: 57, lr: 1.27e-02, grad_scale: 16.0 +2023-03-20 22:22:24,704 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-20 22:22:25,197 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-20 22:22:41,571 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=34381.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:22:45,508 INFO [train.py:901] (0/2) Epoch 13, batch 500, loss[loss=0.1688, simple_loss=0.2431, pruned_loss=0.04727, over 7298.00 frames. ], tot_loss[loss=0.1727, simple_loss=0.2446, pruned_loss=0.05037, over 1322843.94 frames. ], batch size: 66, lr: 1.27e-02, grad_scale: 16.0 +2023-03-20 22:22:51,376 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=34400.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:22:54,452 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4649, 4.3061, 3.9675, 3.6787, 4.0000, 2.6866, 1.6877, 4.4262], + device='cuda:0'), covar=tensor([0.0017, 0.0034, 0.0056, 0.0057, 0.0041, 0.0325, 0.0478, 0.0027], + device='cuda:0'), in_proj_covar=tensor([0.0068, 0.0062, 0.0082, 0.0070, 0.0087, 0.0107, 0.0109, 0.0075], + device='cuda:0'), out_proj_covar=tensor([8.8127e-05, 9.0917e-05, 1.0923e-04, 1.0099e-04, 1.1695e-04, 1.4594e-04, + 1.4939e-04, 1.0037e-04], device='cuda:0') +2023-03-20 22:22:58,935 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.527e+02 2.514e+02 3.021e+02 3.896e+02 7.702e+02, threshold=6.043e+02, percent-clipped=7.0 +2023-03-20 22:22:58,975 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-20 22:23:01,002 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-20 22:23:01,517 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-20 22:23:03,943 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-20 22:23:06,634 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0857, 1.5831, 1.6982, 1.5734, 1.4948, 1.4767, 1.6628, 1.4421], + device='cuda:0'), covar=tensor([0.0441, 0.0421, 0.0115, 0.0056, 0.0524, 0.0341, 0.0131, 0.0254], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0020, 0.0019, 0.0018, 0.0020, 0.0020, 0.0019, 0.0021], + device='cuda:0'), out_proj_covar=tensor([5.1330e-05, 4.9183e-05, 4.5319e-05, 4.1027e-05, 5.0346e-05, 4.7899e-05, + 4.7081e-05, 5.3110e-05], device='cuda:0') +2023-03-20 22:23:08,012 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-20 22:23:11,458 INFO [train.py:901] (0/2) Epoch 13, batch 550, loss[loss=0.1759, simple_loss=0.2498, pruned_loss=0.051, over 7301.00 frames. ], tot_loss[loss=0.1724, simple_loss=0.2444, pruned_loss=0.05016, over 1350632.69 frames. ], batch size: 80, lr: 1.27e-02, grad_scale: 16.0 +2023-03-20 22:23:15,271 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.83 vs. limit=5.0 +2023-03-20 22:23:19,201 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-20 22:23:27,115 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-20 22:23:30,155 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-20 22:23:37,758 INFO [train.py:901] (0/2) Epoch 13, batch 600, loss[loss=0.1736, simple_loss=0.2484, pruned_loss=0.04943, over 7307.00 frames. ], tot_loss[loss=0.1713, simple_loss=0.2433, pruned_loss=0.04963, over 1369241.26 frames. ], batch size: 83, lr: 1.27e-02, grad_scale: 16.0 +2023-03-20 22:23:38,268 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-20 22:23:43,851 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=34501.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:23:50,212 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.492e+02 2.326e+02 2.723e+02 3.243e+02 6.434e+02, threshold=5.446e+02, percent-clipped=1.0 +2023-03-20 22:23:53,287 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-20 22:24:01,843 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-20 22:24:03,307 INFO [train.py:901] (0/2) Epoch 13, batch 650, loss[loss=0.1988, simple_loss=0.2651, pruned_loss=0.06627, over 7297.00 frames. ], tot_loss[loss=0.1719, simple_loss=0.2441, pruned_loss=0.0499, over 1386046.74 frames. ], batch size: 77, lr: 1.27e-02, grad_scale: 16.0 +2023-03-20 22:24:03,487 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7907, 2.0159, 1.6417, 2.5192, 2.0099, 2.1946, 1.4974, 1.7265], + device='cuda:0'), covar=tensor([0.1829, 0.0864, 0.3267, 0.0593, 0.0085, 0.0051, 0.0099, 0.0185], + device='cuda:0'), in_proj_covar=tensor([0.0281, 0.0222, 0.0268, 0.0245, 0.0124, 0.0118, 0.0143, 0.0160], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0001, 0.0001, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:24:18,755 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-20 22:24:28,272 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-20 22:24:28,760 INFO [train.py:901] (0/2) Epoch 13, batch 700, loss[loss=0.1879, simple_loss=0.26, pruned_loss=0.05788, over 7331.00 frames. ], tot_loss[loss=0.1722, simple_loss=0.2442, pruned_loss=0.05009, over 1396752.29 frames. ], batch size: 61, lr: 1.26e-02, grad_scale: 16.0 +2023-03-20 22:24:30,938 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=34593.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:24:31,900 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=34595.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:24:41,586 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.640e+02 2.349e+02 2.874e+02 3.606e+02 6.816e+02, threshold=5.747e+02, percent-clipped=3.0 +2023-03-20 22:24:46,316 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=34622.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:24:46,757 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.0300, 5.5558, 5.5865, 5.4996, 5.2628, 5.2612, 5.6720, 5.4011], + device='cuda:0'), covar=tensor([0.0376, 0.0341, 0.0403, 0.0474, 0.0326, 0.0239, 0.0282, 0.0465], + device='cuda:0'), in_proj_covar=tensor([0.0163, 0.0183, 0.0134, 0.0133, 0.0114, 0.0170, 0.0144, 0.0115], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:24:48,312 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=34626.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:24:52,701 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-20 22:24:52,718 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-20 22:24:54,696 INFO [train.py:901] (0/2) Epoch 13, batch 750, loss[loss=0.1816, simple_loss=0.2549, pruned_loss=0.05415, over 7271.00 frames. ], tot_loss[loss=0.1731, simple_loss=0.2449, pruned_loss=0.05062, over 1404510.39 frames. ], batch size: 52, lr: 1.26e-02, grad_scale: 16.0 +2023-03-20 22:24:56,759 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=34643.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:25:02,418 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=34654.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:25:07,481 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-20 22:25:11,946 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-20 22:25:16,584 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2015, 2.5850, 2.2410, 2.3407, 2.4596, 2.1464, 2.5829, 2.3936], + device='cuda:0'), covar=tensor([0.1284, 0.0455, 0.0743, 0.1241, 0.0624, 0.0367, 0.0948, 0.1142], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0040, 0.0043, 0.0039, 0.0039, 0.0038, 0.0042, 0.0036], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 22:25:17,581 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=34683.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 22:25:18,421 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-20 22:25:19,430 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-20 22:25:19,568 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=34687.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:25:20,030 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5024, 4.3830, 3.7625, 3.6284, 4.0043, 2.4887, 1.5608, 4.3225], + device='cuda:0'), covar=tensor([0.0023, 0.0033, 0.0097, 0.0073, 0.0062, 0.0432, 0.0592, 0.0046], + device='cuda:0'), in_proj_covar=tensor([0.0069, 0.0063, 0.0083, 0.0071, 0.0088, 0.0108, 0.0110, 0.0076], + device='cuda:0'), out_proj_covar=tensor([8.9627e-05, 9.2192e-05, 1.0973e-04, 1.0233e-04, 1.1795e-04, 1.4689e-04, + 1.5023e-04, 1.0225e-04], device='cuda:0') +2023-03-20 22:25:20,417 INFO [train.py:901] (0/2) Epoch 13, batch 800, loss[loss=0.1454, simple_loss=0.2222, pruned_loss=0.03427, over 7345.00 frames. ], tot_loss[loss=0.1728, simple_loss=0.2448, pruned_loss=0.05043, over 1413091.54 frames. ], batch size: 44, lr: 1.26e-02, grad_scale: 16.0 +2023-03-20 22:25:25,936 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=34700.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:25:30,988 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-20 22:25:31,568 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=34710.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 22:25:33,446 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.535e+02 2.193e+02 2.697e+02 3.424e+02 6.128e+02, threshold=5.393e+02, percent-clipped=1.0 +2023-03-20 22:25:45,933 INFO [train.py:901] (0/2) Epoch 13, batch 850, loss[loss=0.2005, simple_loss=0.2713, pruned_loss=0.06483, over 7290.00 frames. ], tot_loss[loss=0.1733, simple_loss=0.2451, pruned_loss=0.05074, over 1420690.84 frames. ], batch size: 68, lr: 1.26e-02, grad_scale: 16.0 +2023-03-20 22:25:50,367 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-20 22:25:50,376 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-20 22:25:50,411 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=34748.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:25:56,393 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-20 22:25:59,907 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-20 22:26:02,563 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=34771.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 22:26:11,306 INFO [train.py:901] (0/2) Epoch 13, batch 900, loss[loss=0.1726, simple_loss=0.2536, pruned_loss=0.04578, over 7147.00 frames. ], tot_loss[loss=0.173, simple_loss=0.2448, pruned_loss=0.05061, over 1424345.56 frames. ], batch size: 98, lr: 1.26e-02, grad_scale: 16.0 +2023-03-20 22:26:17,764 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=34801.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:26:24,751 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.382e+02 2.135e+02 2.636e+02 3.552e+02 6.481e+02, threshold=5.271e+02, percent-clipped=4.0 +2023-03-20 22:26:28,951 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2594, 3.4317, 3.2336, 3.2740, 3.2696, 3.1465, 3.4069, 3.5256], + device='cuda:0'), covar=tensor([0.0347, 0.0247, 0.0351, 0.0378, 0.0377, 0.0480, 0.0533, 0.0444], + device='cuda:0'), in_proj_covar=tensor([0.0096, 0.0100, 0.0094, 0.0102, 0.0095, 0.0081, 0.0076, 0.0080], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:26:29,483 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([0.9905, 0.9925, 1.1868, 1.2662, 1.2513, 1.3782, 1.2558, 1.2745], + device='cuda:0'), covar=tensor([0.1716, 0.2185, 0.0748, 0.1022, 0.2046, 0.1380, 0.1173, 0.1831], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0044, 0.0033, 0.0031, 0.0037, 0.0035, 0.0043, 0.0036], + device='cuda:0'), out_proj_covar=tensor([9.4934e-05, 1.0625e-04, 8.3038e-05, 8.1831e-05, 9.3677e-05, 9.2584e-05, + 1.0486e-04, 9.3995e-05], device='cuda:0') +2023-03-20 22:26:34,927 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1287, 3.7941, 3.7930, 3.8390, 3.2567, 3.7569, 3.9663, 3.6568], + device='cuda:0'), covar=tensor([0.0143, 0.0166, 0.0154, 0.0168, 0.0473, 0.0149, 0.0210, 0.0170], + device='cuda:0'), in_proj_covar=tensor([0.0058, 0.0066, 0.0066, 0.0057, 0.0112, 0.0072, 0.0069, 0.0069], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0001, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:26:37,308 INFO [train.py:901] (0/2) Epoch 13, batch 950, loss[loss=0.2016, simple_loss=0.2727, pruned_loss=0.06526, over 7296.00 frames. ], tot_loss[loss=0.1732, simple_loss=0.2453, pruned_loss=0.05054, over 1430122.34 frames. ], batch size: 86, lr: 1.26e-02, grad_scale: 16.0 +2023-03-20 22:26:38,800 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-20 22:26:42,900 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=34849.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:26:44,561 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1667, 1.2993, 1.2227, 1.3358, 1.1169, 1.0548, 1.2185, 0.9333], + device='cuda:0'), covar=tensor([0.0123, 0.0104, 0.0167, 0.0075, 0.0113, 0.0094, 0.0086, 0.0107], + device='cuda:0'), in_proj_covar=tensor([0.0021, 0.0020, 0.0020, 0.0019, 0.0021, 0.0019, 0.0020, 0.0026], + device='cuda:0'), out_proj_covar=tensor([2.5508e-05, 2.3137e-05, 2.4483e-05, 2.0944e-05, 2.6091e-05, 2.1532e-05, + 2.3352e-05, 3.1684e-05], device='cuda:0') +2023-03-20 22:27:01,864 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-20 22:27:02,798 INFO [train.py:901] (0/2) Epoch 13, batch 1000, loss[loss=0.1616, simple_loss=0.2415, pruned_loss=0.04092, over 7326.00 frames. ], tot_loss[loss=0.1731, simple_loss=0.245, pruned_loss=0.05063, over 1432832.25 frames. ], batch size: 49, lr: 1.26e-02, grad_scale: 16.0 +2023-03-20 22:27:02,950 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0621, 1.1094, 1.3981, 1.4856, 1.2910, 1.6273, 1.4784, 1.3083], + device='cuda:0'), covar=tensor([0.1598, 0.4029, 0.0771, 0.1010, 0.3414, 0.0990, 0.2018, 0.4586], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0044, 0.0034, 0.0032, 0.0038, 0.0036, 0.0044, 0.0037], + device='cuda:0'), out_proj_covar=tensor([9.5365e-05, 1.0647e-04, 8.3758e-05, 8.3056e-05, 9.5131e-05, 9.3346e-05, + 1.0572e-04, 9.5743e-05], device='cuda:0') +2023-03-20 22:27:15,916 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.563e+02 2.342e+02 2.774e+02 3.398e+02 6.564e+02, threshold=5.548e+02, percent-clipped=3.0 +2023-03-20 22:27:22,401 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-20 22:27:26,413 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2669, 2.6688, 2.1535, 2.3225, 2.3406, 1.7996, 2.5456, 2.2942], + device='cuda:0'), covar=tensor([0.0774, 0.0671, 0.1316, 0.1229, 0.1598, 0.1060, 0.1315, 0.1152], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0040, 0.0043, 0.0039, 0.0040, 0.0039, 0.0042, 0.0037], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 22:27:28,814 INFO [train.py:901] (0/2) Epoch 13, batch 1050, loss[loss=0.175, simple_loss=0.2452, pruned_loss=0.05239, over 7210.00 frames. ], tot_loss[loss=0.172, simple_loss=0.2439, pruned_loss=0.04999, over 1433098.85 frames. ], batch size: 45, lr: 1.26e-02, grad_scale: 16.0 +2023-03-20 22:27:33,961 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=34949.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:27:43,253 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-20 22:27:48,361 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-20 22:27:48,889 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=34978.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 22:27:50,885 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=34982.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:27:54,366 INFO [train.py:901] (0/2) Epoch 13, batch 1100, loss[loss=0.1983, simple_loss=0.2661, pruned_loss=0.06529, over 7295.00 frames. ], tot_loss[loss=0.1719, simple_loss=0.2439, pruned_loss=0.04994, over 1434047.79 frames. ], batch size: 86, lr: 1.26e-02, grad_scale: 16.0 +2023-03-20 22:28:05,933 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0705, 4.0324, 3.5666, 3.3557, 3.2547, 2.0378, 1.4506, 4.1019], + device='cuda:0'), covar=tensor([0.0041, 0.0047, 0.0102, 0.0074, 0.0117, 0.0505, 0.0613, 0.0047], + device='cuda:0'), in_proj_covar=tensor([0.0071, 0.0065, 0.0084, 0.0072, 0.0089, 0.0110, 0.0113, 0.0077], + device='cuda:0'), out_proj_covar=tensor([9.2056e-05, 9.3362e-05, 1.1195e-04, 1.0444e-04, 1.1955e-04, 1.4931e-04, + 1.5401e-04, 1.0342e-04], device='cuda:0') +2023-03-20 22:28:07,783 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.551e+02 2.283e+02 2.674e+02 3.489e+02 7.273e+02, threshold=5.348e+02, percent-clipped=5.0 +2023-03-20 22:28:13,767 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.84 vs. limit=2.0 +2023-03-20 22:28:16,917 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-20 22:28:17,423 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 22:28:20,442 INFO [train.py:901] (0/2) Epoch 13, batch 1150, loss[loss=0.1757, simple_loss=0.2511, pruned_loss=0.0501, over 7351.00 frames. ], tot_loss[loss=0.1722, simple_loss=0.244, pruned_loss=0.05023, over 1435311.49 frames. ], batch size: 73, lr: 1.26e-02, grad_scale: 16.0 +2023-03-20 22:28:20,627 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4142, 2.9936, 2.9458, 2.9946, 2.3110, 2.4475, 3.3324, 2.2771], + device='cuda:0'), covar=tensor([0.0198, 0.0228, 0.0214, 0.0245, 0.0331, 0.0495, 0.0250, 0.0909], + device='cuda:0'), in_proj_covar=tensor([0.0281, 0.0286, 0.0237, 0.0293, 0.0294, 0.0293, 0.0281, 0.0286], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 22:28:25,116 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6229, 2.7309, 1.9710, 2.9632, 1.8197, 2.8012, 1.4767, 1.8809], + device='cuda:0'), covar=tensor([0.0216, 0.0424, 0.1673, 0.0366, 0.0328, 0.0384, 0.2372, 0.1508], + device='cuda:0'), in_proj_covar=tensor([0.0209, 0.0218, 0.0296, 0.0216, 0.0238, 0.0218, 0.0262, 0.0281], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 22:28:29,576 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-20 22:28:30,547 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-20 22:28:31,689 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3863, 1.8141, 1.8739, 1.6878, 1.6323, 1.6139, 1.4751, 1.4053], + device='cuda:0'), covar=tensor([0.0324, 0.0205, 0.0146, 0.0082, 0.0331, 0.0213, 0.0227, 0.0247], + device='cuda:0'), in_proj_covar=tensor([0.0022, 0.0020, 0.0021, 0.0020, 0.0020, 0.0019, 0.0020, 0.0022], + device='cuda:0'), out_proj_covar=tensor([5.4291e-05, 5.0068e-05, 4.8256e-05, 4.4724e-05, 5.1359e-05, 4.8361e-05, + 4.9149e-05, 5.5391e-05], device='cuda:0') +2023-03-20 22:28:34,725 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=35066.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 22:28:35,517 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.82 vs. limit=2.0 +2023-03-20 22:28:35,913 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.37 vs. limit=2.0 +2023-03-20 22:28:46,123 INFO [train.py:901] (0/2) Epoch 13, batch 1200, loss[loss=0.2011, simple_loss=0.2699, pruned_loss=0.06615, over 7212.00 frames. ], tot_loss[loss=0.1725, simple_loss=0.2443, pruned_loss=0.05034, over 1436803.31 frames. ], batch size: 93, lr: 1.26e-02, grad_scale: 16.0 +2023-03-20 22:28:47,743 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=35092.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:28:59,136 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.459e+02 2.344e+02 2.798e+02 3.632e+02 7.223e+02, threshold=5.596e+02, percent-clipped=4.0 +2023-03-20 22:29:03,156 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-20 22:29:11,745 INFO [train.py:901] (0/2) Epoch 13, batch 1250, loss[loss=0.1644, simple_loss=0.2447, pruned_loss=0.04199, over 7286.00 frames. ], tot_loss[loss=0.1712, simple_loss=0.2431, pruned_loss=0.04963, over 1435442.56 frames. ], batch size: 70, lr: 1.25e-02, grad_scale: 16.0 +2023-03-20 22:29:19,487 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=35153.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:29:26,840 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-20 22:29:27,712 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.01 vs. limit=2.0 +2023-03-20 22:29:30,870 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-20 22:29:31,867 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-20 22:29:37,103 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0289, 4.0891, 3.5502, 3.3268, 3.4909, 2.3006, 1.5588, 4.0089], + device='cuda:0'), covar=tensor([0.0035, 0.0040, 0.0067, 0.0071, 0.0090, 0.0404, 0.0559, 0.0041], + device='cuda:0'), in_proj_covar=tensor([0.0071, 0.0065, 0.0084, 0.0073, 0.0089, 0.0108, 0.0114, 0.0078], + device='cuda:0'), out_proj_covar=tensor([9.2536e-05, 9.2772e-05, 1.1154e-04, 1.0560e-04, 1.1905e-04, 1.4686e-04, + 1.5378e-04, 1.0377e-04], device='cuda:0') +2023-03-20 22:29:37,975 INFO [train.py:901] (0/2) Epoch 13, batch 1300, loss[loss=0.165, simple_loss=0.2384, pruned_loss=0.04578, over 7325.00 frames. ], tot_loss[loss=0.1709, simple_loss=0.2428, pruned_loss=0.04951, over 1435777.40 frames. ], batch size: 59, lr: 1.25e-02, grad_scale: 16.0 +2023-03-20 22:29:50,990 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.414e+02 2.374e+02 2.864e+02 3.465e+02 9.066e+02, threshold=5.728e+02, percent-clipped=1.0 +2023-03-20 22:29:55,485 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-20 22:29:57,073 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4735, 1.9264, 2.1520, 1.6975, 1.8089, 1.7855, 1.7612, 1.5699], + device='cuda:0'), covar=tensor([0.0308, 0.0215, 0.0149, 0.0069, 0.0282, 0.0168, 0.0128, 0.0196], + device='cuda:0'), in_proj_covar=tensor([0.0022, 0.0019, 0.0021, 0.0020, 0.0020, 0.0020, 0.0020, 0.0021], + device='cuda:0'), out_proj_covar=tensor([5.4734e-05, 4.8988e-05, 4.8463e-05, 4.4366e-05, 5.0381e-05, 4.8715e-05, + 4.8778e-05, 5.4864e-05], device='cuda:0') +2023-03-20 22:29:57,426 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-20 22:30:00,889 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-20 22:30:03,317 INFO [train.py:901] (0/2) Epoch 13, batch 1350, loss[loss=0.1589, simple_loss=0.2296, pruned_loss=0.04407, over 7335.00 frames. ], tot_loss[loss=0.1719, simple_loss=0.244, pruned_loss=0.04995, over 1439374.88 frames. ], batch size: 44, lr: 1.25e-02, grad_scale: 16.0 +2023-03-20 22:30:09,032 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=35249.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:30:12,429 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-20 22:30:19,645 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9525, 3.5278, 3.1495, 3.6669, 2.8581, 2.6657, 3.8973, 2.7849], + device='cuda:0'), covar=tensor([0.0224, 0.0274, 0.0349, 0.0288, 0.0478, 0.0715, 0.0326, 0.1110], + device='cuda:0'), in_proj_covar=tensor([0.0281, 0.0289, 0.0240, 0.0300, 0.0294, 0.0296, 0.0285, 0.0290], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 22:30:23,999 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=35278.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 22:30:25,984 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=35282.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:30:29,434 INFO [train.py:901] (0/2) Epoch 13, batch 1400, loss[loss=0.1641, simple_loss=0.2313, pruned_loss=0.04846, over 7268.00 frames. ], tot_loss[loss=0.1724, simple_loss=0.2441, pruned_loss=0.05029, over 1438631.37 frames. ], batch size: 52, lr: 1.25e-02, grad_scale: 16.0 +2023-03-20 22:30:33,588 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=35297.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:30:42,143 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.592e+02 2.389e+02 2.860e+02 3.521e+02 6.843e+02, threshold=5.719e+02, percent-clipped=2.0 +2023-03-20 22:30:44,722 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-20 22:30:48,286 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=35326.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:30:50,290 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.96 vs. limit=2.0 +2023-03-20 22:30:51,013 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=35330.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:30:52,103 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5316, 2.3822, 2.7758, 2.5402, 2.7870, 2.5098, 2.2753, 2.5636], + device='cuda:0'), covar=tensor([0.1786, 0.0490, 0.1023, 0.1648, 0.0694, 0.1162, 0.2327, 0.1945], + device='cuda:0'), in_proj_covar=tensor([0.0036, 0.0039, 0.0035, 0.0037, 0.0032, 0.0033, 0.0048, 0.0036], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:30:55,452 INFO [train.py:901] (0/2) Epoch 13, batch 1450, loss[loss=0.1639, simple_loss=0.2431, pruned_loss=0.04229, over 7366.00 frames. ], tot_loss[loss=0.1729, simple_loss=0.2446, pruned_loss=0.05054, over 1438045.65 frames. ], batch size: 51, lr: 1.25e-02, grad_scale: 16.0 +2023-03-20 22:30:58,661 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=35345.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 22:31:09,111 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-20 22:31:09,731 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=35366.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 22:31:13,740 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=35374.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:31:20,995 INFO [train.py:901] (0/2) Epoch 13, batch 1500, loss[loss=0.2189, simple_loss=0.2837, pruned_loss=0.07708, over 7314.00 frames. ], tot_loss[loss=0.1728, simple_loss=0.2446, pruned_loss=0.05052, over 1438466.68 frames. ], batch size: 75, lr: 1.25e-02, grad_scale: 16.0 +2023-03-20 22:31:24,915 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-20 22:31:26,907 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2257, 1.4625, 1.2155, 1.4818, 1.2816, 1.0902, 1.2146, 0.7004], + device='cuda:0'), covar=tensor([0.0181, 0.0096, 0.0226, 0.0074, 0.0193, 0.0105, 0.0071, 0.0194], + device='cuda:0'), in_proj_covar=tensor([0.0022, 0.0021, 0.0021, 0.0020, 0.0023, 0.0021, 0.0021, 0.0027], + device='cuda:0'), out_proj_covar=tensor([2.6682e-05, 2.4292e-05, 2.5586e-05, 2.2083e-05, 2.8057e-05, 2.2886e-05, + 2.4594e-05, 3.3014e-05], device='cuda:0') +2023-03-20 22:31:29,913 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3733, 2.7034, 2.0902, 2.5096, 2.4655, 2.1722, 2.6090, 2.4597], + device='cuda:0'), covar=tensor([0.0805, 0.0429, 0.1973, 0.1040, 0.0933, 0.0654, 0.0594, 0.1168], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0041, 0.0046, 0.0040, 0.0040, 0.0041, 0.0043, 0.0038], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0001, 0.0002, 0.0001, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-20 22:31:29,932 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=35406.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 22:31:33,723 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.695e+02 2.282e+02 2.834e+02 3.480e+02 6.162e+02, threshold=5.667e+02, percent-clipped=1.0 +2023-03-20 22:31:33,789 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=35414.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 22:31:45,075 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=35435.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:31:46,864 INFO [train.py:901] (0/2) Epoch 13, batch 1550, loss[loss=0.1764, simple_loss=0.245, pruned_loss=0.05387, over 7257.00 frames. ], tot_loss[loss=0.1739, simple_loss=0.2453, pruned_loss=0.05122, over 1438283.71 frames. ], batch size: 47, lr: 1.25e-02, grad_scale: 16.0 +2023-03-20 22:31:49,465 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-20 22:31:51,966 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=35448.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:32:02,665 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.88 vs. limit=5.0 +2023-03-20 22:32:12,238 INFO [train.py:901] (0/2) Epoch 13, batch 1600, loss[loss=0.1441, simple_loss=0.2133, pruned_loss=0.03739, over 7197.00 frames. ], tot_loss[loss=0.1736, simple_loss=0.2453, pruned_loss=0.0509, over 1438865.38 frames. ], batch size: 39, lr: 1.25e-02, grad_scale: 16.0 +2023-03-20 22:32:17,138 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-20 22:32:17,635 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-20 22:32:21,739 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-20 22:32:25,101 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.433e+02 2.407e+02 2.844e+02 3.510e+02 6.841e+02, threshold=5.688e+02, percent-clipped=1.0 +2023-03-20 22:32:32,161 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-20 22:32:33,821 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=35530.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:32:36,195 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-20 22:32:38,169 INFO [train.py:901] (0/2) Epoch 13, batch 1650, loss[loss=0.1764, simple_loss=0.2508, pruned_loss=0.05099, over 7351.00 frames. ], tot_loss[loss=0.1723, simple_loss=0.2443, pruned_loss=0.05014, over 1439514.56 frames. ], batch size: 73, lr: 1.25e-02, grad_scale: 16.0 +2023-03-20 22:32:44,285 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2369, 1.1160, 1.2710, 1.5333, 1.5262, 1.4285, 1.4718, 1.3674], + device='cuda:0'), covar=tensor([0.1217, 0.2075, 0.0838, 0.1020, 0.2033, 0.3151, 0.1611, 0.4014], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0047, 0.0034, 0.0032, 0.0040, 0.0037, 0.0045, 0.0038], + device='cuda:0'), out_proj_covar=tensor([9.8667e-05, 1.1395e-04, 8.7234e-05, 8.5961e-05, 9.9314e-05, 9.8059e-05, + 1.1035e-04, 9.9126e-05], device='cuda:0') +2023-03-20 22:32:44,678 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-20 22:33:02,013 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 22:33:03,522 INFO [train.py:901] (0/2) Epoch 13, batch 1700, loss[loss=0.175, simple_loss=0.2549, pruned_loss=0.04753, over 7322.00 frames. ], tot_loss[loss=0.1713, simple_loss=0.2436, pruned_loss=0.04949, over 1441606.36 frames. ], batch size: 80, lr: 1.25e-02, grad_scale: 16.0 +2023-03-20 22:33:04,698 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=35591.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:33:06,706 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-20 22:33:08,838 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.87 vs. limit=5.0 +2023-03-20 22:33:17,207 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5146, 3.0751, 3.1170, 3.2563, 2.4590, 2.5229, 3.5246, 2.4521], + device='cuda:0'), covar=tensor([0.0202, 0.0200, 0.0191, 0.0232, 0.0334, 0.0497, 0.0305, 0.0756], + device='cuda:0'), in_proj_covar=tensor([0.0280, 0.0287, 0.0240, 0.0299, 0.0294, 0.0297, 0.0287, 0.0290], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 22:33:17,500 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.397e+02 2.386e+02 2.727e+02 3.355e+02 1.197e+03, threshold=5.455e+02, percent-clipped=3.0 +2023-03-20 22:33:18,517 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-20 22:33:29,977 INFO [train.py:901] (0/2) Epoch 13, batch 1750, loss[loss=0.1705, simple_loss=0.2519, pruned_loss=0.04455, over 7306.00 frames. ], tot_loss[loss=0.1709, simple_loss=0.2433, pruned_loss=0.04929, over 1442272.16 frames. ], batch size: 80, lr: 1.25e-02, grad_scale: 16.0 +2023-03-20 22:33:42,024 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-20 22:33:43,031 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-20 22:33:53,316 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9573, 2.5761, 2.4882, 2.6575, 2.1108, 2.2043, 2.8657, 2.0013], + device='cuda:0'), covar=tensor([0.0239, 0.0224, 0.0216, 0.0284, 0.0325, 0.0425, 0.0325, 0.0804], + device='cuda:0'), in_proj_covar=tensor([0.0280, 0.0287, 0.0242, 0.0299, 0.0295, 0.0296, 0.0287, 0.0292], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 22:33:54,442 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.96 vs. limit=2.0 +2023-03-20 22:33:55,677 INFO [train.py:901] (0/2) Epoch 13, batch 1800, loss[loss=0.1661, simple_loss=0.241, pruned_loss=0.04564, over 7335.00 frames. ], tot_loss[loss=0.1709, simple_loss=0.2432, pruned_loss=0.04929, over 1442338.32 frames. ], batch size: 54, lr: 1.25e-02, grad_scale: 16.0 +2023-03-20 22:34:02,311 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=35701.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 22:34:04,862 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-20 22:34:09,416 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.652e+02 2.292e+02 2.730e+02 3.144e+02 6.344e+02, threshold=5.459e+02, percent-clipped=2.0 +2023-03-20 22:34:10,580 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4947, 2.7851, 2.3599, 3.3395, 3.0246, 2.7169, 2.9141, 2.7623], + device='cuda:0'), covar=tensor([0.1704, 0.0465, 0.2476, 0.0457, 0.0068, 0.0062, 0.0090, 0.0194], + device='cuda:0'), in_proj_covar=tensor([0.0286, 0.0229, 0.0269, 0.0253, 0.0124, 0.0122, 0.0147, 0.0167], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0001, 0.0001, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:34:17,182 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=35730.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:34:19,117 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-20 22:34:19,408 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.02 vs. limit=2.0 +2023-03-20 22:34:21,608 INFO [train.py:901] (0/2) Epoch 13, batch 1850, loss[loss=0.1501, simple_loss=0.2338, pruned_loss=0.03326, over 7235.00 frames. ], tot_loss[loss=0.1709, simple_loss=0.2432, pruned_loss=0.04932, over 1443443.60 frames. ], batch size: 93, lr: 1.24e-02, grad_scale: 16.0 +2023-03-20 22:34:26,184 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=35748.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:34:29,140 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-20 22:34:30,455 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.02 vs. limit=2.0 +2023-03-20 22:34:38,460 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=35771.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:34:43,228 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.90 vs. limit=2.0 +2023-03-20 22:34:46,012 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-20 22:34:47,983 INFO [train.py:901] (0/2) Epoch 13, batch 1900, loss[loss=0.1828, simple_loss=0.2555, pruned_loss=0.0551, over 7300.00 frames. ], tot_loss[loss=0.1717, simple_loss=0.244, pruned_loss=0.04965, over 1441638.65 frames. ], batch size: 86, lr: 1.24e-02, grad_scale: 16.0 +2023-03-20 22:34:51,513 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=35796.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:34:53,099 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7460, 3.7195, 3.2503, 3.1977, 3.0331, 2.2592, 1.7009, 3.7777], + device='cuda:0'), covar=tensor([0.0032, 0.0045, 0.0070, 0.0061, 0.0084, 0.0362, 0.0489, 0.0036], + device='cuda:0'), in_proj_covar=tensor([0.0071, 0.0064, 0.0083, 0.0070, 0.0088, 0.0107, 0.0111, 0.0076], + device='cuda:0'), out_proj_covar=tensor([9.3263e-05, 9.3015e-05, 1.1063e-04, 1.0125e-04, 1.1715e-04, 1.4525e-04, + 1.4988e-04, 1.0138e-04], device='cuda:0') +2023-03-20 22:34:59,501 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=35811.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:35:01,329 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.430e+02 2.515e+02 2.934e+02 3.693e+02 9.378e+02, threshold=5.869e+02, percent-clipped=4.0 +2023-03-20 22:35:09,929 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-20 22:35:10,068 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=35832.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:35:13,458 INFO [train.py:901] (0/2) Epoch 13, batch 1950, loss[loss=0.1422, simple_loss=0.1941, pruned_loss=0.04517, over 6056.00 frames. ], tot_loss[loss=0.1712, simple_loss=0.2437, pruned_loss=0.04929, over 1441500.86 frames. ], batch size: 26, lr: 1.24e-02, grad_scale: 16.0 +2023-03-20 22:35:20,958 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-20 22:35:25,848 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-20 22:35:26,350 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-20 22:35:31,043 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=35872.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:35:37,877 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=35886.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:35:39,345 INFO [train.py:901] (0/2) Epoch 13, batch 2000, loss[loss=0.1497, simple_loss=0.2237, pruned_loss=0.03782, over 7133.00 frames. ], tot_loss[loss=0.1716, simple_loss=0.2441, pruned_loss=0.04959, over 1442514.65 frames. ], batch size: 41, lr: 1.24e-02, grad_scale: 16.0 +2023-03-20 22:35:43,302 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-20 22:35:47,945 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3704, 2.7519, 3.1985, 3.2895, 3.3175, 3.2330, 3.2516, 2.7884], + device='cuda:0'), covar=tensor([0.0045, 0.0168, 0.0062, 0.0063, 0.0057, 0.0062, 0.0088, 0.0106], + device='cuda:0'), in_proj_covar=tensor([0.0032, 0.0044, 0.0038, 0.0038, 0.0038, 0.0040, 0.0043, 0.0048], + device='cuda:0'), out_proj_covar=tensor([7.4891e-05, 1.1891e-04, 1.0228e-04, 9.2899e-05, 9.4121e-05, 9.8456e-05, + 1.1551e-04, 1.2235e-04], device='cuda:0') +2023-03-20 22:35:52,258 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.519e+02 2.324e+02 2.736e+02 3.449e+02 6.925e+02, threshold=5.472e+02, percent-clipped=1.0 +2023-03-20 22:35:52,925 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5811, 1.8636, 2.2156, 1.8207, 1.8121, 1.6117, 1.5828, 1.6339], + device='cuda:0'), covar=tensor([0.0244, 0.0336, 0.0106, 0.0063, 0.0706, 0.0428, 0.0170, 0.0154], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0020, 0.0021, 0.0019, 0.0020, 0.0020, 0.0020, 0.0021], + device='cuda:0'), out_proj_covar=tensor([5.0923e-05, 5.0861e-05, 4.9509e-05, 4.3892e-05, 5.1242e-05, 4.9861e-05, + 4.9510e-05, 5.4527e-05], device='cuda:0') +2023-03-20 22:35:53,208 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.81 vs. limit=2.0 +2023-03-20 22:35:54,306 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-20 22:36:00,959 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1660, 1.1074, 1.3446, 1.5114, 1.3907, 1.4638, 1.5209, 1.3756], + device='cuda:0'), covar=tensor([0.2116, 0.2679, 0.1097, 0.1048, 0.3282, 0.2677, 0.1398, 0.4192], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0049, 0.0035, 0.0033, 0.0040, 0.0039, 0.0045, 0.0039], + device='cuda:0'), out_proj_covar=tensor([1.0038e-04, 1.1667e-04, 8.8886e-05, 8.7669e-05, 1.0169e-04, 1.0185e-04, + 1.1194e-04, 1.0086e-04], device='cuda:0') +2023-03-20 22:36:02,387 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-20 22:36:05,051 INFO [train.py:901] (0/2) Epoch 13, batch 2050, loss[loss=0.1745, simple_loss=0.2407, pruned_loss=0.05414, over 7276.00 frames. ], tot_loss[loss=0.1716, simple_loss=0.244, pruned_loss=0.04963, over 1439106.85 frames. ], batch size: 57, lr: 1.24e-02, grad_scale: 16.0 +2023-03-20 22:36:06,660 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6058, 4.2271, 4.2063, 4.7065, 4.7151, 4.7092, 4.1820, 4.2550], + device='cuda:0'), covar=tensor([0.0813, 0.2432, 0.2167, 0.1152, 0.0672, 0.1188, 0.0637, 0.0865], + device='cuda:0'), in_proj_covar=tensor([0.0182, 0.0293, 0.0239, 0.0227, 0.0170, 0.0288, 0.0158, 0.0205], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:36:14,127 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4528, 3.6669, 3.9891, 3.8430, 3.9036, 3.9810, 4.2330, 3.8669], + device='cuda:0'), covar=tensor([0.0092, 0.0145, 0.0105, 0.0174, 0.0287, 0.0089, 0.0129, 0.0115], + device='cuda:0'), in_proj_covar=tensor([0.0058, 0.0066, 0.0066, 0.0056, 0.0108, 0.0071, 0.0069, 0.0068], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0001, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:36:29,948 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.03 vs. limit=2.0 +2023-03-20 22:36:30,615 INFO [train.py:901] (0/2) Epoch 13, batch 2100, loss[loss=0.1831, simple_loss=0.2628, pruned_loss=0.05172, over 7252.00 frames. ], tot_loss[loss=0.1719, simple_loss=0.2444, pruned_loss=0.04968, over 1440586.91 frames. ], batch size: 55, lr: 1.24e-02, grad_scale: 16.0 +2023-03-20 22:36:36,444 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-36000.pt +2023-03-20 22:36:40,590 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=36001.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 22:36:41,445 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-20 22:36:43,925 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-20 22:36:47,332 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.466e+02 2.503e+02 2.959e+02 3.562e+02 9.049e+02, threshold=5.919e+02, percent-clipped=3.0 +2023-03-20 22:36:55,440 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7861, 3.3746, 3.5029, 3.4099, 2.9474, 3.4156, 3.5511, 3.3420], + device='cuda:0'), covar=tensor([0.0192, 0.0203, 0.0167, 0.0270, 0.0547, 0.0192, 0.0298, 0.0222], + device='cuda:0'), in_proj_covar=tensor([0.0059, 0.0067, 0.0066, 0.0057, 0.0110, 0.0073, 0.0070, 0.0069], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0002, 0.0001, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:36:55,456 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=36030.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:36:59,848 INFO [train.py:901] (0/2) Epoch 13, batch 2150, loss[loss=0.201, simple_loss=0.275, pruned_loss=0.06354, over 6730.00 frames. ], tot_loss[loss=0.1729, simple_loss=0.2452, pruned_loss=0.0503, over 1442345.14 frames. ], batch size: 106, lr: 1.24e-02, grad_scale: 16.0 +2023-03-20 22:37:05,372 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=36049.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 22:37:13,280 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.58 vs. limit=2.0 +2023-03-20 22:37:20,005 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=36078.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:37:25,425 INFO [train.py:901] (0/2) Epoch 13, batch 2200, loss[loss=0.1894, simple_loss=0.2605, pruned_loss=0.05919, over 7343.00 frames. ], tot_loss[loss=0.1741, simple_loss=0.2462, pruned_loss=0.051, over 1441123.65 frames. ], batch size: 63, lr: 1.24e-02, grad_scale: 16.0 +2023-03-20 22:37:27,441 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-20 22:37:27,530 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0768, 4.4060, 4.1693, 4.3805, 4.1279, 4.3280, 4.6906, 4.7616], + device='cuda:0'), covar=tensor([0.0193, 0.0118, 0.0155, 0.0142, 0.0282, 0.0305, 0.0163, 0.0131], + device='cuda:0'), in_proj_covar=tensor([0.0099, 0.0104, 0.0096, 0.0105, 0.0095, 0.0084, 0.0078, 0.0083], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:37:39,138 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.739e+02 2.623e+02 3.188e+02 4.090e+02 7.570e+02, threshold=6.377e+02, percent-clipped=4.0 +2023-03-20 22:37:45,250 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=36127.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:37:51,824 INFO [train.py:901] (0/2) Epoch 13, batch 2250, loss[loss=0.153, simple_loss=0.2286, pruned_loss=0.03871, over 7248.00 frames. ], tot_loss[loss=0.174, simple_loss=0.2463, pruned_loss=0.05084, over 1439746.42 frames. ], batch size: 47, lr: 1.24e-02, grad_scale: 16.0 +2023-03-20 22:37:59,526 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2572, 1.7115, 1.9091, 1.7595, 1.7501, 1.6583, 1.4614, 1.6624], + device='cuda:0'), covar=tensor([0.0429, 0.0178, 0.0190, 0.0069, 0.0339, 0.0401, 0.0237, 0.0195], + device='cuda:0'), in_proj_covar=tensor([0.0021, 0.0020, 0.0022, 0.0020, 0.0021, 0.0021, 0.0020, 0.0021], + device='cuda:0'), out_proj_covar=tensor([5.2517e-05, 4.9972e-05, 5.1046e-05, 4.4556e-05, 5.1563e-05, 5.0596e-05, + 5.0036e-05, 5.4722e-05], device='cuda:0') +2023-03-20 22:38:01,390 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-20 22:38:01,401 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-20 22:38:05,922 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=36167.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:38:12,969 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=36181.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:38:13,841 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-20 22:38:15,406 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=36186.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:38:16,804 INFO [train.py:901] (0/2) Epoch 13, batch 2300, loss[loss=0.1625, simple_loss=0.2348, pruned_loss=0.04514, over 7294.00 frames. ], tot_loss[loss=0.1743, simple_loss=0.2465, pruned_loss=0.05103, over 1438882.87 frames. ], batch size: 66, lr: 1.24e-02, grad_scale: 16.0 +2023-03-20 22:38:30,767 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.358e+02 2.395e+02 2.775e+02 3.184e+02 6.455e+02, threshold=5.551e+02, percent-clipped=1.0 +2023-03-20 22:38:39,297 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.93 vs. limit=5.0 +2023-03-20 22:38:40,902 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=36234.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:38:43,403 INFO [train.py:901] (0/2) Epoch 13, batch 2350, loss[loss=0.1684, simple_loss=0.2466, pruned_loss=0.04514, over 7279.00 frames. ], tot_loss[loss=0.1737, simple_loss=0.2463, pruned_loss=0.05056, over 1441591.64 frames. ], batch size: 70, lr: 1.24e-02, grad_scale: 16.0 +2023-03-20 22:38:45,052 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=36242.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:38:59,497 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-20 22:39:06,130 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-20 22:39:09,117 INFO [train.py:901] (0/2) Epoch 13, batch 2400, loss[loss=0.1738, simple_loss=0.2428, pruned_loss=0.05245, over 7214.00 frames. ], tot_loss[loss=0.1733, simple_loss=0.2463, pruned_loss=0.05017, over 1443613.37 frames. ], batch size: 50, lr: 1.24e-02, grad_scale: 16.0 +2023-03-20 22:39:11,101 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.75 vs. limit=2.0 +2023-03-20 22:39:13,092 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.75 vs. limit=2.0 +2023-03-20 22:39:18,275 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-20 22:39:20,723 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-20 22:39:22,682 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.408e+02 2.253e+02 2.752e+02 3.236e+02 5.663e+02, threshold=5.504e+02, percent-clipped=1.0 +2023-03-20 22:39:34,625 INFO [train.py:901] (0/2) Epoch 13, batch 2450, loss[loss=0.189, simple_loss=0.2573, pruned_loss=0.06029, over 7275.00 frames. ], tot_loss[loss=0.1719, simple_loss=0.2448, pruned_loss=0.04949, over 1443585.63 frames. ], batch size: 64, lr: 1.23e-02, grad_scale: 16.0 +2023-03-20 22:39:35,516 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.58 vs. limit=2.0 +2023-03-20 22:39:45,951 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-20 22:39:46,282 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.99 vs. limit=2.0 +2023-03-20 22:40:00,886 INFO [train.py:901] (0/2) Epoch 13, batch 2500, loss[loss=0.1712, simple_loss=0.249, pruned_loss=0.04669, over 7244.00 frames. ], tot_loss[loss=0.1724, simple_loss=0.2453, pruned_loss=0.04978, over 1444067.22 frames. ], batch size: 89, lr: 1.23e-02, grad_scale: 16.0 +2023-03-20 22:40:10,725 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.95 vs. limit=2.0 +2023-03-20 22:40:11,294 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-20 22:40:14,271 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.589e+02 2.433e+02 2.922e+02 3.588e+02 6.590e+02, threshold=5.845e+02, percent-clipped=3.0 +2023-03-20 22:40:20,385 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=36427.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:40:26,215 INFO [train.py:901] (0/2) Epoch 13, batch 2550, loss[loss=0.1909, simple_loss=0.2589, pruned_loss=0.06142, over 7328.00 frames. ], tot_loss[loss=0.1723, simple_loss=0.245, pruned_loss=0.0498, over 1444605.92 frames. ], batch size: 59, lr: 1.23e-02, grad_scale: 16.0 +2023-03-20 22:40:37,106 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.95 vs. limit=2.0 +2023-03-20 22:40:40,978 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=36467.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:40:45,492 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=36475.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:40:52,338 INFO [train.py:901] (0/2) Epoch 13, batch 2600, loss[loss=0.1631, simple_loss=0.2156, pruned_loss=0.05534, over 5783.00 frames. ], tot_loss[loss=0.172, simple_loss=0.2449, pruned_loss=0.04953, over 1443274.38 frames. ], batch size: 25, lr: 1.23e-02, grad_scale: 16.0 +2023-03-20 22:41:04,999 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.441e+02 2.354e+02 2.833e+02 3.517e+02 6.810e+02, threshold=5.665e+02, percent-clipped=1.0 +2023-03-20 22:41:05,078 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=36515.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:41:15,739 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=36537.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:41:16,682 INFO [train.py:901] (0/2) Epoch 13, batch 2650, loss[loss=0.1632, simple_loss=0.2405, pruned_loss=0.04296, over 7265.00 frames. ], tot_loss[loss=0.1722, simple_loss=0.2451, pruned_loss=0.04965, over 1445364.97 frames. ], batch size: 77, lr: 1.23e-02, grad_scale: 16.0 +2023-03-20 22:41:41,360 INFO [train.py:901] (0/2) Epoch 13, batch 2700, loss[loss=0.2083, simple_loss=0.2654, pruned_loss=0.0756, over 7276.00 frames. ], tot_loss[loss=0.1727, simple_loss=0.2453, pruned_loss=0.05003, over 1443868.62 frames. ], batch size: 70, lr: 1.23e-02, grad_scale: 16.0 +2023-03-20 22:41:49,588 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=36605.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:41:51,106 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=36608.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:41:53,064 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=36612.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:41:54,389 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.842e+02 2.424e+02 2.835e+02 3.228e+02 5.690e+02, threshold=5.669e+02, percent-clipped=1.0 +2023-03-20 22:42:06,221 INFO [train.py:901] (0/2) Epoch 13, batch 2750, loss[loss=0.1638, simple_loss=0.24, pruned_loss=0.04378, over 7126.00 frames. ], tot_loss[loss=0.1723, simple_loss=0.245, pruned_loss=0.04978, over 1443516.42 frames. ], batch size: 41, lr: 1.23e-02, grad_scale: 16.0 +2023-03-20 22:42:06,777 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0530, 4.5197, 4.5933, 4.5859, 4.4699, 4.0272, 4.6685, 4.4550], + device='cuda:0'), covar=tensor([0.0525, 0.0489, 0.0458, 0.0412, 0.0380, 0.0479, 0.0363, 0.0580], + device='cuda:0'), in_proj_covar=tensor([0.0169, 0.0190, 0.0140, 0.0138, 0.0116, 0.0174, 0.0146, 0.0115], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:42:13,432 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3912, 3.8273, 3.4348, 3.7378, 3.5840, 3.5186, 3.6721, 3.9461], + device='cuda:0'), covar=tensor([0.0379, 0.0225, 0.0352, 0.0331, 0.0350, 0.0441, 0.0610, 0.0350], + device='cuda:0'), in_proj_covar=tensor([0.0098, 0.0102, 0.0094, 0.0103, 0.0093, 0.0084, 0.0078, 0.0080], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:42:19,317 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=36666.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:42:20,725 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=36669.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:42:22,996 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=36673.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:42:30,539 INFO [train.py:901] (0/2) Epoch 13, batch 2800, loss[loss=0.1572, simple_loss=0.2206, pruned_loss=0.04691, over 6973.00 frames. ], tot_loss[loss=0.1721, simple_loss=0.2448, pruned_loss=0.04973, over 1444493.86 frames. ], batch size: 35, lr: 1.23e-02, grad_scale: 16.0 +2023-03-20 22:42:35,949 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=36700.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 22:42:43,217 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-13.pt +2023-03-20 22:43:01,672 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-20 22:43:01,730 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-20 22:43:01,756 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-20 22:43:01,774 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-20 22:43:01,959 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-20 22:43:01,974 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-20 22:43:01,985 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-20 22:43:02,084 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-20 22:43:02,162 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-20 22:43:02,178 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-20 22:43:02,194 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-20 22:43:02,727 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-20 22:43:02,732 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-20 22:43:02,740 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-20 22:43:02,889 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-20 22:43:02,892 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-20 22:43:02,950 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-20 22:43:05,229 INFO [train.py:901] (0/2) Epoch 14, batch 0, loss[loss=0.1719, simple_loss=0.2495, pruned_loss=0.04711, over 7336.00 frames. ], tot_loss[loss=0.1719, simple_loss=0.2495, pruned_loss=0.04711, over 7336.00 frames. ], batch size: 75, lr: 1.19e-02, grad_scale: 16.0 +2023-03-20 22:43:05,231 INFO [train.py:926] (0/2) Computing validation loss +2023-03-20 22:43:15,940 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4515, 1.6093, 1.7060, 1.6754, 1.7977, 1.8867, 1.7259, 1.7202], + device='cuda:0'), covar=tensor([0.1266, 0.1809, 0.0718, 0.0644, 0.1555, 0.1922, 0.0920, 0.2553], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0045, 0.0033, 0.0030, 0.0037, 0.0037, 0.0043, 0.0036], + device='cuda:0'), out_proj_covar=tensor([9.5259e-05, 1.0952e-04, 8.3993e-05, 8.3061e-05, 9.5528e-05, 9.7744e-05, + 1.0739e-04, 9.5338e-05], device='cuda:0') +2023-03-20 22:43:16,638 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1430, 3.6775, 3.6432, 3.7908, 3.7226, 3.7228, 3.7900, 3.4070], + device='cuda:0'), covar=tensor([0.0079, 0.0190, 0.0153, 0.0139, 0.0344, 0.0126, 0.0175, 0.0205], + device='cuda:0'), in_proj_covar=tensor([0.0062, 0.0070, 0.0071, 0.0060, 0.0119, 0.0079, 0.0074, 0.0074], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0001, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:43:16,901 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5179, 2.7896, 1.7137, 3.5335, 2.0936, 2.6416, 1.4438, 1.7461], + device='cuda:0'), covar=tensor([0.0193, 0.0513, 0.1758, 0.0435, 0.0303, 0.0534, 0.2713, 0.1387], + device='cuda:0'), in_proj_covar=tensor([0.0215, 0.0226, 0.0299, 0.0218, 0.0238, 0.0223, 0.0262, 0.0280], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 22:43:19,090 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.2196, 5.4329, 5.3191, 5.4381, 4.8720, 4.9738, 5.4485, 4.9932], + device='cuda:0'), covar=tensor([0.0235, 0.0356, 0.0468, 0.0342, 0.0521, 0.0291, 0.0276, 0.0646], + device='cuda:0'), in_proj_covar=tensor([0.0170, 0.0191, 0.0140, 0.0137, 0.0118, 0.0175, 0.0147, 0.0116], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:43:22,240 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2864, 3.1698, 3.4103, 3.1700, 3.4823, 3.1259, 2.6799, 3.2570], + device='cuda:0'), covar=tensor([0.1297, 0.0576, 0.1373, 0.1708, 0.0695, 0.1167, 0.2158, 0.1488], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0040, 0.0035, 0.0037, 0.0034, 0.0032, 0.0049, 0.0037], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:43:31,789 INFO [train.py:935] (0/2) Epoch 14, validation: loss=0.1698, simple_loss=0.2569, pruned_loss=0.04131, over 1622729.00 frames. +2023-03-20 22:43:31,790 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-20 22:43:32,792 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.549e+02 2.144e+02 2.550e+02 3.051e+02 8.229e+02, threshold=5.101e+02, percent-clipped=3.0 +2023-03-20 22:43:38,820 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-20 22:43:51,001 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-20 22:43:56,599 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=36761.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 22:43:57,422 INFO [train.py:901] (0/2) Epoch 14, batch 50, loss[loss=0.1731, simple_loss=0.2437, pruned_loss=0.05127, over 7321.00 frames. ], tot_loss[loss=0.1688, simple_loss=0.2421, pruned_loss=0.04774, over 328497.69 frames. ], batch size: 61, lr: 1.19e-02, grad_scale: 16.0 +2023-03-20 22:43:57,476 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-20 22:44:00,027 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-20 22:44:02,580 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-20 22:44:02,710 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0925, 0.9877, 1.3162, 1.3022, 1.2646, 1.3786, 0.8893, 1.2613], + device='cuda:0'), covar=tensor([0.1053, 0.2257, 0.0740, 0.0717, 0.0873, 0.1047, 0.0718, 0.2838], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0044, 0.0033, 0.0030, 0.0038, 0.0037, 0.0043, 0.0036], + device='cuda:0'), out_proj_covar=tensor([9.5968e-05, 1.0938e-04, 8.5407e-05, 8.3013e-05, 9.6099e-05, 9.7135e-05, + 1.0799e-04, 9.5253e-05], device='cuda:0') +2023-03-20 22:44:05,041 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.85 vs. limit=5.0 +2023-03-20 22:44:19,154 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=36805.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 22:44:21,246 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-20 22:44:21,727 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-20 22:44:21,959 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-20 22:44:23,688 INFO [train.py:901] (0/2) Epoch 14, batch 100, loss[loss=0.1832, simple_loss=0.2547, pruned_loss=0.05586, over 7277.00 frames. ], tot_loss[loss=0.1691, simple_loss=0.2422, pruned_loss=0.04797, over 572842.39 frames. ], batch size: 47, lr: 1.19e-02, grad_scale: 16.0 +2023-03-20 22:44:24,633 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.278e+02 2.289e+02 2.637e+02 3.349e+02 8.885e+02, threshold=5.275e+02, percent-clipped=4.0 +2023-03-20 22:44:30,020 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.95 vs. limit=2.0 +2023-03-20 22:44:36,383 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=36837.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:44:49,241 INFO [train.py:901] (0/2) Epoch 14, batch 150, loss[loss=0.15, simple_loss=0.2193, pruned_loss=0.0403, over 7137.00 frames. ], tot_loss[loss=0.1682, simple_loss=0.2415, pruned_loss=0.04744, over 764609.57 frames. ], batch size: 41, lr: 1.19e-02, grad_scale: 16.0 +2023-03-20 22:44:50,866 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=36866.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 22:45:00,635 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=36885.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:45:14,409 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=36911.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:45:15,260 INFO [train.py:901] (0/2) Epoch 14, batch 200, loss[loss=0.2085, simple_loss=0.2846, pruned_loss=0.06625, over 6608.00 frames. ], tot_loss[loss=0.1697, simple_loss=0.2427, pruned_loss=0.04838, over 911820.76 frames. ], batch size: 106, lr: 1.19e-02, grad_scale: 16.0 +2023-03-20 22:45:16,279 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.419e+02 2.238e+02 2.665e+02 3.334e+02 5.724e+02, threshold=5.331e+02, percent-clipped=5.0 +2023-03-20 22:45:21,438 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-20 22:45:26,017 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-20 22:45:32,054 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-20 22:45:38,681 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=36958.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:45:40,117 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=36961.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:45:41,010 INFO [train.py:901] (0/2) Epoch 14, batch 250, loss[loss=0.1649, simple_loss=0.248, pruned_loss=0.04096, over 7348.00 frames. ], tot_loss[loss=0.1696, simple_loss=0.2428, pruned_loss=0.04821, over 1030383.99 frames. ], batch size: 63, lr: 1.19e-02, grad_scale: 16.0 +2023-03-20 22:45:41,606 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=36964.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:45:43,620 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=36968.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:45:44,537 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-20 22:45:45,674 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=36972.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:45:51,773 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0225, 4.3155, 4.0642, 4.1703, 3.9651, 4.1983, 4.4733, 4.5940], + device='cuda:0'), covar=tensor([0.0161, 0.0097, 0.0149, 0.0144, 0.0237, 0.0150, 0.0217, 0.0144], + device='cuda:0'), in_proj_covar=tensor([0.0098, 0.0102, 0.0093, 0.0103, 0.0093, 0.0082, 0.0078, 0.0079], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:45:59,944 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.99 vs. limit=5.0 +2023-03-20 22:46:03,281 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=37004.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:46:05,739 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-20 22:46:07,723 INFO [train.py:901] (0/2) Epoch 14, batch 300, loss[loss=0.1841, simple_loss=0.2521, pruned_loss=0.05806, over 7328.00 frames. ], tot_loss[loss=0.1694, simple_loss=0.2423, pruned_loss=0.04824, over 1120801.79 frames. ], batch size: 75, lr: 1.18e-02, grad_scale: 16.0 +2023-03-20 22:46:08,350 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3955, 4.3328, 3.8682, 3.6577, 3.8985, 2.6372, 1.9742, 4.3657], + device='cuda:0'), covar=tensor([0.0027, 0.0031, 0.0074, 0.0063, 0.0062, 0.0357, 0.0501, 0.0041], + device='cuda:0'), in_proj_covar=tensor([0.0072, 0.0064, 0.0085, 0.0070, 0.0089, 0.0108, 0.0111, 0.0075], + device='cuda:0'), out_proj_covar=tensor([9.3109e-05, 9.3400e-05, 1.1221e-04, 9.9866e-05, 1.1883e-04, 1.4537e-04, + 1.4879e-04, 1.0007e-04], device='cuda:0') +2023-03-20 22:46:08,720 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.524e+02 2.157e+02 2.665e+02 3.235e+02 4.980e+02, threshold=5.329e+02, percent-clipped=0.0 +2023-03-20 22:46:10,896 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=37019.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:46:14,780 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-20 22:46:19,344 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.0513, 5.5089, 5.5487, 5.4644, 5.1991, 5.1035, 5.6128, 5.2766], + device='cuda:0'), covar=tensor([0.0355, 0.0302, 0.0324, 0.0410, 0.0322, 0.0252, 0.0256, 0.0451], + device='cuda:0'), in_proj_covar=tensor([0.0172, 0.0192, 0.0140, 0.0139, 0.0119, 0.0178, 0.0149, 0.0118], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:46:29,240 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=37056.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 22:46:32,661 INFO [train.py:901] (0/2) Epoch 14, batch 350, loss[loss=0.1707, simple_loss=0.2455, pruned_loss=0.04794, over 7239.00 frames. ], tot_loss[loss=0.1698, simple_loss=0.2423, pruned_loss=0.04864, over 1190345.77 frames. ], batch size: 55, lr: 1.18e-02, grad_scale: 16.0 +2023-03-20 22:46:34,426 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=37065.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:46:48,932 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-20 22:46:58,916 INFO [train.py:901] (0/2) Epoch 14, batch 400, loss[loss=0.1725, simple_loss=0.2479, pruned_loss=0.04856, over 7270.00 frames. ], tot_loss[loss=0.1702, simple_loss=0.2429, pruned_loss=0.04876, over 1247756.59 frames. ], batch size: 77, lr: 1.18e-02, grad_scale: 16.0 +2023-03-20 22:46:59,939 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.519e+02 2.445e+02 2.867e+02 4.022e+02 1.109e+03, threshold=5.735e+02, percent-clipped=7.0 +2023-03-20 22:47:19,914 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9137, 2.0934, 2.0265, 3.2196, 1.4162, 2.9357, 1.4155, 2.8391], + device='cuda:0'), covar=tensor([0.0065, 0.0912, 0.1493, 0.0049, 0.3813, 0.0093, 0.0977, 0.0219], + device='cuda:0'), in_proj_covar=tensor([0.0153, 0.0262, 0.0301, 0.0146, 0.0285, 0.0160, 0.0266, 0.0213], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 22:47:23,791 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=37161.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 22:47:24,743 INFO [train.py:901] (0/2) Epoch 14, batch 450, loss[loss=0.1528, simple_loss=0.2299, pruned_loss=0.03785, over 7211.00 frames. ], tot_loss[loss=0.1694, simple_loss=0.2423, pruned_loss=0.04819, over 1289951.88 frames. ], batch size: 50, lr: 1.18e-02, grad_scale: 16.0 +2023-03-20 22:47:30,316 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-20 22:47:30,329 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-20 22:47:50,861 INFO [train.py:901] (0/2) Epoch 14, batch 500, loss[loss=0.1689, simple_loss=0.2345, pruned_loss=0.0516, over 7259.00 frames. ], tot_loss[loss=0.1695, simple_loss=0.2425, pruned_loss=0.04823, over 1325181.31 frames. ], batch size: 52, lr: 1.18e-02, grad_scale: 16.0 +2023-03-20 22:47:51,846 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.393e+02 2.247e+02 2.615e+02 3.199e+02 8.743e+02, threshold=5.229e+02, percent-clipped=2.0 +2023-03-20 22:48:02,390 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-20 22:48:03,366 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-20 22:48:03,784 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.08 vs. limit=2.0 +2023-03-20 22:48:04,349 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-20 22:48:06,854 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-20 22:48:11,381 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-20 22:48:16,044 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=37261.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:48:16,925 INFO [train.py:901] (0/2) Epoch 14, batch 550, loss[loss=0.1672, simple_loss=0.2454, pruned_loss=0.04454, over 7361.00 frames. ], tot_loss[loss=0.1698, simple_loss=0.2427, pruned_loss=0.04844, over 1349554.96 frames. ], batch size: 63, lr: 1.18e-02, grad_scale: 16.0 +2023-03-20 22:48:17,530 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=37264.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:48:18,976 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=37267.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:48:19,462 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=37268.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:48:19,486 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=37268.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:48:20,992 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9154, 2.5473, 3.0880, 2.8669, 3.1010, 2.8739, 2.4167, 3.2042], + device='cuda:0'), covar=tensor([0.1629, 0.0556, 0.1186, 0.2083, 0.0722, 0.1014, 0.2501, 0.0849], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0040, 0.0036, 0.0037, 0.0033, 0.0031, 0.0048, 0.0036], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:48:22,435 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3670, 3.8331, 3.9677, 3.9152, 3.7545, 3.9070, 4.0551, 3.6310], + device='cuda:0'), covar=tensor([0.0078, 0.0144, 0.0089, 0.0159, 0.0307, 0.0105, 0.0141, 0.0178], + device='cuda:0'), in_proj_covar=tensor([0.0061, 0.0068, 0.0069, 0.0059, 0.0116, 0.0076, 0.0073, 0.0073], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0001, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:48:22,864 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-20 22:48:30,366 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-20 22:48:33,777 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-20 22:48:35,174 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.84 vs. limit=2.0 +2023-03-20 22:48:39,722 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=37309.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:48:41,181 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-20 22:48:41,225 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=37312.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:48:41,663 INFO [train.py:901] (0/2) Epoch 14, batch 600, loss[loss=0.1712, simple_loss=0.2383, pruned_loss=0.05203, over 7308.00 frames. ], tot_loss[loss=0.1702, simple_loss=0.243, pruned_loss=0.0487, over 1371542.51 frames. ], batch size: 49, lr: 1.18e-02, grad_scale: 16.0 +2023-03-20 22:48:42,269 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=37314.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:48:42,681 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.746e+02 2.422e+02 2.800e+02 3.690e+02 8.358e+02, threshold=5.600e+02, percent-clipped=9.0 +2023-03-20 22:48:43,297 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=37316.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:48:47,973 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6058, 3.9615, 3.7086, 4.0009, 3.5405, 4.0035, 4.1448, 4.2530], + device='cuda:0'), covar=tensor([0.0246, 0.0183, 0.0226, 0.0167, 0.0422, 0.0197, 0.0246, 0.0187], + device='cuda:0'), in_proj_covar=tensor([0.0101, 0.0103, 0.0095, 0.0107, 0.0096, 0.0085, 0.0081, 0.0080], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:48:50,477 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=37329.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:48:57,798 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-20 22:49:04,333 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=37356.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 22:49:06,312 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=37360.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:49:06,749 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-20 22:49:07,727 INFO [train.py:901] (0/2) Epoch 14, batch 650, loss[loss=0.1704, simple_loss=0.2447, pruned_loss=0.04805, over 7264.00 frames. ], tot_loss[loss=0.1701, simple_loss=0.2428, pruned_loss=0.04867, over 1385913.44 frames. ], batch size: 52, lr: 1.18e-02, grad_scale: 16.0 +2023-03-20 22:49:08,337 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=37364.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:49:23,718 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-20 22:49:28,484 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=37404.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 22:49:29,518 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6218, 4.1301, 3.9840, 4.5927, 4.5092, 4.5640, 3.9111, 4.0396], + device='cuda:0'), covar=tensor([0.0623, 0.2644, 0.2106, 0.1159, 0.0708, 0.1267, 0.0747, 0.1119], + device='cuda:0'), in_proj_covar=tensor([0.0177, 0.0297, 0.0238, 0.0226, 0.0173, 0.0288, 0.0160, 0.0205], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:49:31,725 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2738, 2.8736, 2.6738, 3.0556, 2.5255, 2.3891, 3.1650, 2.3963], + device='cuda:0'), covar=tensor([0.0286, 0.0246, 0.0331, 0.0287, 0.0358, 0.0551, 0.0317, 0.0884], + device='cuda:0'), in_proj_covar=tensor([0.0288, 0.0293, 0.0245, 0.0306, 0.0297, 0.0297, 0.0295, 0.0289], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 22:49:32,079 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-20 22:49:33,571 INFO [train.py:901] (0/2) Epoch 14, batch 700, loss[loss=0.1735, simple_loss=0.2515, pruned_loss=0.0478, over 7289.00 frames. ], tot_loss[loss=0.1706, simple_loss=0.2431, pruned_loss=0.04903, over 1396735.61 frames. ], batch size: 68, lr: 1.18e-02, grad_scale: 16.0 +2023-03-20 22:49:34,579 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.428e+02 2.290e+02 2.741e+02 3.295e+02 7.204e+02, threshold=5.481e+02, percent-clipped=3.0 +2023-03-20 22:49:39,378 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.00 vs. limit=2.0 +2023-03-20 22:49:39,708 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=37425.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:49:46,768 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6480, 3.1929, 3.0248, 3.4326, 2.8255, 2.7389, 3.5986, 2.6708], + device='cuda:0'), covar=tensor([0.0251, 0.0200, 0.0382, 0.0281, 0.0479, 0.0636, 0.0255, 0.1199], + device='cuda:0'), in_proj_covar=tensor([0.0287, 0.0290, 0.0244, 0.0305, 0.0297, 0.0296, 0.0294, 0.0289], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 22:49:54,641 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5790, 4.9893, 5.0770, 4.9917, 4.8205, 4.5579, 5.0868, 4.8386], + device='cuda:0'), covar=tensor([0.0345, 0.0347, 0.0340, 0.0447, 0.0342, 0.0272, 0.0305, 0.0460], + device='cuda:0'), in_proj_covar=tensor([0.0168, 0.0190, 0.0136, 0.0136, 0.0117, 0.0173, 0.0148, 0.0115], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:49:55,096 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-20 22:49:55,587 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-20 22:49:58,155 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=37461.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 22:49:59,027 INFO [train.py:901] (0/2) Epoch 14, batch 750, loss[loss=0.186, simple_loss=0.2547, pruned_loss=0.05868, over 7328.00 frames. ], tot_loss[loss=0.171, simple_loss=0.2437, pruned_loss=0.04913, over 1409195.17 frames. ], batch size: 59, lr: 1.18e-02, grad_scale: 16.0 +2023-03-20 22:50:00,677 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7864, 4.1009, 3.7782, 4.1956, 3.7878, 4.0480, 4.3571, 4.4807], + device='cuda:0'), covar=tensor([0.0250, 0.0155, 0.0189, 0.0153, 0.0346, 0.0228, 0.0217, 0.0145], + device='cuda:0'), in_proj_covar=tensor([0.0100, 0.0102, 0.0095, 0.0105, 0.0095, 0.0085, 0.0080, 0.0078], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:50:09,050 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-20 22:50:14,017 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-20 22:50:21,200 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-20 22:50:22,172 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-20 22:50:22,736 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=37509.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 22:50:24,697 INFO [train.py:901] (0/2) Epoch 14, batch 800, loss[loss=0.1414, simple_loss=0.2269, pruned_loss=0.02795, over 7350.00 frames. ], tot_loss[loss=0.1702, simple_loss=0.2429, pruned_loss=0.0487, over 1418167.25 frames. ], batch size: 44, lr: 1.18e-02, grad_scale: 16.0 +2023-03-20 22:50:25,670 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.705e+02 2.326e+02 2.652e+02 3.286e+02 4.306e+02, threshold=5.303e+02, percent-clipped=0.0 +2023-03-20 22:50:33,736 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-20 22:50:50,152 INFO [train.py:901] (0/2) Epoch 14, batch 850, loss[loss=0.1746, simple_loss=0.2468, pruned_loss=0.05113, over 7322.00 frames. ], tot_loss[loss=0.1702, simple_loss=0.2432, pruned_loss=0.04865, over 1424809.02 frames. ], batch size: 75, lr: 1.18e-02, grad_scale: 16.0 +2023-03-20 22:50:52,151 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-20 22:50:52,155 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-20 22:50:52,245 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=37567.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:50:57,231 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-20 22:51:00,902 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-20 22:51:14,065 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6610, 3.1011, 3.1397, 3.3450, 2.6434, 2.7078, 3.4595, 2.7256], + device='cuda:0'), covar=tensor([0.0182, 0.0217, 0.0279, 0.0239, 0.0324, 0.0512, 0.0264, 0.0847], + device='cuda:0'), in_proj_covar=tensor([0.0288, 0.0290, 0.0244, 0.0307, 0.0296, 0.0299, 0.0293, 0.0291], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 22:51:16,871 INFO [train.py:901] (0/2) Epoch 14, batch 900, loss[loss=0.1789, simple_loss=0.2516, pruned_loss=0.05312, over 7282.00 frames. ], tot_loss[loss=0.1702, simple_loss=0.2434, pruned_loss=0.04854, over 1427769.56 frames. ], batch size: 57, lr: 1.17e-02, grad_scale: 16.0 +2023-03-20 22:51:17,508 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=37614.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:51:17,857 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.554e+02 2.227e+02 2.893e+02 3.538e+02 7.539e+02, threshold=5.787e+02, percent-clipped=4.0 +2023-03-20 22:51:17,929 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=37615.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:51:22,547 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=37624.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:51:24,797 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-20 22:51:34,187 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.54 vs. limit=5.0 +2023-03-20 22:51:37,923 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-20 22:51:40,454 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=37660.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:51:41,340 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=37662.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:51:41,786 INFO [train.py:901] (0/2) Epoch 14, batch 950, loss[loss=0.1836, simple_loss=0.2522, pruned_loss=0.05747, over 7259.00 frames. ], tot_loss[loss=0.1703, simple_loss=0.2434, pruned_loss=0.04857, over 1431974.75 frames. ], batch size: 52, lr: 1.17e-02, grad_scale: 16.0 +2023-03-20 22:52:01,343 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0186, 4.3135, 3.9497, 4.3179, 4.0026, 4.3468, 4.5764, 4.6565], + device='cuda:0'), covar=tensor([0.0198, 0.0144, 0.0219, 0.0159, 0.0272, 0.0157, 0.0233, 0.0170], + device='cuda:0'), in_proj_covar=tensor([0.0097, 0.0099, 0.0093, 0.0102, 0.0093, 0.0081, 0.0077, 0.0077], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:52:03,271 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-20 22:52:05,810 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=37708.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:52:08,355 INFO [train.py:901] (0/2) Epoch 14, batch 1000, loss[loss=0.1637, simple_loss=0.25, pruned_loss=0.0387, over 7330.00 frames. ], tot_loss[loss=0.1688, simple_loss=0.2422, pruned_loss=0.04771, over 1434150.17 frames. ], batch size: 61, lr: 1.17e-02, grad_scale: 16.0 +2023-03-20 22:52:09,824 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.461e+02 2.281e+02 2.630e+02 3.242e+02 8.955e+02, threshold=5.260e+02, percent-clipped=1.0 +2023-03-20 22:52:11,895 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=37720.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:52:20,703 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.05 vs. limit=2.0 +2023-03-20 22:52:22,908 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-20 22:52:34,161 INFO [train.py:901] (0/2) Epoch 14, batch 1050, loss[loss=0.1766, simple_loss=0.2482, pruned_loss=0.05247, over 7284.00 frames. ], tot_loss[loss=0.1677, simple_loss=0.241, pruned_loss=0.04717, over 1435251.22 frames. ], batch size: 70, lr: 1.17e-02, grad_scale: 16.0 +2023-03-20 22:52:46,470 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-20 22:52:51,531 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-20 22:53:00,401 INFO [train.py:901] (0/2) Epoch 14, batch 1100, loss[loss=0.1578, simple_loss=0.2369, pruned_loss=0.03929, over 7327.00 frames. ], tot_loss[loss=0.1677, simple_loss=0.2409, pruned_loss=0.04725, over 1436872.76 frames. ], batch size: 75, lr: 1.17e-02, grad_scale: 16.0 +2023-03-20 22:53:01,902 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.605e+02 2.212e+02 2.615e+02 3.217e+02 6.533e+02, threshold=5.230e+02, percent-clipped=5.0 +2023-03-20 22:53:03,571 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5148, 2.2459, 2.1977, 2.0782, 1.8707, 1.7005, 1.8379, 1.7688], + device='cuda:0'), covar=tensor([0.0334, 0.0200, 0.0129, 0.0060, 0.0437, 0.0309, 0.0146, 0.0325], + device='cuda:0'), in_proj_covar=tensor([0.0020, 0.0021, 0.0021, 0.0019, 0.0020, 0.0019, 0.0020, 0.0022], + device='cuda:0'), out_proj_covar=tensor([5.1190e-05, 5.1910e-05, 4.8947e-05, 4.4742e-05, 5.0445e-05, 4.8643e-05, + 4.9916e-05, 5.6368e-05], device='cuda:0') +2023-03-20 22:53:07,009 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9754, 2.1480, 2.0540, 3.6055, 1.4028, 3.1967, 1.3751, 2.8762], + device='cuda:0'), covar=tensor([0.0081, 0.1032, 0.1666, 0.0075, 0.3774, 0.0073, 0.0963, 0.0249], + device='cuda:0'), in_proj_covar=tensor([0.0157, 0.0259, 0.0298, 0.0150, 0.0286, 0.0160, 0.0269, 0.0212], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 22:53:12,468 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9702, 3.6828, 3.7223, 3.7429, 3.6214, 3.6348, 3.9069, 3.4227], + device='cuda:0'), covar=tensor([0.0121, 0.0138, 0.0120, 0.0131, 0.0304, 0.0100, 0.0142, 0.0173], + device='cuda:0'), in_proj_covar=tensor([0.0062, 0.0068, 0.0069, 0.0058, 0.0115, 0.0076, 0.0072, 0.0073], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0001, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:53:18,983 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.15 vs. limit=5.0 +2023-03-20 22:53:20,071 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-20 22:53:20,595 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 22:53:26,186 INFO [train.py:901] (0/2) Epoch 14, batch 1150, loss[loss=0.1863, simple_loss=0.251, pruned_loss=0.06081, over 7224.00 frames. ], tot_loss[loss=0.1686, simple_loss=0.2416, pruned_loss=0.04775, over 1437917.19 frames. ], batch size: 45, lr: 1.17e-02, grad_scale: 16.0 +2023-03-20 22:53:32,757 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-20 22:53:33,260 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-20 22:53:35,850 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0451, 0.9666, 1.3267, 1.3258, 1.4004, 1.3960, 1.0801, 1.1662], + device='cuda:0'), covar=tensor([0.2336, 0.2637, 0.0658, 0.1210, 0.1396, 0.1253, 0.1796, 0.2587], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0046, 0.0034, 0.0031, 0.0039, 0.0036, 0.0043, 0.0039], + device='cuda:0'), out_proj_covar=tensor([9.9945e-05, 1.1347e-04, 8.8884e-05, 8.7606e-05, 1.0030e-04, 9.8403e-05, + 1.1007e-04, 1.0216e-04], device='cuda:0') +2023-03-20 22:53:42,704 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.81 vs. limit=2.0 +2023-03-20 22:53:50,656 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.11 vs. limit=2.0 +2023-03-20 22:53:51,813 INFO [train.py:901] (0/2) Epoch 14, batch 1200, loss[loss=0.1715, simple_loss=0.2507, pruned_loss=0.04617, over 7379.00 frames. ], tot_loss[loss=0.1694, simple_loss=0.2422, pruned_loss=0.04825, over 1438851.73 frames. ], batch size: 65, lr: 1.17e-02, grad_scale: 16.0 +2023-03-20 22:53:53,340 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.643e+02 2.300e+02 2.689e+02 3.229e+02 6.828e+02, threshold=5.378e+02, percent-clipped=3.0 +2023-03-20 22:53:56,460 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0713, 3.5699, 4.0307, 4.0357, 4.0917, 3.9340, 3.9902, 3.8260], + device='cuda:0'), covar=tensor([0.0027, 0.0081, 0.0031, 0.0032, 0.0029, 0.0034, 0.0033, 0.0040], + device='cuda:0'), in_proj_covar=tensor([0.0032, 0.0044, 0.0039, 0.0037, 0.0038, 0.0039, 0.0042, 0.0047], + device='cuda:0'), out_proj_covar=tensor([7.5100e-05, 1.1927e-04, 1.0156e-04, 8.7275e-05, 9.1742e-05, 9.5446e-05, + 1.1107e-04, 1.1660e-04], device='cuda:0') +2023-03-20 22:53:57,410 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=37924.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:54:05,091 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-20 22:54:18,271 INFO [train.py:901] (0/2) Epoch 14, batch 1250, loss[loss=0.1918, simple_loss=0.2608, pruned_loss=0.06142, over 7217.00 frames. ], tot_loss[loss=0.1682, simple_loss=0.2412, pruned_loss=0.04764, over 1438524.75 frames. ], batch size: 93, lr: 1.17e-02, grad_scale: 16.0 +2023-03-20 22:54:22,838 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=37972.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:54:26,495 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=37979.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:54:30,391 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-20 22:54:34,408 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-20 22:54:35,414 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-20 22:54:36,056 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3558, 2.3840, 2.0060, 2.5032, 2.6397, 2.5442, 2.5445, 2.5729], + device='cuda:0'), covar=tensor([0.1554, 0.0552, 0.2863, 0.0482, 0.0069, 0.0055, 0.0161, 0.0177], + device='cuda:0'), in_proj_covar=tensor([0.0281, 0.0222, 0.0273, 0.0258, 0.0129, 0.0123, 0.0152, 0.0176], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0001, 0.0001, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:54:43,643 INFO [train.py:901] (0/2) Epoch 14, batch 1300, loss[loss=0.1586, simple_loss=0.2398, pruned_loss=0.0387, over 7286.00 frames. ], tot_loss[loss=0.1688, simple_loss=0.2417, pruned_loss=0.048, over 1440756.52 frames. ], batch size: 68, lr: 1.17e-02, grad_scale: 16.0 +2023-03-20 22:54:45,116 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.489e+02 2.177e+02 2.872e+02 3.523e+02 8.377e+02, threshold=5.744e+02, percent-clipped=2.0 +2023-03-20 22:54:47,279 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=38020.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:54:58,002 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=38040.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:54:58,796 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-20 22:55:01,319 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-20 22:55:05,259 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-20 22:55:09,672 INFO [train.py:901] (0/2) Epoch 14, batch 1350, loss[loss=0.1629, simple_loss=0.2285, pruned_loss=0.04869, over 7259.00 frames. ], tot_loss[loss=0.1683, simple_loss=0.2412, pruned_loss=0.04766, over 1441696.75 frames. ], batch size: 47, lr: 1.17e-02, grad_scale: 16.0 +2023-03-20 22:55:12,217 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=38068.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:55:15,573 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-20 22:55:22,969 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.78 vs. limit=2.0 +2023-03-20 22:55:35,272 INFO [train.py:901] (0/2) Epoch 14, batch 1400, loss[loss=0.1647, simple_loss=0.2382, pruned_loss=0.04556, over 7262.00 frames. ], tot_loss[loss=0.1685, simple_loss=0.2411, pruned_loss=0.0479, over 1440984.81 frames. ], batch size: 64, lr: 1.17e-02, grad_scale: 16.0 +2023-03-20 22:55:36,731 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.635e+02 2.367e+02 2.853e+02 3.401e+02 7.956e+02, threshold=5.705e+02, percent-clipped=5.0 +2023-03-20 22:55:49,482 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-20 22:55:56,858 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1531, 4.6067, 4.6464, 4.6181, 4.5845, 4.2369, 4.7077, 4.5739], + device='cuda:0'), covar=tensor([0.0431, 0.0409, 0.0384, 0.0421, 0.0281, 0.0340, 0.0328, 0.0442], + device='cuda:0'), in_proj_covar=tensor([0.0170, 0.0194, 0.0141, 0.0139, 0.0120, 0.0176, 0.0151, 0.0118], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:56:00,690 INFO [train.py:901] (0/2) Epoch 14, batch 1450, loss[loss=0.1861, simple_loss=0.2549, pruned_loss=0.05869, over 7322.00 frames. ], tot_loss[loss=0.1692, simple_loss=0.2422, pruned_loss=0.04816, over 1442881.48 frames. ], batch size: 75, lr: 1.17e-02, grad_scale: 16.0 +2023-03-20 22:56:13,082 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-20 22:56:23,269 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1609, 1.0352, 1.6671, 1.3857, 1.3579, 1.1420, 1.2723, 1.2924], + device='cuda:0'), covar=tensor([0.1978, 0.2720, 0.0599, 0.1377, 0.1180, 0.4117, 0.1420, 0.4000], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0048, 0.0035, 0.0032, 0.0039, 0.0037, 0.0046, 0.0041], + device='cuda:0'), out_proj_covar=tensor([1.0182e-04, 1.1743e-04, 9.2132e-05, 9.1131e-05, 1.0268e-04, 1.0162e-04, + 1.1592e-04, 1.0670e-04], device='cuda:0') +2023-03-20 22:56:26,654 INFO [train.py:901] (0/2) Epoch 14, batch 1500, loss[loss=0.1476, simple_loss=0.2263, pruned_loss=0.03443, over 7312.00 frames. ], tot_loss[loss=0.1684, simple_loss=0.2418, pruned_loss=0.04755, over 1443305.86 frames. ], batch size: 59, lr: 1.17e-02, grad_scale: 16.0 +2023-03-20 22:56:28,108 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.257e+02 2.452e+02 2.832e+02 3.488e+02 6.170e+02, threshold=5.665e+02, percent-clipped=2.0 +2023-03-20 22:56:28,895 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.84 vs. limit=2.0 +2023-03-20 22:56:30,083 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-20 22:56:33,160 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.77 vs. limit=2.0 +2023-03-20 22:56:52,220 INFO [train.py:901] (0/2) Epoch 14, batch 1550, loss[loss=0.165, simple_loss=0.2389, pruned_loss=0.04551, over 7304.00 frames. ], tot_loss[loss=0.1681, simple_loss=0.241, pruned_loss=0.0476, over 1441391.84 frames. ], batch size: 83, lr: 1.17e-02, grad_scale: 16.0 +2023-03-20 22:56:54,268 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-20 22:56:54,396 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2898, 1.1073, 1.6798, 1.3974, 1.4804, 1.3062, 1.4473, 1.2506], + device='cuda:0'), covar=tensor([0.1126, 0.3712, 0.0689, 0.1609, 0.0970, 0.3996, 0.1744, 0.3336], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0047, 0.0035, 0.0033, 0.0040, 0.0036, 0.0045, 0.0040], + device='cuda:0'), out_proj_covar=tensor([1.0161e-04, 1.1667e-04, 9.1501e-05, 9.1328e-05, 1.0325e-04, 1.0125e-04, + 1.1546e-04, 1.0617e-04], device='cuda:0') +2023-03-20 22:57:08,527 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1502, 1.5274, 1.3747, 1.3983, 1.2114, 1.1766, 1.1556, 1.0615], + device='cuda:0'), covar=tensor([0.0140, 0.0103, 0.0140, 0.0112, 0.0292, 0.0123, 0.0237, 0.0138], + device='cuda:0'), in_proj_covar=tensor([0.0022, 0.0021, 0.0020, 0.0019, 0.0023, 0.0020, 0.0021, 0.0027], + device='cuda:0'), out_proj_covar=tensor([2.6803e-05, 2.3768e-05, 2.4296e-05, 2.1536e-05, 2.7687e-05, 2.2755e-05, + 2.4441e-05, 3.2939e-05], device='cuda:0') +2023-03-20 22:57:18,489 INFO [train.py:901] (0/2) Epoch 14, batch 1600, loss[loss=0.1618, simple_loss=0.2322, pruned_loss=0.04567, over 7324.00 frames. ], tot_loss[loss=0.1678, simple_loss=0.2408, pruned_loss=0.04737, over 1442909.59 frames. ], batch size: 75, lr: 1.16e-02, grad_scale: 16.0 +2023-03-20 22:57:19,971 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.415e+02 2.154e+02 2.623e+02 3.275e+02 5.102e+02, threshold=5.247e+02, percent-clipped=0.0 +2023-03-20 22:57:26,051 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-20 22:57:27,026 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-20 22:57:28,196 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8108, 2.2466, 1.7339, 2.7535, 2.3860, 2.8155, 2.0270, 1.9951], + device='cuda:0'), covar=tensor([0.1684, 0.0721, 0.3048, 0.0488, 0.0097, 0.0073, 0.0132, 0.0178], + device='cuda:0'), in_proj_covar=tensor([0.0281, 0.0223, 0.0271, 0.0257, 0.0131, 0.0125, 0.0152, 0.0177], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0001, 0.0001, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 22:57:29,599 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=38335.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:57:30,044 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-20 22:57:40,021 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-20 22:57:42,167 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=38360.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:57:43,451 INFO [train.py:901] (0/2) Epoch 14, batch 1650, loss[loss=0.1616, simple_loss=0.2355, pruned_loss=0.04387, over 7326.00 frames. ], tot_loss[loss=0.1679, simple_loss=0.2408, pruned_loss=0.04752, over 1443754.16 frames. ], batch size: 44, lr: 1.16e-02, grad_scale: 8.0 +2023-03-20 22:57:44,447 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-20 22:57:52,625 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-20 22:58:01,186 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2286, 1.5979, 1.9183, 1.6362, 1.4215, 1.7729, 1.4181, 1.2703], + device='cuda:0'), covar=tensor([0.0428, 0.0238, 0.0090, 0.0085, 0.0397, 0.0207, 0.0249, 0.0338], + device='cuda:0'), in_proj_covar=tensor([0.0021, 0.0021, 0.0021, 0.0020, 0.0020, 0.0020, 0.0022, 0.0023], + device='cuda:0'), out_proj_covar=tensor([5.3721e-05, 5.2571e-05, 5.1033e-05, 4.6682e-05, 5.2108e-05, 4.9655e-05, + 5.2726e-05, 5.7809e-05], device='cuda:0') +2023-03-20 22:58:09,544 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 22:58:09,997 INFO [train.py:901] (0/2) Epoch 14, batch 1700, loss[loss=0.1768, simple_loss=0.2558, pruned_loss=0.04886, over 7282.00 frames. ], tot_loss[loss=0.1688, simple_loss=0.2417, pruned_loss=0.04796, over 1441867.24 frames. ], batch size: 77, lr: 1.16e-02, grad_scale: 8.0 +2023-03-20 22:58:12,011 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.225e+02 2.418e+02 2.831e+02 3.650e+02 6.823e+02, threshold=5.661e+02, percent-clipped=1.0 +2023-03-20 22:58:13,085 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-20 22:58:14,214 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=38421.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 22:58:23,762 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-20 22:58:35,222 INFO [train.py:901] (0/2) Epoch 14, batch 1750, loss[loss=0.1532, simple_loss=0.2208, pruned_loss=0.04282, over 7198.00 frames. ], tot_loss[loss=0.1677, simple_loss=0.2406, pruned_loss=0.04742, over 1438113.10 frames. ], batch size: 39, lr: 1.16e-02, grad_scale: 8.0 +2023-03-20 22:58:36,521 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4573, 4.3459, 3.9876, 3.6207, 3.8470, 2.6554, 1.9865, 4.4580], + device='cuda:0'), covar=tensor([0.0023, 0.0035, 0.0056, 0.0057, 0.0067, 0.0322, 0.0488, 0.0036], + device='cuda:0'), in_proj_covar=tensor([0.0075, 0.0066, 0.0087, 0.0073, 0.0090, 0.0110, 0.0113, 0.0075], + device='cuda:0'), out_proj_covar=tensor([9.5856e-05, 9.5990e-05, 1.1420e-04, 1.0268e-04, 1.1888e-04, 1.4841e-04, + 1.5146e-04, 9.8931e-05], device='cuda:0') +2023-03-20 22:58:48,475 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-20 22:58:48,988 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-20 22:59:01,588 INFO [train.py:901] (0/2) Epoch 14, batch 1800, loss[loss=0.1845, simple_loss=0.2537, pruned_loss=0.05764, over 7320.00 frames. ], tot_loss[loss=0.1684, simple_loss=0.2414, pruned_loss=0.0477, over 1440965.40 frames. ], batch size: 83, lr: 1.16e-02, grad_scale: 8.0 +2023-03-20 22:59:03,605 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.490e+02 2.294e+02 2.677e+02 3.178e+02 6.312e+02, threshold=5.355e+02, percent-clipped=2.0 +2023-03-20 22:59:11,168 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-20 22:59:25,455 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-20 22:59:27,484 INFO [train.py:901] (0/2) Epoch 14, batch 1850, loss[loss=0.1561, simple_loss=0.2337, pruned_loss=0.03928, over 7340.00 frames. ], tot_loss[loss=0.1675, simple_loss=0.2407, pruned_loss=0.04717, over 1440345.96 frames. ], batch size: 73, lr: 1.16e-02, grad_scale: 8.0 +2023-03-20 22:59:35,640 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-20 22:59:52,575 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-20 22:59:53,599 INFO [train.py:901] (0/2) Epoch 14, batch 1900, loss[loss=0.1879, simple_loss=0.2582, pruned_loss=0.05883, over 7350.00 frames. ], tot_loss[loss=0.167, simple_loss=0.2403, pruned_loss=0.04684, over 1440213.74 frames. ], batch size: 54, lr: 1.16e-02, grad_scale: 8.0 +2023-03-20 22:59:55,556 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.421e+02 2.311e+02 2.784e+02 3.401e+02 7.573e+02, threshold=5.568e+02, percent-clipped=6.0 +2023-03-20 22:59:55,699 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9562, 3.8742, 3.5148, 3.2876, 3.2508, 2.1089, 1.9317, 3.9680], + device='cuda:0'), covar=tensor([0.0032, 0.0037, 0.0075, 0.0051, 0.0095, 0.0423, 0.0493, 0.0044], + device='cuda:0'), in_proj_covar=tensor([0.0075, 0.0067, 0.0087, 0.0073, 0.0091, 0.0110, 0.0113, 0.0075], + device='cuda:0'), out_proj_covar=tensor([9.5580e-05, 9.6666e-05, 1.1495e-04, 1.0249e-04, 1.1984e-04, 1.4870e-04, + 1.5129e-04, 9.8777e-05], device='cuda:0') +2023-03-20 23:00:04,709 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=38635.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:00:18,423 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-20 23:00:19,864 INFO [train.py:901] (0/2) Epoch 14, batch 1950, loss[loss=0.1912, simple_loss=0.2591, pruned_loss=0.06168, over 7377.00 frames. ], tot_loss[loss=0.1674, simple_loss=0.2407, pruned_loss=0.04704, over 1441849.81 frames. ], batch size: 65, lr: 1.16e-02, grad_scale: 8.0 +2023-03-20 23:00:21,009 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2188, 4.0494, 3.7032, 3.3851, 3.4089, 2.2825, 1.7700, 4.1842], + device='cuda:0'), covar=tensor([0.0023, 0.0036, 0.0053, 0.0057, 0.0077, 0.0397, 0.0507, 0.0038], + device='cuda:0'), in_proj_covar=tensor([0.0074, 0.0066, 0.0086, 0.0072, 0.0090, 0.0109, 0.0112, 0.0075], + device='cuda:0'), out_proj_covar=tensor([9.3977e-05, 9.5785e-05, 1.1414e-04, 1.0159e-04, 1.1849e-04, 1.4713e-04, + 1.5002e-04, 9.7990e-05], device='cuda:0') +2023-03-20 23:00:25,032 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7486, 3.1446, 3.5076, 3.6768, 3.4934, 3.6104, 3.4637, 3.4862], + device='cuda:0'), covar=tensor([0.0026, 0.0093, 0.0037, 0.0038, 0.0044, 0.0032, 0.0045, 0.0053], + device='cuda:0'), in_proj_covar=tensor([0.0034, 0.0046, 0.0040, 0.0038, 0.0039, 0.0040, 0.0043, 0.0050], + device='cuda:0'), out_proj_covar=tensor([7.8241e-05, 1.2241e-04, 1.0592e-04, 8.9187e-05, 9.3400e-05, 9.5954e-05, + 1.1271e-04, 1.2337e-04], device='cuda:0') +2023-03-20 23:00:28,923 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-20 23:00:29,960 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=38683.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:00:33,782 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-20 23:00:34,317 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-20 23:00:44,914 INFO [train.py:901] (0/2) Epoch 14, batch 2000, loss[loss=0.1665, simple_loss=0.2486, pruned_loss=0.04216, over 7219.00 frames. ], tot_loss[loss=0.1685, simple_loss=0.2415, pruned_loss=0.0477, over 1442957.28 frames. ], batch size: 93, lr: 1.16e-02, grad_scale: 8.0 +2023-03-20 23:00:46,507 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=38716.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:00:46,883 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.473e+02 2.269e+02 2.724e+02 3.233e+02 7.728e+02, threshold=5.449e+02, percent-clipped=4.0 +2023-03-20 23:00:50,459 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-20 23:01:02,183 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-20 23:01:03,515 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.22 vs. limit=2.0 +2023-03-20 23:01:10,338 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-20 23:01:11,304 INFO [train.py:901] (0/2) Epoch 14, batch 2050, loss[loss=0.154, simple_loss=0.2297, pruned_loss=0.03913, over 7266.00 frames. ], tot_loss[loss=0.1688, simple_loss=0.2419, pruned_loss=0.04783, over 1443890.09 frames. ], batch size: 52, lr: 1.16e-02, grad_scale: 8.0 +2023-03-20 23:01:26,502 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-20 23:01:36,575 INFO [train.py:901] (0/2) Epoch 14, batch 2100, loss[loss=0.1523, simple_loss=0.2378, pruned_loss=0.0334, over 7234.00 frames. ], tot_loss[loss=0.1692, simple_loss=0.2425, pruned_loss=0.04797, over 1444835.14 frames. ], batch size: 93, lr: 1.16e-02, grad_scale: 8.0 +2023-03-20 23:01:39,201 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.563e+02 2.353e+02 2.685e+02 3.230e+02 6.902e+02, threshold=5.371e+02, percent-clipped=1.0 +2023-03-20 23:01:42,265 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=38823.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:01:42,645 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-20 23:01:46,170 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-20 23:01:50,957 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.94 vs. limit=5.0 +2023-03-20 23:01:58,158 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0908, 2.4524, 2.0977, 3.7398, 1.4850, 3.3254, 1.2603, 3.0373], + device='cuda:0'), covar=tensor([0.0081, 0.0833, 0.1603, 0.0061, 0.4185, 0.0089, 0.1092, 0.0198], + device='cuda:0'), in_proj_covar=tensor([0.0159, 0.0257, 0.0295, 0.0150, 0.0284, 0.0158, 0.0265, 0.0211], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 23:02:02,513 INFO [train.py:901] (0/2) Epoch 14, batch 2150, loss[loss=0.1576, simple_loss=0.2359, pruned_loss=0.03964, over 7331.00 frames. ], tot_loss[loss=0.1691, simple_loss=0.2425, pruned_loss=0.04784, over 1445455.82 frames. ], batch size: 61, lr: 1.16e-02, grad_scale: 8.0 +2023-03-20 23:02:13,061 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=38884.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:02:21,463 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-20 23:02:26,622 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.97 vs. limit=2.0 +2023-03-20 23:02:28,848 INFO [train.py:901] (0/2) Epoch 14, batch 2200, loss[loss=0.1898, simple_loss=0.2644, pruned_loss=0.05757, over 7119.00 frames. ], tot_loss[loss=0.1699, simple_loss=0.2434, pruned_loss=0.04823, over 1446371.32 frames. ], batch size: 98, lr: 1.16e-02, grad_scale: 8.0 +2023-03-20 23:02:30,371 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-20 23:02:30,843 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.439e+02 2.207e+02 2.666e+02 3.268e+02 4.931e+02, threshold=5.333e+02, percent-clipped=0.0 +2023-03-20 23:02:32,455 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2820, 4.8098, 4.8677, 4.7762, 4.7516, 4.3970, 4.8594, 4.6856], + device='cuda:0'), covar=tensor([0.0382, 0.0300, 0.0223, 0.0346, 0.0251, 0.0272, 0.0243, 0.0423], + device='cuda:0'), in_proj_covar=tensor([0.0166, 0.0191, 0.0134, 0.0142, 0.0119, 0.0174, 0.0150, 0.0117], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:02:53,829 INFO [train.py:901] (0/2) Epoch 14, batch 2250, loss[loss=0.1822, simple_loss=0.2509, pruned_loss=0.05677, over 7252.00 frames. ], tot_loss[loss=0.1691, simple_loss=0.2425, pruned_loss=0.04786, over 1444360.89 frames. ], batch size: 47, lr: 1.15e-02, grad_scale: 8.0 +2023-03-20 23:02:59,990 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5106, 4.2121, 3.9784, 3.6935, 4.0253, 2.4780, 2.0710, 4.4866], + device='cuda:0'), covar=tensor([0.0019, 0.0076, 0.0060, 0.0057, 0.0049, 0.0398, 0.0510, 0.0032], + device='cuda:0'), in_proj_covar=tensor([0.0075, 0.0068, 0.0088, 0.0074, 0.0093, 0.0112, 0.0116, 0.0076], + device='cuda:0'), out_proj_covar=tensor([9.5326e-05, 9.8918e-05, 1.1656e-04, 1.0332e-04, 1.2244e-04, 1.5079e-04, + 1.5462e-04, 9.9676e-05], device='cuda:0') +2023-03-20 23:03:04,528 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-20 23:03:05,026 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-20 23:03:17,876 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-20 23:03:20,296 INFO [train.py:901] (0/2) Epoch 14, batch 2300, loss[loss=0.17, simple_loss=0.245, pruned_loss=0.04756, over 7322.00 frames. ], tot_loss[loss=0.1688, simple_loss=0.2426, pruned_loss=0.04755, over 1445135.09 frames. ], batch size: 59, lr: 1.15e-02, grad_scale: 8.0 +2023-03-20 23:03:21,942 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=39016.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:03:22,306 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.345e+02 2.210e+02 2.631e+02 3.311e+02 5.841e+02, threshold=5.262e+02, percent-clipped=4.0 +2023-03-20 23:03:25,704 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.52 vs. limit=5.0 +2023-03-20 23:03:39,531 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0763, 2.6190, 2.0813, 3.0018, 2.5819, 2.8411, 2.2812, 2.5377], + device='cuda:0'), covar=tensor([0.1560, 0.0599, 0.2432, 0.0364, 0.0066, 0.0054, 0.0132, 0.0174], + device='cuda:0'), in_proj_covar=tensor([0.0294, 0.0231, 0.0281, 0.0266, 0.0137, 0.0132, 0.0158, 0.0184], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:03:45,472 INFO [train.py:901] (0/2) Epoch 14, batch 2350, loss[loss=0.1737, simple_loss=0.2548, pruned_loss=0.04633, over 7260.00 frames. ], tot_loss[loss=0.1683, simple_loss=0.2419, pruned_loss=0.04732, over 1444459.41 frames. ], batch size: 64, lr: 1.15e-02, grad_scale: 8.0 +2023-03-20 23:03:46,031 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=39064.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:04:06,143 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-20 23:04:11,731 INFO [train.py:901] (0/2) Epoch 14, batch 2400, loss[loss=0.1636, simple_loss=0.2396, pruned_loss=0.0438, over 7345.00 frames. ], tot_loss[loss=0.168, simple_loss=0.2416, pruned_loss=0.04727, over 1442045.50 frames. ], batch size: 73, lr: 1.15e-02, grad_scale: 8.0 +2023-03-20 23:04:12,249 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-20 23:04:13,734 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.544e+02 2.325e+02 2.690e+02 3.201e+02 5.267e+02, threshold=5.381e+02, percent-clipped=1.0 +2023-03-20 23:04:14,922 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1239, 2.3864, 2.0343, 3.7143, 1.4964, 3.2961, 1.2160, 3.0251], + device='cuda:0'), covar=tensor([0.0058, 0.0883, 0.1767, 0.0063, 0.4086, 0.0084, 0.1160, 0.0301], + device='cuda:0'), in_proj_covar=tensor([0.0155, 0.0252, 0.0290, 0.0147, 0.0279, 0.0156, 0.0259, 0.0206], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 23:04:16,721 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-03-20 23:04:22,285 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-20 23:04:24,741 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-20 23:04:37,364 INFO [train.py:901] (0/2) Epoch 14, batch 2450, loss[loss=0.1665, simple_loss=0.2405, pruned_loss=0.04619, over 7328.00 frames. ], tot_loss[loss=0.1675, simple_loss=0.241, pruned_loss=0.04694, over 1442461.78 frames. ], batch size: 54, lr: 1.15e-02, grad_scale: 8.0 +2023-03-20 23:04:46,238 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=39179.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:04:46,257 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9909, 3.5991, 3.7484, 3.7658, 3.5919, 3.5971, 3.7574, 3.2697], + device='cuda:0'), covar=tensor([0.0112, 0.0146, 0.0126, 0.0146, 0.0334, 0.0112, 0.0192, 0.0218], + device='cuda:0'), in_proj_covar=tensor([0.0063, 0.0069, 0.0071, 0.0061, 0.0122, 0.0081, 0.0076, 0.0076], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:04:49,229 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=39185.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:04:52,581 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-20 23:05:03,376 INFO [train.py:901] (0/2) Epoch 14, batch 2500, loss[loss=0.1696, simple_loss=0.2437, pruned_loss=0.04779, over 7258.00 frames. ], tot_loss[loss=0.1673, simple_loss=0.2408, pruned_loss=0.04691, over 1442016.45 frames. ], batch size: 89, lr: 1.15e-02, grad_scale: 8.0 +2023-03-20 23:05:05,415 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.736e+02 2.502e+02 3.022e+02 3.546e+02 7.037e+02, threshold=6.044e+02, percent-clipped=3.0 +2023-03-20 23:05:16,362 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.00 vs. limit=2.0 +2023-03-20 23:05:18,674 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-20 23:05:20,279 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=39246.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:05:29,212 INFO [train.py:901] (0/2) Epoch 14, batch 2550, loss[loss=0.1788, simple_loss=0.2515, pruned_loss=0.05303, over 7314.00 frames. ], tot_loss[loss=0.1677, simple_loss=0.2412, pruned_loss=0.04707, over 1442806.76 frames. ], batch size: 83, lr: 1.15e-02, grad_scale: 8.0 +2023-03-20 23:05:40,518 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=39284.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:05:48,923 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6061, 4.0145, 4.2543, 4.2212, 4.0996, 4.1693, 4.5169, 3.9505], + device='cuda:0'), covar=tensor([0.0097, 0.0120, 0.0088, 0.0104, 0.0309, 0.0095, 0.0106, 0.0136], + device='cuda:0'), in_proj_covar=tensor([0.0064, 0.0071, 0.0071, 0.0062, 0.0123, 0.0081, 0.0075, 0.0076], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:05:54,848 INFO [train.py:901] (0/2) Epoch 14, batch 2600, loss[loss=0.1271, simple_loss=0.1908, pruned_loss=0.03168, over 7040.00 frames. ], tot_loss[loss=0.1675, simple_loss=0.241, pruned_loss=0.04698, over 1442020.55 frames. ], batch size: 35, lr: 1.15e-02, grad_scale: 8.0 +2023-03-20 23:05:56,768 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.537e+02 2.181e+02 2.670e+02 3.207e+02 6.854e+02, threshold=5.339e+02, percent-clipped=3.0 +2023-03-20 23:06:10,672 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=39345.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:06:16,397 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0355, 3.5242, 3.6222, 3.6480, 3.4993, 3.5183, 3.8447, 3.3463], + device='cuda:0'), covar=tensor([0.0091, 0.0189, 0.0121, 0.0146, 0.0394, 0.0128, 0.0161, 0.0171], + device='cuda:0'), in_proj_covar=tensor([0.0063, 0.0072, 0.0071, 0.0062, 0.0123, 0.0080, 0.0075, 0.0076], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:06:19,253 INFO [train.py:901] (0/2) Epoch 14, batch 2650, loss[loss=0.1711, simple_loss=0.2533, pruned_loss=0.04438, over 7259.00 frames. ], tot_loss[loss=0.1676, simple_loss=0.2412, pruned_loss=0.04697, over 1443360.13 frames. ], batch size: 89, lr: 1.15e-02, grad_scale: 8.0 +2023-03-20 23:06:44,697 INFO [train.py:901] (0/2) Epoch 14, batch 2700, loss[loss=0.1518, simple_loss=0.2337, pruned_loss=0.03493, over 7357.00 frames. ], tot_loss[loss=0.1671, simple_loss=0.2408, pruned_loss=0.0467, over 1440157.50 frames. ], batch size: 63, lr: 1.15e-02, grad_scale: 8.0 +2023-03-20 23:06:46,670 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.412e+02 2.100e+02 2.474e+02 3.007e+02 6.884e+02, threshold=4.948e+02, percent-clipped=2.0 +2023-03-20 23:07:09,771 INFO [train.py:901] (0/2) Epoch 14, batch 2750, loss[loss=0.1587, simple_loss=0.235, pruned_loss=0.04122, over 7309.00 frames. ], tot_loss[loss=0.1678, simple_loss=0.2412, pruned_loss=0.04727, over 1440498.36 frames. ], batch size: 80, lr: 1.15e-02, grad_scale: 8.0 +2023-03-20 23:07:09,892 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=39463.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:07:14,317 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9436, 2.3361, 1.8036, 3.0747, 2.5714, 2.8481, 2.4725, 2.5744], + device='cuda:0'), covar=tensor([0.1778, 0.0795, 0.2969, 0.0669, 0.0108, 0.0066, 0.0135, 0.0168], + device='cuda:0'), in_proj_covar=tensor([0.0286, 0.0228, 0.0275, 0.0257, 0.0137, 0.0129, 0.0154, 0.0182], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:07:17,607 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=39479.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:07:33,828 INFO [train.py:901] (0/2) Epoch 14, batch 2800, loss[loss=0.1611, simple_loss=0.2288, pruned_loss=0.04665, over 7206.00 frames. ], tot_loss[loss=0.168, simple_loss=0.2417, pruned_loss=0.04716, over 1442684.92 frames. ], batch size: 50, lr: 1.15e-02, grad_scale: 8.0 +2023-03-20 23:07:35,754 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.700e+02 2.191e+02 2.661e+02 3.280e+02 1.218e+03, threshold=5.323e+02, percent-clipped=4.0 +2023-03-20 23:07:39,324 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=39524.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:07:40,718 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=39527.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:07:46,566 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-14.pt +2023-03-20 23:08:05,560 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-20 23:08:05,616 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-20 23:08:05,641 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-20 23:08:05,659 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-20 23:08:05,841 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-20 23:08:05,857 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-20 23:08:05,868 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-20 23:08:05,967 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-20 23:08:06,393 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-20 23:08:06,408 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-20 23:08:06,425 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-20 23:08:06,593 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-20 23:08:06,599 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-20 23:08:06,607 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-20 23:08:06,748 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-20 23:08:06,751 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-20 23:08:06,810 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-20 23:08:09,222 INFO [train.py:901] (0/2) Epoch 15, batch 0, loss[loss=0.1585, simple_loss=0.2384, pruned_loss=0.03928, over 7260.00 frames. ], tot_loss[loss=0.1585, simple_loss=0.2384, pruned_loss=0.03928, over 7260.00 frames. ], batch size: 89, lr: 1.11e-02, grad_scale: 8.0 +2023-03-20 23:08:09,224 INFO [train.py:926] (0/2) Computing validation loss +2023-03-20 23:08:13,256 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0082, 3.1314, 2.7688, 2.9873, 3.0452, 2.5374, 3.3595, 3.1518], + device='cuda:0'), covar=tensor([0.0591, 0.1258, 0.1124, 0.1919, 0.3967, 0.0726, 0.0428, 0.0713], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0041, 0.0047, 0.0042, 0.0043, 0.0043, 0.0043, 0.0040], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:08:28,516 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3917, 3.6963, 3.9127, 3.9997, 4.1013, 3.8950, 4.1302, 3.8965], + device='cuda:0'), covar=tensor([0.0089, 0.0161, 0.0136, 0.0139, 0.0329, 0.0124, 0.0174, 0.0125], + device='cuda:0'), in_proj_covar=tensor([0.0063, 0.0070, 0.0071, 0.0061, 0.0122, 0.0079, 0.0075, 0.0075], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0001, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:08:30,748 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3608, 1.9824, 1.7752, 1.7389, 1.4439, 1.6609, 1.8842, 1.5836], + device='cuda:0'), covar=tensor([0.0293, 0.0494, 0.0181, 0.0106, 0.0547, 0.0384, 0.0193, 0.0249], + device='cuda:0'), in_proj_covar=tensor([0.0022, 0.0023, 0.0022, 0.0021, 0.0021, 0.0021, 0.0023, 0.0024], + device='cuda:0'), out_proj_covar=tensor([5.5820e-05, 5.7003e-05, 5.1662e-05, 4.8850e-05, 5.3773e-05, 5.2706e-05, + 5.4645e-05, 6.0357e-05], device='cuda:0') +2023-03-20 23:08:32,928 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3119, 4.0676, 3.5330, 3.6128, 3.9467, 2.3911, 1.8400, 4.3275], + device='cuda:0'), covar=tensor([0.0017, 0.0032, 0.0078, 0.0060, 0.0037, 0.0433, 0.0533, 0.0031], + device='cuda:0'), in_proj_covar=tensor([0.0074, 0.0067, 0.0087, 0.0073, 0.0093, 0.0111, 0.0114, 0.0077], + device='cuda:0'), out_proj_covar=tensor([9.3879e-05, 9.7424e-05, 1.1427e-04, 1.0210e-04, 1.2239e-04, 1.4984e-04, + 1.5256e-04, 1.0034e-04], device='cuda:0') +2023-03-20 23:08:35,274 INFO [train.py:935] (0/2) Epoch 15, validation: loss=0.1689, simple_loss=0.2557, pruned_loss=0.04109, over 1622729.00 frames. +2023-03-20 23:08:35,274 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-20 23:08:37,362 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=39541.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:08:42,281 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-20 23:08:52,814 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-20 23:08:59,295 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-20 23:09:00,284 INFO [train.py:901] (0/2) Epoch 15, batch 50, loss[loss=0.1479, simple_loss=0.2299, pruned_loss=0.03297, over 7291.00 frames. ], tot_loss[loss=0.1661, simple_loss=0.2398, pruned_loss=0.04623, over 324340.02 frames. ], batch size: 86, lr: 1.11e-02, grad_scale: 8.0 +2023-03-20 23:09:01,286 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-20 23:09:04,252 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-20 23:09:16,182 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.268e+02 2.158e+02 2.629e+02 3.348e+02 6.220e+02, threshold=5.258e+02, percent-clipped=5.0 +2023-03-20 23:09:21,379 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-20 23:09:21,920 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-20 23:09:26,890 INFO [train.py:901] (0/2) Epoch 15, batch 100, loss[loss=0.1678, simple_loss=0.2364, pruned_loss=0.04961, over 7328.00 frames. ], tot_loss[loss=0.1677, simple_loss=0.2421, pruned_loss=0.04669, over 573302.89 frames. ], batch size: 44, lr: 1.11e-02, grad_scale: 8.0 +2023-03-20 23:09:28,020 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=39639.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 23:09:28,440 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=39640.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:09:29,523 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5204, 2.2753, 2.1893, 3.8554, 1.5935, 3.4974, 1.4570, 3.1118], + device='cuda:0'), covar=tensor([0.0066, 0.0889, 0.1717, 0.0053, 0.3922, 0.0094, 0.0980, 0.0262], + device='cuda:0'), in_proj_covar=tensor([0.0161, 0.0260, 0.0302, 0.0152, 0.0282, 0.0160, 0.0268, 0.0214], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 23:09:43,765 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.93 vs. limit=5.0 +2023-03-20 23:09:44,482 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.0492, 4.6023, 4.4911, 5.0380, 5.0133, 5.0832, 4.1382, 4.7238], + device='cuda:0'), covar=tensor([0.0756, 0.2388, 0.2170, 0.1212, 0.0767, 0.1332, 0.0827, 0.0900], + device='cuda:0'), in_proj_covar=tensor([0.0191, 0.0303, 0.0244, 0.0238, 0.0177, 0.0293, 0.0165, 0.0208], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:09:52,631 INFO [train.py:901] (0/2) Epoch 15, batch 150, loss[loss=0.1718, simple_loss=0.2512, pruned_loss=0.04624, over 7313.00 frames. ], tot_loss[loss=0.1658, simple_loss=0.2398, pruned_loss=0.04584, over 765121.82 frames. ], batch size: 49, lr: 1.11e-02, grad_scale: 8.0 +2023-03-20 23:09:59,353 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=39700.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 23:10:08,388 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.494e+02 2.211e+02 2.562e+02 3.102e+02 7.128e+02, threshold=5.124e+02, percent-clipped=3.0 +2023-03-20 23:10:18,272 INFO [train.py:901] (0/2) Epoch 15, batch 200, loss[loss=0.1432, simple_loss=0.2177, pruned_loss=0.03439, over 7153.00 frames. ], tot_loss[loss=0.166, simple_loss=0.2398, pruned_loss=0.04608, over 914604.89 frames. ], batch size: 41, lr: 1.11e-02, grad_scale: 8.0 +2023-03-20 23:10:18,652 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-20 23:10:22,303 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-20 23:10:26,797 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-20 23:10:30,451 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4780, 2.9783, 3.0898, 3.3409, 2.8560, 2.5998, 3.3355, 2.7038], + device='cuda:0'), covar=tensor([0.0265, 0.0325, 0.0325, 0.0335, 0.0388, 0.0463, 0.0375, 0.1047], + device='cuda:0'), in_proj_covar=tensor([0.0300, 0.0302, 0.0253, 0.0318, 0.0308, 0.0301, 0.0307, 0.0295], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 23:10:32,807 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-20 23:10:33,403 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=39767.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:10:43,923 INFO [train.py:901] (0/2) Epoch 15, batch 250, loss[loss=0.1315, simple_loss=0.2039, pruned_loss=0.02951, over 7180.00 frames. ], tot_loss[loss=0.1658, simple_loss=0.2392, pruned_loss=0.04618, over 1031562.36 frames. ], batch size: 39, lr: 1.11e-02, grad_scale: 8.0 +2023-03-20 23:10:46,571 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-20 23:10:57,010 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5556, 3.0253, 2.3836, 2.8236, 2.7904, 2.1969, 2.9067, 2.7222], + device='cuda:0'), covar=tensor([0.0655, 0.0397, 0.1272, 0.1256, 0.1424, 0.1230, 0.0873, 0.1137], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0041, 0.0048, 0.0041, 0.0042, 0.0044, 0.0044, 0.0041], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:10:58,963 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=39815.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:10:59,824 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.607e+02 2.227e+02 2.655e+02 3.135e+02 9.236e+02, threshold=5.310e+02, percent-clipped=3.0 +2023-03-20 23:11:00,932 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=39819.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:11:05,494 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=39828.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:11:07,845 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-20 23:11:09,811 INFO [train.py:901] (0/2) Epoch 15, batch 300, loss[loss=0.1786, simple_loss=0.2513, pruned_loss=0.05292, over 7278.00 frames. ], tot_loss[loss=0.1647, simple_loss=0.238, pruned_loss=0.04572, over 1119779.83 frames. ], batch size: 66, lr: 1.11e-02, grad_scale: 8.0 +2023-03-20 23:11:11,856 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=39841.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:11:16,244 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-20 23:11:30,553 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=39876.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:11:32,066 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7369, 2.0045, 1.9050, 3.2457, 1.4486, 2.9913, 1.2923, 2.7931], + device='cuda:0'), covar=tensor([0.0068, 0.0847, 0.1619, 0.0070, 0.3896, 0.0087, 0.1047, 0.0175], + device='cuda:0'), in_proj_covar=tensor([0.0158, 0.0259, 0.0302, 0.0153, 0.0282, 0.0160, 0.0268, 0.0214], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 23:11:32,545 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5269, 2.8584, 2.4715, 2.8572, 2.8240, 2.2189, 2.7876, 2.7244], + device='cuda:0'), covar=tensor([0.1394, 0.1505, 0.1643, 0.1276, 0.0832, 0.0998, 0.1273, 0.1346], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0042, 0.0049, 0.0042, 0.0043, 0.0044, 0.0046, 0.0041], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:11:35,980 INFO [train.py:901] (0/2) Epoch 15, batch 350, loss[loss=0.1534, simple_loss=0.2406, pruned_loss=0.03311, over 7308.00 frames. ], tot_loss[loss=0.1643, simple_loss=0.2382, pruned_loss=0.04515, over 1193720.55 frames. ], batch size: 59, lr: 1.11e-02, grad_scale: 8.0 +2023-03-20 23:11:37,056 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=39889.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:11:50,881 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.396e+02 2.229e+02 2.631e+02 3.067e+02 9.607e+02, threshold=5.263e+02, percent-clipped=2.0 +2023-03-20 23:11:50,919 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-20 23:12:00,948 INFO [train.py:901] (0/2) Epoch 15, batch 400, loss[loss=0.1804, simple_loss=0.2562, pruned_loss=0.05227, over 7138.00 frames. ], tot_loss[loss=0.1656, simple_loss=0.2397, pruned_loss=0.04577, over 1249978.04 frames. ], batch size: 98, lr: 1.11e-02, grad_scale: 8.0 +2023-03-20 23:12:02,506 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=39940.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:12:26,855 INFO [train.py:901] (0/2) Epoch 15, batch 450, loss[loss=0.1546, simple_loss=0.2328, pruned_loss=0.03814, over 7331.00 frames. ], tot_loss[loss=0.1664, simple_loss=0.2405, pruned_loss=0.0462, over 1294314.31 frames. ], batch size: 75, lr: 1.11e-02, grad_scale: 8.0 +2023-03-20 23:12:27,418 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=39988.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:12:30,296 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-20 23:12:30,796 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-20 23:12:30,855 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=39995.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 23:12:33,623 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-40000.pt +2023-03-20 23:12:45,601 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.293e+02 2.173e+02 2.880e+02 3.214e+02 6.134e+02, threshold=5.760e+02, percent-clipped=1.0 +2023-03-20 23:12:49,610 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6554, 2.6910, 3.6887, 3.7620, 3.6741, 3.6295, 3.5746, 3.2407], + device='cuda:0'), covar=tensor([0.0031, 0.0149, 0.0036, 0.0035, 0.0040, 0.0045, 0.0047, 0.0078], + device='cuda:0'), in_proj_covar=tensor([0.0033, 0.0045, 0.0041, 0.0038, 0.0039, 0.0040, 0.0043, 0.0050], + device='cuda:0'), out_proj_covar=tensor([7.5681e-05, 1.2014e-04, 1.0480e-04, 8.8573e-05, 9.0757e-05, 9.5641e-05, + 1.0994e-04, 1.2068e-04], device='cuda:0') +2023-03-20 23:12:56,077 INFO [train.py:901] (0/2) Epoch 15, batch 500, loss[loss=0.1694, simple_loss=0.2461, pruned_loss=0.04633, over 7275.00 frames. ], tot_loss[loss=0.1662, simple_loss=0.2402, pruned_loss=0.04609, over 1325299.79 frames. ], batch size: 77, lr: 1.10e-02, grad_scale: 8.0 +2023-03-20 23:13:01,544 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.00 vs. limit=2.0 +2023-03-20 23:13:06,706 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-20 23:13:08,635 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-20 23:13:09,570 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-20 23:13:11,546 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-20 23:13:16,619 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-20 23:13:21,553 INFO [train.py:901] (0/2) Epoch 15, batch 550, loss[loss=0.151, simple_loss=0.2334, pruned_loss=0.03434, over 7327.00 frames. ], tot_loss[loss=0.1663, simple_loss=0.24, pruned_loss=0.0463, over 1349433.52 frames. ], batch size: 61, lr: 1.10e-02, grad_scale: 8.0 +2023-03-20 23:13:28,018 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-20 23:13:36,003 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-20 23:13:36,487 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.576e+02 2.123e+02 2.573e+02 3.365e+02 8.835e+02, threshold=5.146e+02, percent-clipped=2.0 +2023-03-20 23:13:38,252 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=40119.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:13:39,241 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7216, 4.0117, 3.6984, 3.8099, 3.7821, 4.0044, 4.2763, 4.3086], + device='cuda:0'), covar=tensor([0.0226, 0.0167, 0.0201, 0.0194, 0.0274, 0.0258, 0.0270, 0.0190], + device='cuda:0'), in_proj_covar=tensor([0.0100, 0.0105, 0.0094, 0.0105, 0.0096, 0.0087, 0.0081, 0.0079], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:13:40,147 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-20 23:13:40,205 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=40123.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:13:47,336 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-20 23:13:47,831 INFO [train.py:901] (0/2) Epoch 15, batch 600, loss[loss=0.1653, simple_loss=0.2364, pruned_loss=0.04707, over 7205.00 frames. ], tot_loss[loss=0.1657, simple_loss=0.2396, pruned_loss=0.04591, over 1370305.07 frames. ], batch size: 50, lr: 1.10e-02, grad_scale: 8.0 +2023-03-20 23:13:49,177 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=6.03 vs. limit=5.0 +2023-03-20 23:14:02,896 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-20 23:14:02,939 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=40167.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:14:04,968 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=40171.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:14:11,948 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-20 23:14:12,943 INFO [train.py:901] (0/2) Epoch 15, batch 650, loss[loss=0.1347, simple_loss=0.2003, pruned_loss=0.03451, over 7032.00 frames. ], tot_loss[loss=0.1656, simple_loss=0.2393, pruned_loss=0.04601, over 1386346.44 frames. ], batch size: 35, lr: 1.10e-02, grad_scale: 8.0 +2023-03-20 23:14:17,053 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=40195.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:14:23,984 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1047, 3.9990, 4.1187, 3.8694, 3.8939, 4.2320, 4.0832, 3.8400], + device='cuda:0'), covar=tensor([0.0032, 0.0050, 0.0032, 0.0039, 0.0042, 0.0025, 0.0027, 0.0055], + device='cuda:0'), in_proj_covar=tensor([0.0033, 0.0045, 0.0041, 0.0037, 0.0039, 0.0040, 0.0043, 0.0049], + device='cuda:0'), out_proj_covar=tensor([7.5693e-05, 1.1822e-04, 1.0438e-04, 8.7592e-05, 9.2114e-05, 9.4537e-05, + 1.1042e-04, 1.1875e-04], device='cuda:0') +2023-03-20 23:14:29,413 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.498e+02 2.119e+02 2.439e+02 2.967e+02 4.703e+02, threshold=4.878e+02, percent-clipped=0.0 +2023-03-20 23:14:29,455 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-20 23:14:30,074 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=40218.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:14:37,451 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-20 23:14:39,385 INFO [train.py:901] (0/2) Epoch 15, batch 700, loss[loss=0.1542, simple_loss=0.2283, pruned_loss=0.0401, over 7311.00 frames. ], tot_loss[loss=0.1662, simple_loss=0.2399, pruned_loss=0.04622, over 1399653.36 frames. ], batch size: 49, lr: 1.10e-02, grad_scale: 8.0 +2023-03-20 23:14:41,007 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4447, 2.2592, 2.0087, 1.9712, 1.6443, 1.6809, 1.8516, 1.6426], + device='cuda:0'), covar=tensor([0.0336, 0.0220, 0.0119, 0.0102, 0.0608, 0.0331, 0.0216, 0.0201], + device='cuda:0'), in_proj_covar=tensor([0.0023, 0.0022, 0.0021, 0.0020, 0.0020, 0.0020, 0.0022, 0.0024], + device='cuda:0'), out_proj_covar=tensor([5.6811e-05, 5.4350e-05, 5.0939e-05, 4.6543e-05, 5.2615e-05, 5.1547e-05, + 5.3131e-05, 5.9542e-05], device='cuda:0') +2023-03-20 23:14:49,021 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=40256.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:14:51,798 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.94 vs. limit=2.0 +2023-03-20 23:14:57,140 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=40272.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:15:00,608 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=40279.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:15:00,970 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-20 23:15:01,955 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-20 23:15:02,037 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.0336, 5.5564, 5.5542, 5.5453, 5.2783, 5.1014, 5.6256, 5.2513], + device='cuda:0'), covar=tensor([0.0410, 0.0329, 0.0283, 0.0396, 0.0338, 0.0270, 0.0245, 0.0568], + device='cuda:0'), in_proj_covar=tensor([0.0175, 0.0200, 0.0146, 0.0147, 0.0125, 0.0185, 0.0156, 0.0124], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:15:04,421 INFO [train.py:901] (0/2) Epoch 15, batch 750, loss[loss=0.1577, simple_loss=0.2272, pruned_loss=0.04409, over 7134.00 frames. ], tot_loss[loss=0.1663, simple_loss=0.2402, pruned_loss=0.04623, over 1408680.74 frames. ], batch size: 41, lr: 1.10e-02, grad_scale: 8.0 +2023-03-20 23:15:09,113 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=40295.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 23:15:13,383 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-20 23:15:17,630 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-20 23:15:20,573 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.345e+02 2.287e+02 2.600e+02 3.173e+02 8.989e+02, threshold=5.199e+02, percent-clipped=3.0 +2023-03-20 23:15:22,281 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-20 23:15:27,546 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=40330.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:15:29,108 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=40333.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:15:29,502 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-20 23:15:30,546 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-20 23:15:31,021 INFO [train.py:901] (0/2) Epoch 15, batch 800, loss[loss=0.1724, simple_loss=0.2455, pruned_loss=0.04964, over 7306.00 frames. ], tot_loss[loss=0.1651, simple_loss=0.2389, pruned_loss=0.04568, over 1413353.73 frames. ], batch size: 86, lr: 1.10e-02, grad_scale: 8.0 +2023-03-20 23:15:34,093 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=40343.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 23:15:41,027 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-20 23:15:43,142 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0785, 3.0124, 2.9830, 2.9955, 3.1934, 2.7508, 2.6131, 3.1418], + device='cuda:0'), covar=tensor([0.2025, 0.0695, 0.1625, 0.2314, 0.1182, 0.1613, 0.2198, 0.1589], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0044, 0.0038, 0.0039, 0.0036, 0.0034, 0.0052, 0.0039], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:15:56,612 INFO [train.py:901] (0/2) Epoch 15, batch 850, loss[loss=0.1664, simple_loss=0.2478, pruned_loss=0.04253, over 7271.00 frames. ], tot_loss[loss=0.1658, simple_loss=0.2395, pruned_loss=0.04606, over 1419728.46 frames. ], batch size: 55, lr: 1.10e-02, grad_scale: 16.0 +2023-03-20 23:15:59,322 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=40391.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:16:00,218 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-20 23:16:00,652 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-20 23:16:06,028 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-20 23:16:09,438 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-20 23:16:12,447 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.364e+02 2.208e+02 2.627e+02 3.087e+02 7.163e+02, threshold=5.253e+02, percent-clipped=1.0 +2023-03-20 23:16:14,840 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.67 vs. limit=5.0 +2023-03-20 23:16:15,624 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=40423.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:16:22,546 INFO [train.py:901] (0/2) Epoch 15, batch 900, loss[loss=0.1346, simple_loss=0.204, pruned_loss=0.03256, over 6916.00 frames. ], tot_loss[loss=0.1649, simple_loss=0.2386, pruned_loss=0.04563, over 1424482.34 frames. ], batch size: 35, lr: 1.10e-02, grad_scale: 16.0 +2023-03-20 23:16:39,499 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=40471.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:16:39,541 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=40471.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:16:47,483 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-20 23:16:48,943 INFO [train.py:901] (0/2) Epoch 15, batch 950, loss[loss=0.1819, simple_loss=0.2561, pruned_loss=0.05386, over 6700.00 frames. ], tot_loss[loss=0.1655, simple_loss=0.239, pruned_loss=0.04594, over 1427654.65 frames. ], batch size: 107, lr: 1.10e-02, grad_scale: 16.0 +2023-03-20 23:16:51,562 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=40492.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:16:54,960 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9176, 4.4563, 4.4851, 4.4426, 4.4456, 4.0416, 4.5570, 4.3373], + device='cuda:0'), covar=tensor([0.0472, 0.0440, 0.0365, 0.0507, 0.0288, 0.0387, 0.0342, 0.0467], + device='cuda:0'), in_proj_covar=tensor([0.0173, 0.0197, 0.0144, 0.0146, 0.0122, 0.0182, 0.0157, 0.0119], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:17:03,788 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.511e+02 2.192e+02 2.721e+02 3.328e+02 8.028e+02, threshold=5.442e+02, percent-clipped=2.0 +2023-03-20 23:17:04,878 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=40519.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:17:10,328 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-20 23:17:13,160 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.04 vs. limit=2.0 +2023-03-20 23:17:13,800 INFO [train.py:901] (0/2) Epoch 15, batch 1000, loss[loss=0.1368, simple_loss=0.2153, pruned_loss=0.02918, over 7343.00 frames. ], tot_loss[loss=0.1658, simple_loss=0.2394, pruned_loss=0.04614, over 1427518.96 frames. ], batch size: 44, lr: 1.10e-02, grad_scale: 16.0 +2023-03-20 23:17:14,660 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.79 vs. limit=2.0 +2023-03-20 23:17:20,420 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2781, 1.8359, 1.7794, 1.7430, 1.5414, 1.4757, 1.7337, 1.6165], + device='cuda:0'), covar=tensor([0.0528, 0.0221, 0.0243, 0.0093, 0.0615, 0.0410, 0.0225, 0.0225], + device='cuda:0'), in_proj_covar=tensor([0.0023, 0.0022, 0.0022, 0.0020, 0.0022, 0.0021, 0.0023, 0.0025], + device='cuda:0'), out_proj_covar=tensor([5.8539e-05, 5.6143e-05, 5.3385e-05, 4.7864e-05, 5.6165e-05, 5.3102e-05, + 5.5678e-05, 6.1883e-05], device='cuda:0') +2023-03-20 23:17:20,875 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=40551.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:17:21,953 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=40553.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:17:31,580 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-20 23:17:33,661 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=40574.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:17:40,098 INFO [train.py:901] (0/2) Epoch 15, batch 1050, loss[loss=0.1662, simple_loss=0.2451, pruned_loss=0.04368, over 7289.00 frames. ], tot_loss[loss=0.1654, simple_loss=0.239, pruned_loss=0.04595, over 1433133.19 frames. ], batch size: 68, lr: 1.10e-02, grad_scale: 16.0 +2023-03-20 23:17:52,528 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-20 23:17:55,536 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.580e+02 2.140e+02 2.605e+02 3.056e+02 5.614e+02, threshold=5.210e+02, percent-clipped=1.0 +2023-03-20 23:17:56,559 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-20 23:18:01,108 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=40628.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:18:05,662 INFO [train.py:901] (0/2) Epoch 15, batch 1100, loss[loss=0.1575, simple_loss=0.2363, pruned_loss=0.03935, over 7312.00 frames. ], tot_loss[loss=0.1653, simple_loss=0.2389, pruned_loss=0.0458, over 1435588.05 frames. ], batch size: 80, lr: 1.10e-02, grad_scale: 16.0 +2023-03-20 23:18:14,010 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=40651.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:18:20,054 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8966, 2.6107, 3.0353, 3.0943, 2.9709, 2.7852, 2.3743, 2.9460], + device='cuda:0'), covar=tensor([0.1767, 0.0419, 0.1119, 0.1191, 0.0828, 0.0873, 0.2313, 0.1352], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0042, 0.0037, 0.0037, 0.0035, 0.0033, 0.0050, 0.0038], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:18:25,403 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-20 23:18:25,415 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 23:18:31,613 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=40686.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:18:32,047 INFO [train.py:901] (0/2) Epoch 15, batch 1150, loss[loss=0.1549, simple_loss=0.2303, pruned_loss=0.03971, over 7280.00 frames. ], tot_loss[loss=0.1655, simple_loss=0.2391, pruned_loss=0.04596, over 1435607.40 frames. ], batch size: 66, lr: 1.10e-02, grad_scale: 16.0 +2023-03-20 23:18:38,366 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.44 vs. limit=2.0 +2023-03-20 23:18:38,562 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-20 23:18:38,577 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-20 23:18:44,716 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=40712.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:18:47,054 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.479e+02 2.199e+02 2.637e+02 2.987e+02 6.275e+02, threshold=5.274e+02, percent-clipped=1.0 +2023-03-20 23:18:58,331 INFO [train.py:901] (0/2) Epoch 15, batch 1200, loss[loss=0.1827, simple_loss=0.2557, pruned_loss=0.05481, over 7297.00 frames. ], tot_loss[loss=0.1651, simple_loss=0.2388, pruned_loss=0.04566, over 1438100.86 frames. ], batch size: 57, lr: 1.10e-02, grad_scale: 16.0 +2023-03-20 23:18:58,743 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.82 vs. limit=2.0 +2023-03-20 23:19:10,005 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-20 23:19:21,753 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-20 23:19:23,440 INFO [train.py:901] (0/2) Epoch 15, batch 1250, loss[loss=0.2011, simple_loss=0.2778, pruned_loss=0.06218, over 7135.00 frames. ], tot_loss[loss=0.1652, simple_loss=0.239, pruned_loss=0.04572, over 1437749.39 frames. ], batch size: 98, lr: 1.09e-02, grad_scale: 16.0 +2023-03-20 23:19:28,099 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2334, 2.8910, 3.3332, 2.9236, 3.1090, 2.7641, 2.5045, 3.0116], + device='cuda:0'), covar=tensor([0.1137, 0.0587, 0.0888, 0.2792, 0.1034, 0.1283, 0.2606, 0.2559], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0042, 0.0038, 0.0039, 0.0036, 0.0034, 0.0051, 0.0039], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:19:33,771 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-20 23:19:38,391 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-20 23:19:39,344 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.392e+02 2.209e+02 2.625e+02 3.137e+02 7.215e+02, threshold=5.251e+02, percent-clipped=3.0 +2023-03-20 23:19:39,881 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-20 23:19:48,623 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2330, 1.0054, 1.5126, 1.4105, 1.3537, 1.4951, 1.0911, 1.3321], + device='cuda:0'), covar=tensor([0.1414, 0.1920, 0.0493, 0.0621, 0.1004, 0.1017, 0.0688, 0.2000], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0050, 0.0036, 0.0033, 0.0040, 0.0038, 0.0052, 0.0042], + device='cuda:0'), out_proj_covar=tensor([1.1028e-04, 1.2458e-04, 9.7413e-05, 9.5951e-05, 1.0847e-04, 1.0753e-04, + 1.3016e-04, 1.1430e-04], device='cuda:0') +2023-03-20 23:19:49,977 INFO [train.py:901] (0/2) Epoch 15, batch 1300, loss[loss=0.1383, simple_loss=0.2184, pruned_loss=0.02912, over 7331.00 frames. ], tot_loss[loss=0.1654, simple_loss=0.2391, pruned_loss=0.04587, over 1437625.62 frames. ], batch size: 44, lr: 1.09e-02, grad_scale: 16.0 +2023-03-20 23:19:53,079 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3372, 3.0688, 2.8673, 3.2465, 3.2364, 2.3521, 3.3776, 2.9632], + device='cuda:0'), covar=tensor([0.0387, 0.1645, 0.1758, 0.0765, 0.0913, 0.1270, 0.2283, 0.2059], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0041, 0.0047, 0.0042, 0.0041, 0.0043, 0.0043, 0.0040], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:19:55,493 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=40848.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:19:56,977 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=40851.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:20:03,500 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-20 23:20:04,139 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=40865.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:20:05,519 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-20 23:20:07,146 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9999, 3.2424, 2.1139, 3.7657, 2.5949, 2.9301, 1.7519, 2.0780], + device='cuda:0'), covar=tensor([0.0246, 0.0644, 0.1877, 0.0396, 0.0504, 0.0379, 0.2667, 0.1597], + device='cuda:0'), in_proj_covar=tensor([0.0233, 0.0228, 0.0300, 0.0237, 0.0244, 0.0244, 0.0267, 0.0285], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0004, 0.0004], + device='cuda:0') +2023-03-20 23:20:08,550 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=40874.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:20:08,979 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-20 23:20:14,545 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3904, 3.2078, 3.1479, 3.0094, 3.1487, 3.0943, 2.7781, 3.2710], + device='cuda:0'), covar=tensor([0.1157, 0.0436, 0.1507, 0.1609, 0.1200, 0.0937, 0.1926, 0.1880], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0042, 0.0037, 0.0038, 0.0036, 0.0033, 0.0051, 0.0038], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:20:14,950 INFO [train.py:901] (0/2) Epoch 15, batch 1350, loss[loss=0.1228, simple_loss=0.1903, pruned_loss=0.02766, over 7208.00 frames. ], tot_loss[loss=0.1658, simple_loss=0.2395, pruned_loss=0.04603, over 1436066.19 frames. ], batch size: 39, lr: 1.09e-02, grad_scale: 16.0 +2023-03-20 23:20:17,038 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6467, 3.3177, 3.3774, 3.3532, 3.2162, 3.2866, 3.5657, 3.1555], + device='cuda:0'), covar=tensor([0.0151, 0.0189, 0.0155, 0.0218, 0.0441, 0.0136, 0.0191, 0.0198], + device='cuda:0'), in_proj_covar=tensor([0.0064, 0.0072, 0.0072, 0.0063, 0.0127, 0.0081, 0.0076, 0.0077], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:20:20,055 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-20 23:20:21,260 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7323, 3.9932, 3.7595, 3.8861, 3.6919, 3.8864, 4.2677, 4.2752], + device='cuda:0'), covar=tensor([0.0233, 0.0154, 0.0216, 0.0192, 0.0330, 0.0275, 0.0226, 0.0161], + device='cuda:0'), in_proj_covar=tensor([0.0101, 0.0105, 0.0096, 0.0105, 0.0098, 0.0087, 0.0083, 0.0081], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:20:21,732 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=40899.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:20:22,316 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=40900.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:20:31,275 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.585e+02 2.254e+02 2.662e+02 3.076e+02 6.866e+02, threshold=5.325e+02, percent-clipped=2.0 +2023-03-20 23:20:33,056 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.37 vs. limit=2.0 +2023-03-20 23:20:33,826 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=40922.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:20:35,964 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=40926.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:20:36,907 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=40928.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:20:41,369 INFO [train.py:901] (0/2) Epoch 15, batch 1400, loss[loss=0.1685, simple_loss=0.2421, pruned_loss=0.04747, over 7233.00 frames. ], tot_loss[loss=0.166, simple_loss=0.2399, pruned_loss=0.046, over 1437199.34 frames. ], batch size: 45, lr: 1.09e-02, grad_scale: 16.0 +2023-03-20 23:20:46,648 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.05 vs. limit=2.0 +2023-03-20 23:20:52,345 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-20 23:20:53,536 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=40961.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 23:20:56,486 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3986, 3.7606, 4.0069, 3.9704, 3.6992, 3.9443, 4.0939, 3.6620], + device='cuda:0'), covar=tensor([0.0089, 0.0152, 0.0114, 0.0138, 0.0390, 0.0104, 0.0164, 0.0167], + device='cuda:0'), in_proj_covar=tensor([0.0063, 0.0072, 0.0072, 0.0062, 0.0127, 0.0081, 0.0076, 0.0077], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:21:00,883 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=40976.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:21:02,952 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3026, 1.9192, 2.2593, 1.8384, 1.9751, 1.3694, 1.7394, 1.5684], + device='cuda:0'), covar=tensor([0.0612, 0.0587, 0.0145, 0.0100, 0.0407, 0.0464, 0.0258, 0.0284], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0024, 0.0023, 0.0021, 0.0023, 0.0022, 0.0025, 0.0027], + device='cuda:0'), out_proj_covar=tensor([6.2531e-05, 6.0680e-05, 5.6140e-05, 5.0967e-05, 5.8454e-05, 5.6760e-05, + 5.9913e-05, 6.7584e-05], device='cuda:0') +2023-03-20 23:21:06,547 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=40986.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:21:06,954 INFO [train.py:901] (0/2) Epoch 15, batch 1450, loss[loss=0.181, simple_loss=0.2453, pruned_loss=0.05832, over 7270.00 frames. ], tot_loss[loss=0.1659, simple_loss=0.2397, pruned_loss=0.04608, over 1438213.32 frames. ], batch size: 52, lr: 1.09e-02, grad_scale: 16.0 +2023-03-20 23:21:12,489 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=40997.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:21:17,771 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-20 23:21:17,830 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=41007.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:21:17,909 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1257, 1.0841, 1.3311, 1.3755, 1.3515, 1.4039, 1.0475, 1.2465], + device='cuda:0'), covar=tensor([0.0914, 0.3290, 0.0682, 0.0882, 0.1097, 0.0850, 0.0846, 0.2117], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0051, 0.0037, 0.0034, 0.0040, 0.0038, 0.0053, 0.0043], + device='cuda:0'), out_proj_covar=tensor([1.1095e-04, 1.2734e-04, 1.0023e-04, 9.9547e-05, 1.0988e-04, 1.0880e-04, + 1.3316e-04, 1.1711e-04], device='cuda:0') +2023-03-20 23:21:19,074 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.48 vs. limit=2.0 +2023-03-20 23:21:21,519 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-03-20 23:21:22,719 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.574e+02 2.243e+02 2.788e+02 3.650e+02 1.062e+03, threshold=5.577e+02, percent-clipped=7.0 +2023-03-20 23:21:31,119 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=41034.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:21:32,504 INFO [train.py:901] (0/2) Epoch 15, batch 1500, loss[loss=0.1705, simple_loss=0.2386, pruned_loss=0.05124, over 7312.00 frames. ], tot_loss[loss=0.1663, simple_loss=0.2402, pruned_loss=0.04618, over 1440137.51 frames. ], batch size: 83, lr: 1.09e-02, grad_scale: 16.0 +2023-03-20 23:21:33,519 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-20 23:21:36,688 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=41045.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:21:43,284 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=41058.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:21:57,274 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-20 23:21:58,402 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9323, 2.6648, 2.7852, 2.9348, 2.9651, 2.8611, 2.3235, 3.1530], + device='cuda:0'), covar=tensor([0.1605, 0.0950, 0.1989, 0.1942, 0.1026, 0.1434, 0.2412, 0.1293], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0041, 0.0036, 0.0037, 0.0036, 0.0032, 0.0050, 0.0037], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:21:58,766 INFO [train.py:901] (0/2) Epoch 15, batch 1550, loss[loss=0.1528, simple_loss=0.2287, pruned_loss=0.03845, over 7283.00 frames. ], tot_loss[loss=0.1646, simple_loss=0.2388, pruned_loss=0.04524, over 1439674.86 frames. ], batch size: 57, lr: 1.09e-02, grad_scale: 16.0 +2023-03-20 23:22:06,429 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=41102.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:22:08,345 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=41106.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:22:13,692 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.596e+02 2.410e+02 2.850e+02 3.373e+02 6.504e+02, threshold=5.701e+02, percent-clipped=1.0 +2023-03-20 23:22:13,811 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6520, 3.9649, 3.7517, 3.9153, 3.6416, 3.9604, 4.2057, 4.2033], + device='cuda:0'), covar=tensor([0.0226, 0.0151, 0.0184, 0.0176, 0.0309, 0.0211, 0.0222, 0.0189], + device='cuda:0'), in_proj_covar=tensor([0.0101, 0.0104, 0.0095, 0.0105, 0.0098, 0.0086, 0.0083, 0.0080], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:22:23,629 INFO [train.py:901] (0/2) Epoch 15, batch 1600, loss[loss=0.1721, simple_loss=0.2546, pruned_loss=0.0448, over 7374.00 frames. ], tot_loss[loss=0.1644, simple_loss=0.2386, pruned_loss=0.04506, over 1439710.16 frames. ], batch size: 73, lr: 1.09e-02, grad_scale: 16.0 +2023-03-20 23:22:26,662 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-20 23:22:27,645 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-20 23:22:29,192 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=41148.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:22:30,713 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-20 23:22:32,837 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1562, 2.5060, 1.8839, 2.7519, 2.5783, 2.6782, 2.0245, 2.1838], + device='cuda:0'), covar=tensor([0.1521, 0.0712, 0.2944, 0.0473, 0.0076, 0.0047, 0.0149, 0.0130], + device='cuda:0'), in_proj_covar=tensor([0.0283, 0.0234, 0.0276, 0.0260, 0.0136, 0.0128, 0.0162, 0.0183], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:22:37,258 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=41163.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:22:40,157 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-20 23:22:45,186 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-20 23:22:49,253 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2473, 2.8761, 1.9720, 2.7905, 2.7961, 2.9446, 2.2047, 2.3671], + device='cuda:0'), covar=tensor([0.1597, 0.0637, 0.3020, 0.0717, 0.0075, 0.0059, 0.0142, 0.0169], + device='cuda:0'), in_proj_covar=tensor([0.0283, 0.0235, 0.0277, 0.0260, 0.0136, 0.0128, 0.0162, 0.0184], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:22:49,587 INFO [train.py:901] (0/2) Epoch 15, batch 1650, loss[loss=0.1279, simple_loss=0.2014, pruned_loss=0.0272, over 7010.00 frames. ], tot_loss[loss=0.1655, simple_loss=0.2395, pruned_loss=0.04578, over 1438949.30 frames. ], batch size: 35, lr: 1.09e-02, grad_scale: 16.0 +2023-03-20 23:22:52,708 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-20 23:22:54,249 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=41196.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:23:04,928 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.286e+02 2.271e+02 2.648e+02 3.290e+02 5.745e+02, threshold=5.296e+02, percent-clipped=1.0 +2023-03-20 23:23:07,087 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=41221.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:23:10,577 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 23:23:11,222 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5936, 2.0277, 2.1894, 1.6429, 1.9874, 1.4515, 1.7901, 1.2763], + device='cuda:0'), covar=tensor([0.0317, 0.0174, 0.0132, 0.0120, 0.0218, 0.0342, 0.0235, 0.0260], + device='cuda:0'), in_proj_covar=tensor([0.0023, 0.0023, 0.0022, 0.0020, 0.0022, 0.0021, 0.0024, 0.0025], + device='cuda:0'), out_proj_covar=tensor([5.9155e-05, 5.7216e-05, 5.3016e-05, 4.8948e-05, 5.6368e-05, 5.3784e-05, + 5.7595e-05, 6.3393e-05], device='cuda:0') +2023-03-20 23:23:15,220 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-20 23:23:15,708 INFO [train.py:901] (0/2) Epoch 15, batch 1700, loss[loss=0.1611, simple_loss=0.2336, pruned_loss=0.04425, over 7271.00 frames. ], tot_loss[loss=0.1658, simple_loss=0.2397, pruned_loss=0.04596, over 1440082.45 frames. ], batch size: 70, lr: 1.09e-02, grad_scale: 16.0 +2023-03-20 23:23:25,896 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=41256.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 23:23:26,298 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-20 23:23:34,422 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0939, 4.1321, 3.4378, 3.4118, 3.6416, 2.0829, 1.7842, 4.0089], + device='cuda:0'), covar=tensor([0.0047, 0.0050, 0.0127, 0.0083, 0.0093, 0.0653, 0.0620, 0.0084], + device='cuda:0'), in_proj_covar=tensor([0.0077, 0.0068, 0.0088, 0.0073, 0.0095, 0.0113, 0.0113, 0.0080], + device='cuda:0'), out_proj_covar=tensor([9.7871e-05, 9.8282e-05, 1.1627e-04, 1.0093e-04, 1.2451e-04, 1.4895e-04, + 1.4998e-04, 1.0356e-04], device='cuda:0') +2023-03-20 23:23:41,295 INFO [train.py:901] (0/2) Epoch 15, batch 1750, loss[loss=0.1789, simple_loss=0.2482, pruned_loss=0.05486, over 7289.00 frames. ], tot_loss[loss=0.1659, simple_loss=0.24, pruned_loss=0.04592, over 1442018.83 frames. ], batch size: 70, lr: 1.09e-02, grad_scale: 16.0 +2023-03-20 23:23:50,786 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-20 23:23:51,383 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=41307.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:23:51,777 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-20 23:23:56,294 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.319e+02 2.224e+02 2.586e+02 3.166e+02 5.142e+02, threshold=5.173e+02, percent-clipped=0.0 +2023-03-20 23:24:07,016 INFO [train.py:901] (0/2) Epoch 15, batch 1800, loss[loss=0.1648, simple_loss=0.24, pruned_loss=0.04483, over 7320.00 frames. ], tot_loss[loss=0.165, simple_loss=0.2392, pruned_loss=0.04538, over 1442502.78 frames. ], batch size: 49, lr: 1.09e-02, grad_scale: 16.0 +2023-03-20 23:24:12,630 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2647, 3.0028, 1.9763, 3.2450, 3.2567, 3.6816, 2.6765, 2.6729], + device='cuda:0'), covar=tensor([0.1671, 0.0687, 0.3327, 0.0733, 0.0131, 0.0071, 0.0203, 0.0195], + device='cuda:0'), in_proj_covar=tensor([0.0279, 0.0228, 0.0270, 0.0255, 0.0134, 0.0126, 0.0158, 0.0179], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:24:13,997 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-20 23:24:15,584 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=41353.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:24:15,682 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9139, 2.4350, 1.7998, 2.7220, 2.8630, 3.1211, 2.3585, 2.1653], + device='cuda:0'), covar=tensor([0.1616, 0.0669, 0.2949, 0.0581, 0.0080, 0.0057, 0.0128, 0.0181], + device='cuda:0'), in_proj_covar=tensor([0.0278, 0.0228, 0.0269, 0.0255, 0.0134, 0.0126, 0.0157, 0.0179], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:24:16,560 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=41355.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:24:26,493 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-20 23:24:32,447 INFO [train.py:901] (0/2) Epoch 15, batch 1850, loss[loss=0.1694, simple_loss=0.2448, pruned_loss=0.04696, over 7303.00 frames. ], tot_loss[loss=0.1642, simple_loss=0.2384, pruned_loss=0.04504, over 1440627.24 frames. ], batch size: 86, lr: 1.09e-02, grad_scale: 16.0 +2023-03-20 23:24:36,490 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-20 23:24:39,871 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=41401.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:24:48,355 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.233e+02 2.216e+02 2.442e+02 3.300e+02 6.714e+02, threshold=4.883e+02, percent-clipped=2.0 +2023-03-20 23:24:52,451 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-20 23:24:56,128 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0794, 2.6550, 2.7875, 2.9466, 2.1365, 2.3237, 3.0955, 2.3382], + device='cuda:0'), covar=tensor([0.0223, 0.0210, 0.0291, 0.0282, 0.0399, 0.0503, 0.0336, 0.0934], + device='cuda:0'), in_proj_covar=tensor([0.0302, 0.0298, 0.0247, 0.0314, 0.0302, 0.0293, 0.0308, 0.0288], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 23:24:58,916 INFO [train.py:901] (0/2) Epoch 15, batch 1900, loss[loss=0.1964, simple_loss=0.2616, pruned_loss=0.06557, over 7339.00 frames. ], tot_loss[loss=0.1653, simple_loss=0.2393, pruned_loss=0.0456, over 1441845.40 frames. ], batch size: 63, lr: 1.09e-02, grad_scale: 16.0 +2023-03-20 23:25:09,606 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=41458.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:25:16,708 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5674, 2.9544, 3.3933, 3.5877, 3.4984, 3.7236, 3.3373, 3.4240], + device='cuda:0'), covar=tensor([0.0023, 0.0099, 0.0035, 0.0027, 0.0032, 0.0021, 0.0051, 0.0049], + device='cuda:0'), in_proj_covar=tensor([0.0034, 0.0046, 0.0040, 0.0039, 0.0039, 0.0041, 0.0044, 0.0051], + device='cuda:0'), out_proj_covar=tensor([7.6193e-05, 1.2023e-04, 1.0187e-04, 9.1166e-05, 9.0347e-05, 9.5273e-05, + 1.1174e-04, 1.2136e-04], device='cuda:0') +2023-03-20 23:25:18,197 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-20 23:25:24,041 INFO [train.py:901] (0/2) Epoch 15, batch 1950, loss[loss=0.1428, simple_loss=0.2239, pruned_loss=0.03081, over 7283.00 frames. ], tot_loss[loss=0.164, simple_loss=0.2382, pruned_loss=0.04495, over 1439353.00 frames. ], batch size: 66, lr: 1.09e-02, grad_scale: 16.0 +2023-03-20 23:25:29,052 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-20 23:25:33,669 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-20 23:25:34,642 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-20 23:25:40,163 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.628e+02 2.202e+02 2.600e+02 3.272e+02 6.051e+02, threshold=5.201e+02, percent-clipped=3.0 +2023-03-20 23:25:42,276 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=41521.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:25:50,215 INFO [train.py:901] (0/2) Epoch 15, batch 2000, loss[loss=0.1577, simple_loss=0.2368, pruned_loss=0.03936, over 7274.00 frames. ], tot_loss[loss=0.1641, simple_loss=0.2381, pruned_loss=0.04499, over 1440467.33 frames. ], batch size: 70, lr: 1.08e-02, grad_scale: 16.0 +2023-03-20 23:25:51,731 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-20 23:25:55,776 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9103, 5.4449, 5.4581, 5.3768, 5.1407, 5.0095, 5.5390, 5.2512], + device='cuda:0'), covar=tensor([0.0421, 0.0350, 0.0369, 0.0441, 0.0327, 0.0311, 0.0322, 0.0480], + device='cuda:0'), in_proj_covar=tensor([0.0179, 0.0207, 0.0151, 0.0150, 0.0127, 0.0189, 0.0161, 0.0125], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:25:59,832 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=41556.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 23:26:02,219 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-20 23:26:06,242 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=41569.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:26:09,668 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-20 23:26:15,765 INFO [train.py:901] (0/2) Epoch 15, batch 2050, loss[loss=0.1782, simple_loss=0.2475, pruned_loss=0.05445, over 7360.00 frames. ], tot_loss[loss=0.163, simple_loss=0.2372, pruned_loss=0.04437, over 1440821.51 frames. ], batch size: 51, lr: 1.08e-02, grad_scale: 16.0 +2023-03-20 23:26:25,275 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=41604.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:26:29,417 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9107, 3.9673, 3.3940, 3.4154, 3.2469, 2.4832, 1.9456, 3.9873], + device='cuda:0'), covar=tensor([0.0034, 0.0053, 0.0085, 0.0053, 0.0081, 0.0360, 0.0460, 0.0036], + device='cuda:0'), in_proj_covar=tensor([0.0079, 0.0068, 0.0090, 0.0074, 0.0096, 0.0114, 0.0115, 0.0080], + device='cuda:0'), out_proj_covar=tensor([9.9899e-05, 9.9088e-05, 1.1936e-04, 1.0222e-04, 1.2649e-04, 1.5139e-04, + 1.5268e-04, 1.0348e-04], device='cuda:0') +2023-03-20 23:26:31,811 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.491e+02 2.163e+02 2.642e+02 3.130e+02 5.722e+02, threshold=5.285e+02, percent-clipped=2.0 +2023-03-20 23:26:41,041 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9773, 2.6193, 3.2054, 2.9155, 3.1196, 2.8327, 2.7165, 3.1235], + device='cuda:0'), covar=tensor([0.1241, 0.0476, 0.0810, 0.1642, 0.0849, 0.0989, 0.1323, 0.1099], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0042, 0.0036, 0.0038, 0.0035, 0.0034, 0.0050, 0.0038], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:26:41,961 INFO [train.py:901] (0/2) Epoch 15, batch 2100, loss[loss=0.1692, simple_loss=0.2496, pruned_loss=0.0444, over 7299.00 frames. ], tot_loss[loss=0.1634, simple_loss=0.2375, pruned_loss=0.04466, over 1440047.16 frames. ], batch size: 86, lr: 1.08e-02, grad_scale: 16.0 +2023-03-20 23:26:44,502 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-20 23:26:47,415 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-20 23:26:49,953 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=41653.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:26:50,506 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=41654.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:27:07,959 INFO [train.py:901] (0/2) Epoch 15, batch 2150, loss[loss=0.1712, simple_loss=0.2295, pruned_loss=0.05643, over 7306.00 frames. ], tot_loss[loss=0.1638, simple_loss=0.2377, pruned_loss=0.04493, over 1442180.20 frames. ], batch size: 80, lr: 1.08e-02, grad_scale: 16.0 +2023-03-20 23:27:15,069 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=41701.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:27:15,136 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=41701.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:27:22,224 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=41715.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 23:27:23,073 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.454e+02 2.084e+02 2.496e+02 3.208e+02 6.072e+02, threshold=4.993e+02, percent-clipped=2.0 +2023-03-20 23:27:24,644 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8403, 4.1239, 3.9867, 4.0691, 3.9019, 3.9049, 4.2051, 4.2577], + device='cuda:0'), covar=tensor([0.0327, 0.0248, 0.0228, 0.0340, 0.0381, 0.0434, 0.0415, 0.0371], + device='cuda:0'), in_proj_covar=tensor([0.0102, 0.0102, 0.0096, 0.0105, 0.0097, 0.0086, 0.0083, 0.0081], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:27:33,051 INFO [train.py:901] (0/2) Epoch 15, batch 2200, loss[loss=0.1465, simple_loss=0.2181, pruned_loss=0.03745, over 7269.00 frames. ], tot_loss[loss=0.1633, simple_loss=0.2374, pruned_loss=0.04463, over 1442610.95 frames. ], batch size: 47, lr: 1.08e-02, grad_scale: 16.0 +2023-03-20 23:27:34,151 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.0513, 4.6006, 4.5503, 5.0301, 5.0049, 5.0330, 4.3291, 4.6755], + device='cuda:0'), covar=tensor([0.0822, 0.2459, 0.2110, 0.1101, 0.0732, 0.1206, 0.0853, 0.1072], + device='cuda:0'), in_proj_covar=tensor([0.0196, 0.0315, 0.0254, 0.0243, 0.0184, 0.0304, 0.0171, 0.0219], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:27:35,117 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-20 23:27:39,248 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=41749.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:27:43,772 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=41758.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:27:52,406 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7215, 3.0639, 3.6353, 3.7855, 3.7531, 3.8102, 3.6932, 3.5124], + device='cuda:0'), covar=tensor([0.0030, 0.0110, 0.0037, 0.0033, 0.0033, 0.0031, 0.0045, 0.0054], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0047, 0.0042, 0.0039, 0.0040, 0.0042, 0.0045, 0.0051], + device='cuda:0'), out_proj_covar=tensor([7.7027e-05, 1.2261e-04, 1.0507e-04, 9.1270e-05, 9.1510e-05, 9.7549e-05, + 1.1417e-04, 1.2300e-04], device='cuda:0') +2023-03-20 23:27:59,463 INFO [train.py:901] (0/2) Epoch 15, batch 2250, loss[loss=0.1553, simple_loss=0.228, pruned_loss=0.04125, over 7323.00 frames. ], tot_loss[loss=0.1643, simple_loss=0.2381, pruned_loss=0.04527, over 1442121.14 frames. ], batch size: 59, lr: 1.08e-02, grad_scale: 16.0 +2023-03-20 23:28:09,402 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=41806.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:28:10,379 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-20 23:28:10,884 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-20 23:28:14,907 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.381e+02 2.118e+02 2.612e+02 3.099e+02 6.115e+02, threshold=5.224e+02, percent-clipped=1.0 +2023-03-20 23:28:23,435 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-20 23:28:24,922 INFO [train.py:901] (0/2) Epoch 15, batch 2300, loss[loss=0.1364, simple_loss=0.2094, pruned_loss=0.03171, over 7127.00 frames. ], tot_loss[loss=0.1638, simple_loss=0.2378, pruned_loss=0.04491, over 1443547.41 frames. ], batch size: 41, lr: 1.08e-02, grad_scale: 16.0 +2023-03-20 23:28:36,138 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2607, 1.1449, 1.6834, 1.5104, 1.4201, 1.5643, 1.5901, 1.5474], + device='cuda:0'), covar=tensor([0.2858, 0.2084, 0.0518, 0.1235, 0.1518, 0.2042, 0.1303, 0.1818], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0052, 0.0036, 0.0035, 0.0039, 0.0037, 0.0055, 0.0041], + device='cuda:0'), out_proj_covar=tensor([1.1341e-04, 1.2965e-04, 9.9977e-05, 1.0220e-04, 1.0910e-04, 1.0909e-04, + 1.3707e-04, 1.1383e-04], device='cuda:0') +2023-03-20 23:28:51,147 INFO [train.py:901] (0/2) Epoch 15, batch 2350, loss[loss=0.188, simple_loss=0.2601, pruned_loss=0.05795, over 7303.00 frames. ], tot_loss[loss=0.1646, simple_loss=0.2387, pruned_loss=0.04521, over 1443124.00 frames. ], batch size: 83, lr: 1.08e-02, grad_scale: 16.0 +2023-03-20 23:29:01,042 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=41906.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:29:06,450 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.373e+02 2.041e+02 2.774e+02 3.527e+02 1.027e+03, threshold=5.547e+02, percent-clipped=7.0 +2023-03-20 23:29:08,977 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-20 23:29:16,011 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-20 23:29:16,500 INFO [train.py:901] (0/2) Epoch 15, batch 2400, loss[loss=0.1344, simple_loss=0.2103, pruned_loss=0.02926, over 7141.00 frames. ], tot_loss[loss=0.1636, simple_loss=0.2378, pruned_loss=0.04473, over 1441901.71 frames. ], batch size: 41, lr: 1.08e-02, grad_scale: 16.0 +2023-03-20 23:29:22,337 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4736, 1.7465, 2.0943, 1.5958, 1.9096, 1.5864, 1.6536, 1.4270], + device='cuda:0'), covar=tensor([0.0312, 0.0246, 0.0064, 0.0107, 0.0319, 0.0218, 0.0241, 0.0140], + device='cuda:0'), in_proj_covar=tensor([0.0024, 0.0024, 0.0022, 0.0021, 0.0022, 0.0021, 0.0024, 0.0025], + device='cuda:0'), out_proj_covar=tensor([6.1078e-05, 5.9525e-05, 5.2889e-05, 4.9826e-05, 5.6511e-05, 5.4546e-05, + 5.8691e-05, 6.3260e-05], device='cuda:0') +2023-03-20 23:29:22,845 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=41948.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 23:29:27,866 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-20 23:29:30,669 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.45 vs. limit=2.0 +2023-03-20 23:29:30,906 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-20 23:29:33,012 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=41967.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:29:42,901 INFO [train.py:901] (0/2) Epoch 15, batch 2450, loss[loss=0.1552, simple_loss=0.2284, pruned_loss=0.04102, over 7326.00 frames. ], tot_loss[loss=0.1626, simple_loss=0.2369, pruned_loss=0.04418, over 1440781.21 frames. ], batch size: 44, lr: 1.08e-02, grad_scale: 16.0 +2023-03-20 23:29:54,512 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=42009.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 23:29:54,925 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=42010.0, num_to_drop=1, layers_to_drop={3} +2023-03-20 23:29:58,414 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.325e+02 2.135e+02 2.470e+02 3.053e+02 4.844e+02, threshold=4.939e+02, percent-clipped=0.0 +2023-03-20 23:29:58,440 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-20 23:29:58,500 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8697, 4.4108, 4.4003, 4.9145, 4.8416, 4.8572, 4.2134, 4.5329], + device='cuda:0'), covar=tensor([0.0882, 0.2546, 0.2251, 0.1016, 0.0917, 0.1476, 0.0868, 0.1159], + device='cuda:0'), in_proj_covar=tensor([0.0195, 0.0313, 0.0250, 0.0245, 0.0183, 0.0303, 0.0169, 0.0220], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:30:09,677 INFO [train.py:901] (0/2) Epoch 15, batch 2500, loss[loss=0.1978, simple_loss=0.2719, pruned_loss=0.06182, over 7336.00 frames. ], tot_loss[loss=0.1634, simple_loss=0.2376, pruned_loss=0.04461, over 1442581.27 frames. ], batch size: 61, lr: 1.08e-02, grad_scale: 16.0 +2023-03-20 23:30:23,687 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-20 23:30:34,627 INFO [train.py:901] (0/2) Epoch 15, batch 2550, loss[loss=0.1592, simple_loss=0.2351, pruned_loss=0.04165, over 7271.00 frames. ], tot_loss[loss=0.1639, simple_loss=0.238, pruned_loss=0.04489, over 1443912.13 frames. ], batch size: 70, lr: 1.08e-02, grad_scale: 16.0 +2023-03-20 23:30:49,755 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.459e+02 2.206e+02 2.616e+02 3.137e+02 6.441e+02, threshold=5.231e+02, percent-clipped=2.0 +2023-03-20 23:31:00,961 INFO [train.py:901] (0/2) Epoch 15, batch 2600, loss[loss=0.1779, simple_loss=0.25, pruned_loss=0.0529, over 7252.00 frames. ], tot_loss[loss=0.1633, simple_loss=0.2376, pruned_loss=0.04453, over 1444343.00 frames. ], batch size: 55, lr: 1.08e-02, grad_scale: 16.0 +2023-03-20 23:31:25,848 INFO [train.py:901] (0/2) Epoch 15, batch 2650, loss[loss=0.157, simple_loss=0.2378, pruned_loss=0.03817, over 7254.00 frames. ], tot_loss[loss=0.1629, simple_loss=0.2372, pruned_loss=0.04434, over 1441224.53 frames. ], batch size: 55, lr: 1.08e-02, grad_scale: 16.0 +2023-03-20 23:31:41,017 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.465e+02 2.117e+02 2.438e+02 2.904e+02 5.048e+02, threshold=4.876e+02, percent-clipped=0.0 +2023-03-20 23:31:48,227 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.02 vs. limit=2.0 +2023-03-20 23:31:50,885 INFO [train.py:901] (0/2) Epoch 15, batch 2700, loss[loss=0.185, simple_loss=0.2531, pruned_loss=0.05851, over 7331.00 frames. ], tot_loss[loss=0.1628, simple_loss=0.2371, pruned_loss=0.04426, over 1440626.46 frames. ], batch size: 61, lr: 1.08e-02, grad_scale: 16.0 +2023-03-20 23:32:03,238 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=42262.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:32:10,058 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.05 vs. limit=2.0 +2023-03-20 23:32:15,708 INFO [train.py:901] (0/2) Epoch 15, batch 2750, loss[loss=0.1551, simple_loss=0.2319, pruned_loss=0.0391, over 7319.00 frames. ], tot_loss[loss=0.1628, simple_loss=0.237, pruned_loss=0.04426, over 1441524.54 frames. ], batch size: 80, lr: 1.08e-02, grad_scale: 16.0 +2023-03-20 23:32:17,109 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.70 vs. limit=2.0 +2023-03-20 23:32:24,267 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=42304.0, num_to_drop=1, layers_to_drop={0} +2023-03-20 23:32:27,267 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=42310.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 23:32:30,642 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.441e+02 2.237e+02 2.733e+02 3.339e+02 6.919e+02, threshold=5.467e+02, percent-clipped=1.0 +2023-03-20 23:32:40,404 INFO [train.py:901] (0/2) Epoch 15, batch 2800, loss[loss=0.1569, simple_loss=0.2386, pruned_loss=0.03759, over 7293.00 frames. ], tot_loss[loss=0.1635, simple_loss=0.2378, pruned_loss=0.04458, over 1439886.01 frames. ], batch size: 83, lr: 1.07e-02, grad_scale: 16.0 +2023-03-20 23:32:50,601 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=42358.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:32:53,000 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-15.pt +2023-03-20 23:33:09,534 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-20 23:33:09,589 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-20 23:33:09,612 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-20 23:33:09,942 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-20 23:33:10,116 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-20 23:33:10,136 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-20 23:33:10,147 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-20 23:33:10,246 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-20 23:33:10,320 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-20 23:33:10,336 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-20 23:33:10,352 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-20 23:33:10,514 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-20 23:33:10,520 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-20 23:33:10,528 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-20 23:33:11,014 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-20 23:33:11,017 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-20 23:33:11,076 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-20 23:33:13,492 INFO [train.py:901] (0/2) Epoch 16, batch 0, loss[loss=0.1701, simple_loss=0.2471, pruned_loss=0.04651, over 7316.00 frames. ], tot_loss[loss=0.1701, simple_loss=0.2471, pruned_loss=0.04651, over 7316.00 frames. ], batch size: 80, lr: 1.04e-02, grad_scale: 32.0 +2023-03-20 23:33:13,494 INFO [train.py:926] (0/2) Computing validation loss +2023-03-20 23:33:25,455 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0331, 3.5624, 3.6861, 3.7065, 3.7242, 3.6992, 3.8671, 3.4533], + device='cuda:0'), covar=tensor([0.0084, 0.0153, 0.0118, 0.0115, 0.0314, 0.0099, 0.0122, 0.0166], + device='cuda:0'), in_proj_covar=tensor([0.0066, 0.0073, 0.0073, 0.0063, 0.0127, 0.0082, 0.0079, 0.0078], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:33:36,912 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4602, 1.7888, 2.1530, 1.6952, 2.0909, 1.6328, 1.8289, 1.4921], + device='cuda:0'), covar=tensor([0.0363, 0.0308, 0.0119, 0.0191, 0.0434, 0.0396, 0.0197, 0.0215], + device='cuda:0'), in_proj_covar=tensor([0.0023, 0.0023, 0.0022, 0.0020, 0.0020, 0.0020, 0.0023, 0.0023], + device='cuda:0'), out_proj_covar=tensor([5.8221e-05, 5.7401e-05, 5.1984e-05, 4.8366e-05, 5.3618e-05, 5.2072e-05, + 5.5183e-05, 5.9053e-05], device='cuda:0') +2023-03-20 23:33:39,696 INFO [train.py:935] (0/2) Epoch 16, validation: loss=0.1677, simple_loss=0.2552, pruned_loss=0.04012, over 1622729.00 frames. +2023-03-20 23:33:39,697 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-20 23:33:46,800 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-20 23:33:49,475 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=42379.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:33:50,499 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8258, 2.5326, 2.4392, 3.9585, 1.6261, 3.7729, 1.3753, 3.1038], + device='cuda:0'), covar=tensor([0.0060, 0.0835, 0.1348, 0.0063, 0.3753, 0.0089, 0.1049, 0.0214], + device='cuda:0'), in_proj_covar=tensor([0.0161, 0.0254, 0.0293, 0.0154, 0.0282, 0.0166, 0.0267, 0.0211], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 23:33:57,850 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-20 23:34:05,291 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-20 23:34:05,806 INFO [train.py:901] (0/2) Epoch 16, batch 50, loss[loss=0.1735, simple_loss=0.2494, pruned_loss=0.04885, over 7277.00 frames. ], tot_loss[loss=0.162, simple_loss=0.2359, pruned_loss=0.04401, over 322210.19 frames. ], batch size: 57, lr: 1.04e-02, grad_scale: 16.0 +2023-03-20 23:34:07,361 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-20 23:34:09,363 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.463e+02 2.229e+02 2.517e+02 3.017e+02 7.171e+02, threshold=5.034e+02, percent-clipped=2.0 +2023-03-20 23:34:10,398 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-20 23:34:20,022 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2668, 1.3999, 1.3528, 1.3576, 1.3765, 1.2406, 1.2335, 0.9452], + device='cuda:0'), covar=tensor([0.0189, 0.0102, 0.0174, 0.0099, 0.0097, 0.0131, 0.0116, 0.0104], + device='cuda:0'), in_proj_covar=tensor([0.0024, 0.0022, 0.0022, 0.0022, 0.0023, 0.0022, 0.0024, 0.0029], + device='cuda:0'), out_proj_covar=tensor([2.8535e-05, 2.4950e-05, 2.5634e-05, 2.5203e-05, 2.7471e-05, 2.4853e-05, + 2.7304e-05, 3.4360e-05], device='cuda:0') +2023-03-20 23:34:20,536 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=42440.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:34:28,155 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-20 23:34:28,632 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-20 23:34:31,682 INFO [train.py:901] (0/2) Epoch 16, batch 100, loss[loss=0.1659, simple_loss=0.2491, pruned_loss=0.04133, over 7276.00 frames. ], tot_loss[loss=0.1602, simple_loss=0.2344, pruned_loss=0.04305, over 569523.22 frames. ], batch size: 57, lr: 1.04e-02, grad_scale: 16.0 +2023-03-20 23:34:44,760 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7467, 2.6123, 1.9595, 3.0454, 2.0090, 2.6905, 1.5533, 1.7055], + device='cuda:0'), covar=tensor([0.0247, 0.0581, 0.1972, 0.0418, 0.0316, 0.0455, 0.2761, 0.1636], + device='cuda:0'), in_proj_covar=tensor([0.0236, 0.0233, 0.0301, 0.0241, 0.0243, 0.0243, 0.0263, 0.0282], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0004, 0.0004], + device='cuda:0') +2023-03-20 23:34:55,135 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=42506.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:34:56,424 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.84 vs. limit=2.0 +2023-03-20 23:34:56,678 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8254, 2.1386, 1.5909, 2.9316, 2.2785, 2.7633, 2.0637, 2.6066], + device='cuda:0'), covar=tensor([0.2028, 0.1074, 0.3625, 0.0726, 0.0103, 0.0115, 0.0157, 0.0194], + device='cuda:0'), in_proj_covar=tensor([0.0277, 0.0229, 0.0272, 0.0263, 0.0136, 0.0133, 0.0161, 0.0182], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:34:57,507 INFO [train.py:901] (0/2) Epoch 16, batch 150, loss[loss=0.1887, simple_loss=0.2577, pruned_loss=0.05982, over 7329.00 frames. ], tot_loss[loss=0.163, simple_loss=0.237, pruned_loss=0.04449, over 765237.12 frames. ], batch size: 75, lr: 1.04e-02, grad_scale: 16.0 +2023-03-20 23:35:01,031 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.394e+02 2.163e+02 2.502e+02 3.239e+02 5.751e+02, threshold=5.003e+02, percent-clipped=1.0 +2023-03-20 23:35:23,801 INFO [train.py:901] (0/2) Epoch 16, batch 200, loss[loss=0.1593, simple_loss=0.2404, pruned_loss=0.03906, over 7309.00 frames. ], tot_loss[loss=0.1639, simple_loss=0.2383, pruned_loss=0.04474, over 916228.25 frames. ], batch size: 80, lr: 1.04e-02, grad_scale: 16.0 +2023-03-20 23:35:24,425 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=42562.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:35:26,914 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=42567.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:35:29,829 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-20 23:35:34,892 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-20 23:35:40,879 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-20 23:35:45,708 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=42604.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 23:35:48,704 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=42610.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:35:49,162 INFO [train.py:901] (0/2) Epoch 16, batch 250, loss[loss=0.1728, simple_loss=0.2433, pruned_loss=0.0511, over 7293.00 frames. ], tot_loss[loss=0.1635, simple_loss=0.2383, pruned_loss=0.04434, over 1035751.67 frames. ], batch size: 49, lr: 1.04e-02, grad_scale: 16.0 +2023-03-20 23:35:52,735 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.263e+02 2.075e+02 2.482e+02 2.956e+02 5.105e+02, threshold=4.965e+02, percent-clipped=1.0 +2023-03-20 23:35:54,770 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-20 23:36:09,596 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=42649.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:36:11,025 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=42652.0, num_to_drop=1, layers_to_drop={1} +2023-03-20 23:36:15,470 INFO [train.py:901] (0/2) Epoch 16, batch 300, loss[loss=0.1555, simple_loss=0.2268, pruned_loss=0.04215, over 7213.00 frames. ], tot_loss[loss=0.1632, simple_loss=0.2377, pruned_loss=0.04437, over 1126068.21 frames. ], batch size: 45, lr: 1.04e-02, grad_scale: 16.0 +2023-03-20 23:36:15,988 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-20 23:36:24,934 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-20 23:36:40,090 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=42710.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:36:40,458 INFO [train.py:901] (0/2) Epoch 16, batch 350, loss[loss=0.1313, simple_loss=0.1917, pruned_loss=0.03543, over 6492.00 frames. ], tot_loss[loss=0.1627, simple_loss=0.2371, pruned_loss=0.0442, over 1197111.75 frames. ], batch size: 28, lr: 1.04e-02, grad_scale: 16.0 +2023-03-20 23:36:44,533 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.472e+02 2.340e+02 2.627e+02 3.104e+02 4.812e+02, threshold=5.255e+02, percent-clipped=0.0 +2023-03-20 23:36:53,835 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=42735.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:36:56,966 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4073, 1.7196, 2.0055, 1.6621, 1.9139, 1.8001, 1.6289, 1.3185], + device='cuda:0'), covar=tensor([0.0394, 0.0254, 0.0089, 0.0099, 0.0325, 0.0161, 0.0229, 0.0160], + device='cuda:0'), in_proj_covar=tensor([0.0023, 0.0022, 0.0021, 0.0020, 0.0020, 0.0019, 0.0022, 0.0023], + device='cuda:0'), out_proj_covar=tensor([5.8420e-05, 5.6194e-05, 5.0618e-05, 4.7732e-05, 5.2250e-05, 5.0008e-05, + 5.4345e-05, 5.7898e-05], device='cuda:0') +2023-03-20 23:36:58,878 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-20 23:36:59,513 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0729, 2.4828, 1.8671, 2.7493, 2.5819, 2.7386, 2.4003, 2.5882], + device='cuda:0'), covar=tensor([0.1615, 0.0693, 0.2819, 0.0614, 0.0096, 0.0051, 0.0172, 0.0195], + device='cuda:0'), in_proj_covar=tensor([0.0284, 0.0232, 0.0275, 0.0267, 0.0141, 0.0135, 0.0164, 0.0187], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:37:06,935 INFO [train.py:901] (0/2) Epoch 16, batch 400, loss[loss=0.1612, simple_loss=0.2387, pruned_loss=0.04192, over 7358.00 frames. ], tot_loss[loss=0.1629, simple_loss=0.2372, pruned_loss=0.04428, over 1251116.29 frames. ], batch size: 63, lr: 1.04e-02, grad_scale: 16.0 +2023-03-20 23:37:22,248 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5270, 4.2279, 4.1652, 4.6798, 4.6440, 4.6280, 3.7933, 4.2151], + device='cuda:0'), covar=tensor([0.0756, 0.2439, 0.2170, 0.1010, 0.0716, 0.1275, 0.1069, 0.1128], + device='cuda:0'), in_proj_covar=tensor([0.0191, 0.0311, 0.0244, 0.0241, 0.0180, 0.0295, 0.0171, 0.0218], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:37:33,188 INFO [train.py:901] (0/2) Epoch 16, batch 450, loss[loss=0.1697, simple_loss=0.2491, pruned_loss=0.04516, over 7231.00 frames. ], tot_loss[loss=0.162, simple_loss=0.2362, pruned_loss=0.04392, over 1289805.03 frames. ], batch size: 93, lr: 1.04e-02, grad_scale: 16.0 +2023-03-20 23:37:37,217 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.512e+02 2.066e+02 2.412e+02 3.102e+02 6.597e+02, threshold=4.823e+02, percent-clipped=2.0 +2023-03-20 23:37:40,765 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-20 23:37:40,778 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-20 23:37:57,544 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=42859.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:37:58,450 INFO [train.py:901] (0/2) Epoch 16, batch 500, loss[loss=0.1469, simple_loss=0.2222, pruned_loss=0.0358, over 7334.00 frames. ], tot_loss[loss=0.1621, simple_loss=0.2365, pruned_loss=0.04385, over 1326721.75 frames. ], batch size: 75, lr: 1.04e-02, grad_scale: 8.0 +2023-03-20 23:37:59,075 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=42862.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:38:00,180 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9266, 2.9332, 2.7684, 2.9394, 2.2615, 2.4972, 3.0192, 2.4150], + device='cuda:0'), covar=tensor([0.0255, 0.0370, 0.0286, 0.0345, 0.0379, 0.0522, 0.0375, 0.1011], + device='cuda:0'), in_proj_covar=tensor([0.0307, 0.0310, 0.0253, 0.0328, 0.0309, 0.0301, 0.0317, 0.0294], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 23:38:07,796 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=42879.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:38:13,816 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-20 23:38:15,300 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-20 23:38:16,275 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-20 23:38:17,849 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1504, 4.4491, 4.1510, 4.2863, 3.9852, 4.4609, 4.6769, 4.7261], + device='cuda:0'), covar=tensor([0.0196, 0.0120, 0.0190, 0.0150, 0.0289, 0.0152, 0.0231, 0.0153], + device='cuda:0'), in_proj_covar=tensor([0.0103, 0.0102, 0.0098, 0.0105, 0.0097, 0.0088, 0.0086, 0.0084], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:38:18,284 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-20 23:38:22,996 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-20 23:38:24,947 INFO [train.py:901] (0/2) Epoch 16, batch 550, loss[loss=0.1828, simple_loss=0.2535, pruned_loss=0.05607, over 7347.00 frames. ], tot_loss[loss=0.1624, simple_loss=0.2367, pruned_loss=0.04407, over 1351526.63 frames. ], batch size: 63, lr: 1.04e-02, grad_scale: 8.0 +2023-03-20 23:38:28,879 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.372e+02 2.137e+02 2.590e+02 3.192e+02 4.525e+02, threshold=5.179e+02, percent-clipped=0.0 +2023-03-20 23:38:29,561 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=42920.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:38:30,300 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.86 vs. limit=5.0 +2023-03-20 23:38:32,295 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.00 vs. limit=2.0 +2023-03-20 23:38:33,968 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-20 23:38:39,696 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=42940.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:38:42,598 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-20 23:38:45,617 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-20 23:38:50,135 INFO [train.py:901] (0/2) Epoch 16, batch 600, loss[loss=0.1742, simple_loss=0.2469, pruned_loss=0.05078, over 7275.00 frames. ], tot_loss[loss=0.1614, simple_loss=0.2358, pruned_loss=0.04354, over 1372234.44 frames. ], batch size: 57, lr: 1.04e-02, grad_scale: 8.0 +2023-03-20 23:38:53,029 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-20 23:39:09,162 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-20 23:39:14,034 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=43005.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:39:17,008 INFO [train.py:901] (0/2) Epoch 16, batch 650, loss[loss=0.1778, simple_loss=0.2548, pruned_loss=0.05042, over 7274.00 frames. ], tot_loss[loss=0.1628, simple_loss=0.2372, pruned_loss=0.04421, over 1386648.94 frames. ], batch size: 57, lr: 1.04e-02, grad_scale: 8.0 +2023-03-20 23:39:18,029 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-20 23:39:21,006 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.512e+02 2.149e+02 2.474e+02 3.215e+02 8.384e+02, threshold=4.948e+02, percent-clipped=2.0 +2023-03-20 23:39:29,238 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=43035.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:39:35,691 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-20 23:39:41,496 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9865, 4.2585, 4.0453, 4.0750, 3.9921, 4.2060, 4.5695, 4.5453], + device='cuda:0'), covar=tensor([0.0208, 0.0133, 0.0171, 0.0170, 0.0229, 0.0182, 0.0165, 0.0159], + device='cuda:0'), in_proj_covar=tensor([0.0104, 0.0104, 0.0099, 0.0105, 0.0098, 0.0089, 0.0086, 0.0084], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:39:42,907 INFO [train.py:901] (0/2) Epoch 16, batch 700, loss[loss=0.1553, simple_loss=0.2334, pruned_loss=0.03856, over 7261.00 frames. ], tot_loss[loss=0.1627, simple_loss=0.2368, pruned_loss=0.0443, over 1397530.44 frames. ], batch size: 64, lr: 1.03e-02, grad_scale: 8.0 +2023-03-20 23:39:44,367 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-20 23:39:54,680 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=43083.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:40:08,153 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-20 23:40:08,624 INFO [train.py:901] (0/2) Epoch 16, batch 750, loss[loss=0.1749, simple_loss=0.2508, pruned_loss=0.04949, over 7129.00 frames. ], tot_loss[loss=0.1636, simple_loss=0.2378, pruned_loss=0.04467, over 1407457.04 frames. ], batch size: 98, lr: 1.03e-02, grad_scale: 8.0 +2023-03-20 23:40:08,632 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-20 23:40:09,552 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-20 23:40:12,702 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.251e+02 2.316e+02 2.793e+02 3.292e+02 6.692e+02, threshold=5.586e+02, percent-clipped=4.0 +2023-03-20 23:40:22,722 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-20 23:40:27,217 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-20 23:40:32,695 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-20 23:40:34,277 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-20 23:40:34,748 INFO [train.py:901] (0/2) Epoch 16, batch 800, loss[loss=0.1512, simple_loss=0.2319, pruned_loss=0.03525, over 7246.00 frames. ], tot_loss[loss=0.1628, simple_loss=0.2373, pruned_loss=0.04415, over 1415476.04 frames. ], batch size: 89, lr: 1.03e-02, grad_scale: 8.0 +2023-03-20 23:40:35,373 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=43162.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:40:35,383 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7123, 2.4027, 2.8081, 2.3859, 2.8192, 2.6120, 2.3886, 2.7217], + device='cuda:0'), covar=tensor([0.1876, 0.0506, 0.1115, 0.2103, 0.0956, 0.0973, 0.2176, 0.1183], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0047, 0.0038, 0.0039, 0.0037, 0.0034, 0.0053, 0.0041], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:40:41,780 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.2494, 4.7604, 4.7240, 5.2068, 5.2064, 5.2267, 4.6045, 4.8405], + device='cuda:0'), covar=tensor([0.0669, 0.2505, 0.2154, 0.1003, 0.0702, 0.1111, 0.0612, 0.0900], + device='cuda:0'), in_proj_covar=tensor([0.0195, 0.0312, 0.0250, 0.0240, 0.0183, 0.0298, 0.0173, 0.0218], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:40:45,278 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-20 23:40:59,845 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=43210.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:41:00,282 INFO [train.py:901] (0/2) Epoch 16, batch 850, loss[loss=0.1599, simple_loss=0.2405, pruned_loss=0.03964, over 7362.00 frames. ], tot_loss[loss=0.1622, simple_loss=0.2371, pruned_loss=0.04362, over 1423340.51 frames. ], batch size: 73, lr: 1.03e-02, grad_scale: 8.0 +2023-03-20 23:41:02,354 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=43215.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:41:04,281 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.587e+02 2.177e+02 2.571e+02 2.944e+02 4.753e+02, threshold=5.142e+02, percent-clipped=0.0 +2023-03-20 23:41:05,840 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-20 23:41:05,849 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-20 23:41:12,011 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-20 23:41:13,086 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=43235.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:41:15,568 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-20 23:41:21,355 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1461, 1.2845, 1.2103, 1.2269, 1.2872, 1.1948, 1.2260, 0.7498], + device='cuda:0'), covar=tensor([0.0163, 0.0127, 0.0132, 0.0108, 0.0171, 0.0099, 0.0112, 0.0158], + device='cuda:0'), in_proj_covar=tensor([0.0024, 0.0022, 0.0021, 0.0022, 0.0022, 0.0022, 0.0023, 0.0029], + device='cuda:0'), out_proj_covar=tensor([2.7953e-05, 2.5018e-05, 2.5386e-05, 2.5143e-05, 2.7104e-05, 2.4773e-05, + 2.6563e-05, 3.3864e-05], device='cuda:0') +2023-03-20 23:41:26,687 INFO [train.py:901] (0/2) Epoch 16, batch 900, loss[loss=0.1591, simple_loss=0.2289, pruned_loss=0.04466, over 7323.00 frames. ], tot_loss[loss=0.1616, simple_loss=0.2363, pruned_loss=0.04344, over 1425172.56 frames. ], batch size: 59, lr: 1.03e-02, grad_scale: 8.0 +2023-03-20 23:41:36,834 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1864, 1.3670, 1.2318, 1.2664, 1.3329, 1.0805, 1.3311, 0.7297], + device='cuda:0'), covar=tensor([0.0133, 0.0090, 0.0096, 0.0103, 0.0148, 0.0180, 0.0112, 0.0118], + device='cuda:0'), in_proj_covar=tensor([0.0023, 0.0022, 0.0021, 0.0022, 0.0022, 0.0022, 0.0023, 0.0028], + device='cuda:0'), out_proj_covar=tensor([2.7386e-05, 2.4723e-05, 2.4752e-05, 2.4796e-05, 2.6712e-05, 2.4331e-05, + 2.6257e-05, 3.3279e-05], device='cuda:0') +2023-03-20 23:41:49,038 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=43305.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:41:52,503 INFO [train.py:901] (0/2) Epoch 16, batch 950, loss[loss=0.1666, simple_loss=0.2434, pruned_loss=0.04494, over 7288.00 frames. ], tot_loss[loss=0.1619, simple_loss=0.2365, pruned_loss=0.0437, over 1429073.73 frames. ], batch size: 86, lr: 1.03e-02, grad_scale: 8.0 +2023-03-20 23:41:52,531 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-20 23:41:56,556 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.675e+02 2.071e+02 2.459e+02 3.042e+02 5.230e+02, threshold=4.919e+02, percent-clipped=1.0 +2023-03-20 23:42:14,267 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=43353.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:42:15,240 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-20 23:42:18,323 INFO [train.py:901] (0/2) Epoch 16, batch 1000, loss[loss=0.1601, simple_loss=0.2347, pruned_loss=0.04275, over 7303.00 frames. ], tot_loss[loss=0.1615, simple_loss=0.2363, pruned_loss=0.04333, over 1429771.69 frames. ], batch size: 80, lr: 1.03e-02, grad_scale: 8.0 +2023-03-20 23:42:30,165 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1350, 0.7391, 1.4970, 1.4550, 1.3586, 1.5732, 1.3088, 1.3057], + device='cuda:0'), covar=tensor([0.1135, 0.2534, 0.1350, 0.1046, 0.1442, 0.0995, 0.0866, 0.2098], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0049, 0.0036, 0.0035, 0.0038, 0.0038, 0.0054, 0.0040], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 23:42:36,574 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-20 23:42:44,383 INFO [train.py:901] (0/2) Epoch 16, batch 1050, loss[loss=0.1647, simple_loss=0.2343, pruned_loss=0.04751, over 7351.00 frames. ], tot_loss[loss=0.1612, simple_loss=0.2361, pruned_loss=0.04312, over 1432474.79 frames. ], batch size: 63, lr: 1.03e-02, grad_scale: 8.0 +2023-03-20 23:42:48,407 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.382e+02 2.268e+02 2.632e+02 3.225e+02 5.905e+02, threshold=5.263e+02, percent-clipped=2.0 +2023-03-20 23:42:52,772 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6849, 3.4361, 3.6349, 3.7502, 3.9224, 3.7553, 3.6901, 3.6072], + device='cuda:0'), covar=tensor([0.0044, 0.0078, 0.0035, 0.0034, 0.0024, 0.0034, 0.0043, 0.0048], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0048, 0.0043, 0.0041, 0.0041, 0.0044, 0.0045, 0.0053], + device='cuda:0'), out_proj_covar=tensor([8.1428e-05, 1.2395e-04, 1.0768e-04, 9.1569e-05, 9.2731e-05, 1.0239e-04, + 1.1426e-04, 1.2223e-04], device='cuda:0') +2023-03-20 23:42:58,722 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-20 23:43:00,324 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2334, 0.9823, 1.6042, 1.5666, 1.4263, 1.6556, 1.3535, 1.5910], + device='cuda:0'), covar=tensor([0.1324, 0.2268, 0.0718, 0.0810, 0.1375, 0.1088, 0.0865, 0.1198], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0049, 0.0036, 0.0035, 0.0038, 0.0039, 0.0054, 0.0040], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 23:43:03,246 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-20 23:43:10,296 INFO [train.py:901] (0/2) Epoch 16, batch 1100, loss[loss=0.1684, simple_loss=0.2449, pruned_loss=0.04593, over 7349.00 frames. ], tot_loss[loss=0.1608, simple_loss=0.236, pruned_loss=0.04286, over 1435733.54 frames. ], batch size: 61, lr: 1.03e-02, grad_scale: 8.0 +2023-03-20 23:43:32,721 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-20 23:43:33,237 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 23:43:36,286 INFO [train.py:901] (0/2) Epoch 16, batch 1150, loss[loss=0.1304, simple_loss=0.2119, pruned_loss=0.02447, over 7341.00 frames. ], tot_loss[loss=0.1603, simple_loss=0.2355, pruned_loss=0.04255, over 1438944.82 frames. ], batch size: 44, lr: 1.03e-02, grad_scale: 8.0 +2023-03-20 23:43:38,543 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=43514.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:43:39,047 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=43515.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:43:40,854 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.302e+02 2.117e+02 2.537e+02 3.053e+02 5.220e+02, threshold=5.075e+02, percent-clipped=0.0 +2023-03-20 23:43:46,487 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-20 23:43:47,032 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-20 23:43:49,068 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=43535.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:44:01,967 INFO [train.py:901] (0/2) Epoch 16, batch 1200, loss[loss=0.1343, simple_loss=0.2106, pruned_loss=0.02896, over 7146.00 frames. ], tot_loss[loss=0.1616, simple_loss=0.2366, pruned_loss=0.04331, over 1439448.39 frames. ], batch size: 41, lr: 1.03e-02, grad_scale: 8.0 +2023-03-20 23:44:03,061 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=43563.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:44:03,191 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4224, 3.2060, 3.0547, 2.9838, 2.6536, 2.4830, 3.2107, 2.5627], + device='cuda:0'), covar=tensor([0.0251, 0.0279, 0.0293, 0.0338, 0.0386, 0.0555, 0.0421, 0.0992], + device='cuda:0'), in_proj_covar=tensor([0.0303, 0.0305, 0.0250, 0.0322, 0.0300, 0.0294, 0.0308, 0.0287], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 23:44:05,100 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7114, 4.0283, 3.8406, 3.9783, 3.6694, 4.1045, 4.3275, 4.3969], + device='cuda:0'), covar=tensor([0.0218, 0.0136, 0.0207, 0.0177, 0.0328, 0.0189, 0.0217, 0.0170], + device='cuda:0'), in_proj_covar=tensor([0.0102, 0.0103, 0.0096, 0.0104, 0.0096, 0.0088, 0.0085, 0.0082], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:44:06,670 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4225, 1.8525, 1.6013, 1.6252, 1.8494, 1.4207, 1.4528, 1.1393], + device='cuda:0'), covar=tensor([0.0176, 0.0123, 0.0168, 0.0099, 0.0171, 0.0150, 0.0142, 0.0148], + device='cuda:0'), in_proj_covar=tensor([0.0024, 0.0022, 0.0022, 0.0023, 0.0023, 0.0022, 0.0024, 0.0029], + device='cuda:0'), out_proj_covar=tensor([2.8420e-05, 2.5502e-05, 2.5823e-05, 2.5479e-05, 2.7402e-05, 2.4722e-05, + 2.7717e-05, 3.4166e-05], device='cuda:0') +2023-03-20 23:44:09,821 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=43575.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:44:13,703 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=43583.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:44:18,700 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-20 23:44:27,278 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3743, 1.0835, 1.5828, 1.6540, 1.4567, 1.7535, 1.4628, 1.6600], + device='cuda:0'), covar=tensor([0.0843, 0.2272, 0.1056, 0.0747, 0.3417, 0.1364, 0.0889, 0.1606], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0049, 0.0037, 0.0035, 0.0038, 0.0039, 0.0054, 0.0041], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 23:44:27,797 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6606, 1.6493, 2.1259, 1.7683, 1.9588, 1.7233, 1.5942, 1.4471], + device='cuda:0'), covar=tensor([0.0187, 0.0433, 0.0137, 0.0088, 0.0349, 0.0284, 0.0207, 0.0255], + device='cuda:0'), in_proj_covar=tensor([0.0023, 0.0023, 0.0022, 0.0020, 0.0021, 0.0020, 0.0023, 0.0024], + device='cuda:0'), out_proj_covar=tensor([5.8527e-05, 5.8512e-05, 5.3148e-05, 4.8331e-05, 5.4226e-05, 5.1827e-05, + 5.6834e-05, 6.0614e-05], device='cuda:0') +2023-03-20 23:44:28,674 INFO [train.py:901] (0/2) Epoch 16, batch 1250, loss[loss=0.1621, simple_loss=0.2424, pruned_loss=0.04085, over 7253.00 frames. ], tot_loss[loss=0.1615, simple_loss=0.2365, pruned_loss=0.04321, over 1440127.27 frames. ], batch size: 55, lr: 1.03e-02, grad_scale: 8.0 +2023-03-20 23:44:32,680 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.619e+02 2.248e+02 2.795e+02 3.324e+02 5.636e+02, threshold=5.591e+02, percent-clipped=2.0 +2023-03-20 23:44:42,922 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-20 23:44:46,526 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-20 23:44:47,971 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-20 23:44:50,436 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.82 vs. limit=5.0 +2023-03-20 23:44:54,625 INFO [train.py:901] (0/2) Epoch 16, batch 1300, loss[loss=0.1562, simple_loss=0.2338, pruned_loss=0.03933, over 7282.00 frames. ], tot_loss[loss=0.1613, simple_loss=0.2363, pruned_loss=0.04314, over 1441566.04 frames. ], batch size: 70, lr: 1.03e-02, grad_scale: 8.0 +2023-03-20 23:45:00,736 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1590, 0.8245, 1.4535, 1.4740, 1.3926, 1.6198, 1.3373, 1.3987], + device='cuda:0'), covar=tensor([0.0852, 0.3798, 0.0717, 0.1127, 0.0937, 0.1453, 0.0708, 0.1974], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0048, 0.0037, 0.0035, 0.0038, 0.0039, 0.0054, 0.0041], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 23:45:11,961 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-20 23:45:14,406 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-20 23:45:17,885 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-20 23:45:20,342 INFO [train.py:901] (0/2) Epoch 16, batch 1350, loss[loss=0.1705, simple_loss=0.2459, pruned_loss=0.04753, over 7220.00 frames. ], tot_loss[loss=0.1612, simple_loss=0.2361, pruned_loss=0.04314, over 1439214.68 frames. ], batch size: 93, lr: 1.03e-02, grad_scale: 8.0 +2023-03-20 23:45:24,371 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.464e+02 2.176e+02 2.636e+02 3.037e+02 1.340e+03, threshold=5.272e+02, percent-clipped=1.0 +2023-03-20 23:45:27,066 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-20 23:45:46,727 INFO [train.py:901] (0/2) Epoch 16, batch 1400, loss[loss=0.1663, simple_loss=0.2415, pruned_loss=0.04557, over 7347.00 frames. ], tot_loss[loss=0.1613, simple_loss=0.2361, pruned_loss=0.04319, over 1439880.09 frames. ], batch size: 73, lr: 1.03e-02, grad_scale: 8.0 +2023-03-20 23:46:00,597 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=43787.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:46:01,530 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-20 23:46:08,194 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.95 vs. limit=2.0 +2023-03-20 23:46:12,962 INFO [train.py:901] (0/2) Epoch 16, batch 1450, loss[loss=0.1625, simple_loss=0.2346, pruned_loss=0.04518, over 7326.00 frames. ], tot_loss[loss=0.1606, simple_loss=0.2353, pruned_loss=0.04292, over 1439537.66 frames. ], batch size: 54, lr: 1.03e-02, grad_scale: 8.0 +2023-03-20 23:46:16,889 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.415e+02 2.136e+02 2.517e+02 2.991e+02 6.215e+02, threshold=5.033e+02, percent-clipped=2.0 +2023-03-20 23:46:25,577 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-20 23:46:32,228 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=43848.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:46:39,415 INFO [train.py:901] (0/2) Epoch 16, batch 1500, loss[loss=0.1489, simple_loss=0.2347, pruned_loss=0.03156, over 7256.00 frames. ], tot_loss[loss=0.161, simple_loss=0.236, pruned_loss=0.04305, over 1441088.48 frames. ], batch size: 89, lr: 1.03e-02, grad_scale: 8.0 +2023-03-20 23:46:42,500 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-20 23:46:44,076 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=43870.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:47:03,863 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=43909.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:47:04,721 INFO [train.py:901] (0/2) Epoch 16, batch 1550, loss[loss=0.1399, simple_loss=0.2254, pruned_loss=0.02725, over 7237.00 frames. ], tot_loss[loss=0.1612, simple_loss=0.2364, pruned_loss=0.04301, over 1442031.96 frames. ], batch size: 89, lr: 1.02e-02, grad_scale: 8.0 +2023-03-20 23:47:05,764 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-20 23:47:09,380 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.128e+02 2.273e+02 2.649e+02 3.291e+02 5.530e+02, threshold=5.297e+02, percent-clipped=1.0 +2023-03-20 23:47:25,183 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3476, 4.2408, 4.1285, 3.7162, 3.9382, 2.7686, 2.1756, 4.4252], + device='cuda:0'), covar=tensor([0.0051, 0.0065, 0.0069, 0.0081, 0.0075, 0.0384, 0.0493, 0.0048], + device='cuda:0'), in_proj_covar=tensor([0.0082, 0.0069, 0.0092, 0.0077, 0.0099, 0.0115, 0.0115, 0.0082], + device='cuda:0'), out_proj_covar=tensor([1.0319e-04, 9.8245e-05, 1.2104e-04, 1.0500e-04, 1.2873e-04, 1.5056e-04, + 1.5092e-04, 1.0483e-04], device='cuda:0') +2023-03-20 23:47:30,253 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=43959.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:47:31,150 INFO [train.py:901] (0/2) Epoch 16, batch 1600, loss[loss=0.1542, simple_loss=0.2303, pruned_loss=0.03905, over 7346.00 frames. ], tot_loss[loss=0.1609, simple_loss=0.2358, pruned_loss=0.04298, over 1443413.35 frames. ], batch size: 63, lr: 1.02e-02, grad_scale: 8.0 +2023-03-20 23:47:35,819 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=43970.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:47:36,170 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-20 23:47:37,147 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-20 23:47:39,666 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-20 23:47:50,302 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4149, 4.3792, 4.2473, 3.7330, 3.8302, 2.8840, 2.4203, 4.4276], + device='cuda:0'), covar=tensor([0.0033, 0.0044, 0.0039, 0.0054, 0.0066, 0.0321, 0.0391, 0.0038], + device='cuda:0'), in_proj_covar=tensor([0.0081, 0.0068, 0.0091, 0.0077, 0.0099, 0.0114, 0.0114, 0.0082], + device='cuda:0'), out_proj_covar=tensor([1.0275e-04, 9.7072e-05, 1.2019e-04, 1.0469e-04, 1.2877e-04, 1.4946e-04, + 1.5007e-04, 1.0454e-04], device='cuda:0') +2023-03-20 23:47:51,033 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-44000.pt +2023-03-20 23:47:54,815 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-20 23:47:57,460 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9287, 2.2884, 1.7608, 2.8336, 2.4736, 2.9359, 2.5070, 2.2726], + device='cuda:0'), covar=tensor([0.1791, 0.0868, 0.2982, 0.0501, 0.0093, 0.0066, 0.0183, 0.0233], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0230, 0.0266, 0.0256, 0.0142, 0.0133, 0.0162, 0.0186], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0001, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:47:59,314 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-20 23:48:00,782 INFO [train.py:901] (0/2) Epoch 16, batch 1650, loss[loss=0.1891, simple_loss=0.2533, pruned_loss=0.06249, over 7247.00 frames. ], tot_loss[loss=0.1602, simple_loss=0.2351, pruned_loss=0.04272, over 1440203.30 frames. ], batch size: 55, lr: 1.02e-02, grad_scale: 8.0 +2023-03-20 23:48:03,167 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.44 vs. limit=2.0 +2023-03-20 23:48:04,870 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.383e+02 2.044e+02 2.444e+02 3.043e+02 5.639e+02, threshold=4.888e+02, percent-clipped=2.0 +2023-03-20 23:48:05,563 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=44020.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:48:07,490 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-20 23:48:25,194 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-20 23:48:26,661 INFO [train.py:901] (0/2) Epoch 16, batch 1700, loss[loss=0.1295, simple_loss=0.2042, pruned_loss=0.02734, over 7306.00 frames. ], tot_loss[loss=0.1605, simple_loss=0.2353, pruned_loss=0.0429, over 1438801.17 frames. ], batch size: 42, lr: 1.02e-02, grad_scale: 8.0 +2023-03-20 23:48:28,121 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.65 vs. limit=5.0 +2023-03-20 23:48:29,249 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-20 23:48:39,374 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-20 23:48:52,737 INFO [train.py:901] (0/2) Epoch 16, batch 1750, loss[loss=0.1629, simple_loss=0.2436, pruned_loss=0.04108, over 7264.00 frames. ], tot_loss[loss=0.1599, simple_loss=0.235, pruned_loss=0.04244, over 1438974.75 frames. ], batch size: 89, lr: 1.02e-02, grad_scale: 8.0 +2023-03-20 23:48:56,791 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.516e+02 2.154e+02 2.480e+02 3.011e+02 7.790e+02, threshold=4.959e+02, percent-clipped=2.0 +2023-03-20 23:49:05,028 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-20 23:49:05,537 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-20 23:49:05,808 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.30 vs. limit=2.0 +2023-03-20 23:49:09,685 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=44143.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:49:13,779 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-03-20 23:49:18,533 INFO [train.py:901] (0/2) Epoch 16, batch 1800, loss[loss=0.1449, simple_loss=0.21, pruned_loss=0.03988, over 6966.00 frames. ], tot_loss[loss=0.1609, simple_loss=0.2357, pruned_loss=0.04301, over 1439345.33 frames. ], batch size: 35, lr: 1.02e-02, grad_scale: 8.0 +2023-03-20 23:49:22,422 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.68 vs. limit=2.0 +2023-03-20 23:49:23,129 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=44170.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:49:26,646 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-20 23:49:33,343 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.89 vs. limit=2.0 +2023-03-20 23:49:39,918 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-20 23:49:45,065 INFO [train.py:901] (0/2) Epoch 16, batch 1850, loss[loss=0.1657, simple_loss=0.2357, pruned_loss=0.04786, over 7339.00 frames. ], tot_loss[loss=0.1611, simple_loss=0.236, pruned_loss=0.04305, over 1439877.98 frames. ], batch size: 73, lr: 1.02e-02, grad_scale: 8.0 +2023-03-20 23:49:48,567 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=44218.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:49:48,969 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.288e+02 2.195e+02 2.582e+02 3.217e+02 6.301e+02, threshold=5.165e+02, percent-clipped=5.0 +2023-03-20 23:49:51,004 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-20 23:50:08,078 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-20 23:50:10,095 INFO [train.py:901] (0/2) Epoch 16, batch 1900, loss[loss=0.1483, simple_loss=0.2284, pruned_loss=0.03415, over 7300.00 frames. ], tot_loss[loss=0.1604, simple_loss=0.2355, pruned_loss=0.0427, over 1439616.09 frames. ], batch size: 68, lr: 1.02e-02, grad_scale: 8.0 +2023-03-20 23:50:12,156 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=44265.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:50:33,879 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-20 23:50:33,951 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5134, 3.6765, 3.5566, 3.6240, 3.5586, 3.4774, 3.6652, 3.7827], + device='cuda:0'), covar=tensor([0.0429, 0.0297, 0.0332, 0.0391, 0.0395, 0.0439, 0.0571, 0.0515], + device='cuda:0'), in_proj_covar=tensor([0.0106, 0.0105, 0.0099, 0.0107, 0.0098, 0.0091, 0.0087, 0.0085], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:50:36,349 INFO [train.py:901] (0/2) Epoch 16, batch 1950, loss[loss=0.1525, simple_loss=0.2325, pruned_loss=0.03626, over 7305.00 frames. ], tot_loss[loss=0.1601, simple_loss=0.2354, pruned_loss=0.04243, over 1442749.30 frames. ], batch size: 80, lr: 1.02e-02, grad_scale: 8.0 +2023-03-20 23:50:38,444 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=44315.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:50:40,373 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.314e+02 2.194e+02 2.559e+02 3.068e+02 6.748e+02, threshold=5.118e+02, percent-clipped=3.0 +2023-03-20 23:50:43,556 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2871, 1.1030, 1.6204, 1.5870, 1.5581, 1.7660, 1.4159, 1.7198], + device='cuda:0'), covar=tensor([0.1798, 0.2125, 0.0729, 0.1176, 0.0811, 0.3291, 0.1317, 0.1914], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0051, 0.0036, 0.0035, 0.0039, 0.0041, 0.0054, 0.0042], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-20 23:50:44,417 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-20 23:50:49,027 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-20 23:50:49,523 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-20 23:51:02,196 INFO [train.py:901] (0/2) Epoch 16, batch 2000, loss[loss=0.1443, simple_loss=0.2198, pruned_loss=0.03437, over 7362.00 frames. ], tot_loss[loss=0.1601, simple_loss=0.2352, pruned_loss=0.04251, over 1441711.07 frames. ], batch size: 65, lr: 1.02e-02, grad_scale: 8.0 +2023-03-20 23:51:05,747 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-20 23:51:05,872 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6668, 3.1320, 2.2962, 2.8727, 2.8009, 2.3456, 3.1441, 2.8615], + device='cuda:0'), covar=tensor([0.0892, 0.0607, 0.2386, 0.1739, 0.1609, 0.0867, 0.0594, 0.1003], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0043, 0.0049, 0.0043, 0.0042, 0.0044, 0.0044, 0.0042], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:51:15,539 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3577, 3.3558, 3.4131, 3.2136, 2.6752, 2.8394, 3.5596, 2.8781], + device='cuda:0'), covar=tensor([0.0237, 0.0285, 0.0274, 0.0302, 0.0376, 0.0504, 0.0303, 0.1002], + device='cuda:0'), in_proj_covar=tensor([0.0301, 0.0305, 0.0245, 0.0326, 0.0298, 0.0292, 0.0312, 0.0285], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 23:51:16,915 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-20 23:51:24,978 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.81 vs. limit=5.0 +2023-03-20 23:51:25,767 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-20 23:51:28,247 INFO [train.py:901] (0/2) Epoch 16, batch 2050, loss[loss=0.1554, simple_loss=0.2281, pruned_loss=0.04134, over 7194.00 frames. ], tot_loss[loss=0.1608, simple_loss=0.2358, pruned_loss=0.04292, over 1443340.88 frames. ], batch size: 50, lr: 1.02e-02, grad_scale: 8.0 +2023-03-20 23:51:32,325 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.233e+02 2.247e+02 2.734e+02 3.151e+02 5.434e+02, threshold=5.468e+02, percent-clipped=2.0 +2023-03-20 23:51:45,052 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=44443.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:51:54,000 INFO [train.py:901] (0/2) Epoch 16, batch 2100, loss[loss=0.1562, simple_loss=0.2303, pruned_loss=0.04101, over 7282.00 frames. ], tot_loss[loss=0.1616, simple_loss=0.2368, pruned_loss=0.04324, over 1444892.97 frames. ], batch size: 68, lr: 1.02e-02, grad_scale: 8.0 +2023-03-20 23:52:00,442 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-20 23:52:04,552 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-20 23:52:06,182 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4357, 3.3336, 3.4396, 2.9235, 2.5002, 2.7826, 3.2702, 2.8476], + device='cuda:0'), covar=tensor([0.0216, 0.0244, 0.0268, 0.0269, 0.0427, 0.0563, 0.0345, 0.0986], + device='cuda:0'), in_proj_covar=tensor([0.0305, 0.0308, 0.0248, 0.0329, 0.0302, 0.0296, 0.0315, 0.0288], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 23:52:09,573 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=44491.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:52:11,231 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.92 vs. limit=5.0 +2023-03-20 23:52:19,295 INFO [train.py:901] (0/2) Epoch 16, batch 2150, loss[loss=0.1769, simple_loss=0.2571, pruned_loss=0.04837, over 7278.00 frames. ], tot_loss[loss=0.161, simple_loss=0.2363, pruned_loss=0.04289, over 1444900.18 frames. ], batch size: 57, lr: 1.02e-02, grad_scale: 8.0 +2023-03-20 23:52:23,162 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.579e+02 2.371e+02 2.688e+02 3.151e+02 5.325e+02, threshold=5.377e+02, percent-clipped=0.0 +2023-03-20 23:52:27,925 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=44527.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:52:34,666 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-03-20 23:52:39,479 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=44550.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:52:44,914 INFO [train.py:901] (0/2) Epoch 16, batch 2200, loss[loss=0.1898, simple_loss=0.2572, pruned_loss=0.0612, over 7321.00 frames. ], tot_loss[loss=0.1615, simple_loss=0.2364, pruned_loss=0.04331, over 1439799.77 frames. ], batch size: 80, lr: 1.02e-02, grad_scale: 8.0 +2023-03-20 23:52:47,081 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=44565.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:52:49,658 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-20 23:52:59,360 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=44588.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:53:08,361 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9336, 2.6149, 1.7039, 2.7779, 3.2424, 2.9931, 2.2712, 2.3715], + device='cuda:0'), covar=tensor([0.2055, 0.0702, 0.3142, 0.0618, 0.0162, 0.0096, 0.0210, 0.0234], + device='cuda:0'), in_proj_covar=tensor([0.0277, 0.0228, 0.0267, 0.0261, 0.0145, 0.0135, 0.0165, 0.0187], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:53:11,852 INFO [train.py:901] (0/2) Epoch 16, batch 2250, loss[loss=0.1761, simple_loss=0.2485, pruned_loss=0.05184, over 7273.00 frames. ], tot_loss[loss=0.161, simple_loss=0.2358, pruned_loss=0.04306, over 1439124.58 frames. ], batch size: 70, lr: 1.02e-02, grad_scale: 8.0 +2023-03-20 23:53:11,991 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=44611.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:53:12,927 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=44613.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:53:14,030 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=44615.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:53:15,907 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.421e+02 2.027e+02 2.444e+02 2.860e+02 7.319e+02, threshold=4.889e+02, percent-clipped=1.0 +2023-03-20 23:53:22,477 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-20 23:53:22,948 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-20 23:53:36,691 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-20 23:53:37,647 INFO [train.py:901] (0/2) Epoch 16, batch 2300, loss[loss=0.1536, simple_loss=0.2303, pruned_loss=0.03843, over 7355.00 frames. ], tot_loss[loss=0.1612, simple_loss=0.2363, pruned_loss=0.04305, over 1439920.46 frames. ], batch size: 73, lr: 1.02e-02, grad_scale: 8.0 +2023-03-20 23:53:38,736 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=44663.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:53:40,279 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8143, 4.3774, 4.3305, 4.7855, 4.7987, 4.7823, 4.0111, 4.3490], + device='cuda:0'), covar=tensor([0.0699, 0.2673, 0.2377, 0.0983, 0.0641, 0.1367, 0.0813, 0.1291], + device='cuda:0'), in_proj_covar=tensor([0.0202, 0.0322, 0.0252, 0.0246, 0.0187, 0.0307, 0.0179, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:53:56,703 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7708, 3.8239, 3.6898, 3.9351, 3.7483, 3.7512, 4.1440, 4.2216], + device='cuda:0'), covar=tensor([0.0207, 0.0183, 0.0224, 0.0157, 0.0268, 0.0425, 0.0264, 0.0203], + device='cuda:0'), in_proj_covar=tensor([0.0104, 0.0105, 0.0100, 0.0105, 0.0097, 0.0089, 0.0086, 0.0084], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:54:00,810 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7239, 2.1686, 2.3970, 1.7432, 2.3530, 1.6877, 1.6091, 1.5642], + device='cuda:0'), covar=tensor([0.0339, 0.0164, 0.0127, 0.0217, 0.0427, 0.0473, 0.0376, 0.0314], + device='cuda:0'), in_proj_covar=tensor([0.0024, 0.0023, 0.0021, 0.0021, 0.0021, 0.0020, 0.0024, 0.0025], + device='cuda:0'), out_proj_covar=tensor([5.9997e-05, 5.8519e-05, 5.2941e-05, 5.1038e-05, 5.5644e-05, 5.2308e-05, + 5.8162e-05, 6.2053e-05], device='cuda:0') +2023-03-20 23:54:03,686 INFO [train.py:901] (0/2) Epoch 16, batch 2350, loss[loss=0.1356, simple_loss=0.1973, pruned_loss=0.03693, over 6097.00 frames. ], tot_loss[loss=0.1609, simple_loss=0.2359, pruned_loss=0.04296, over 1439670.83 frames. ], batch size: 26, lr: 1.02e-02, grad_scale: 8.0 +2023-03-20 23:54:07,744 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.213e+02 2.179e+02 2.590e+02 3.054e+02 7.266e+02, threshold=5.180e+02, percent-clipped=2.0 +2023-03-20 23:54:24,462 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-20 23:54:29,412 INFO [train.py:901] (0/2) Epoch 16, batch 2400, loss[loss=0.1753, simple_loss=0.2547, pruned_loss=0.04798, over 7248.00 frames. ], tot_loss[loss=0.1604, simple_loss=0.2357, pruned_loss=0.0426, over 1442315.91 frames. ], batch size: 55, lr: 1.02e-02, grad_scale: 8.0 +2023-03-20 23:54:30,429 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-20 23:54:42,155 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-20 23:54:44,720 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-20 23:54:53,080 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.1744, 5.6605, 5.7151, 5.6613, 5.4454, 5.3357, 5.7943, 5.5085], + device='cuda:0'), covar=tensor([0.0369, 0.0279, 0.0294, 0.0333, 0.0230, 0.0220, 0.0246, 0.0311], + device='cuda:0'), in_proj_covar=tensor([0.0185, 0.0211, 0.0153, 0.0154, 0.0129, 0.0192, 0.0165, 0.0126], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:54:55,559 INFO [train.py:901] (0/2) Epoch 16, batch 2450, loss[loss=0.1677, simple_loss=0.2441, pruned_loss=0.04563, over 7321.00 frames. ], tot_loss[loss=0.1606, simple_loss=0.2357, pruned_loss=0.04281, over 1441618.92 frames. ], batch size: 59, lr: 1.01e-02, grad_scale: 8.0 +2023-03-20 23:54:59,655 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.397e+02 2.079e+02 2.447e+02 2.964e+02 4.807e+02, threshold=4.894e+02, percent-clipped=0.0 +2023-03-20 23:55:05,478 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.96 vs. limit=2.0 +2023-03-20 23:55:10,798 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-20 23:55:19,891 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5642, 4.3005, 4.1992, 3.9070, 4.1073, 2.9649, 1.9439, 4.5666], + device='cuda:0'), covar=tensor([0.0025, 0.0079, 0.0050, 0.0049, 0.0048, 0.0298, 0.0475, 0.0028], + device='cuda:0'), in_proj_covar=tensor([0.0081, 0.0069, 0.0090, 0.0075, 0.0099, 0.0113, 0.0114, 0.0081], + device='cuda:0'), out_proj_covar=tensor([1.0215e-04, 9.6868e-05, 1.1886e-04, 1.0237e-04, 1.2656e-04, 1.4747e-04, + 1.4996e-04, 1.0385e-04], device='cuda:0') +2023-03-20 23:55:21,250 INFO [train.py:901] (0/2) Epoch 16, batch 2500, loss[loss=0.2025, simple_loss=0.2701, pruned_loss=0.06745, over 6643.00 frames. ], tot_loss[loss=0.1614, simple_loss=0.2364, pruned_loss=0.04326, over 1438951.47 frames. ], batch size: 106, lr: 1.01e-02, grad_scale: 16.0 +2023-03-20 23:55:32,928 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=44883.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:55:35,753 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-20 23:55:39,430 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9249, 2.8948, 2.5516, 2.7166, 2.2140, 2.2961, 2.8651, 2.2442], + device='cuda:0'), covar=tensor([0.0351, 0.0366, 0.0239, 0.0388, 0.0383, 0.0535, 0.0468, 0.0820], + device='cuda:0'), in_proj_covar=tensor([0.0308, 0.0314, 0.0248, 0.0331, 0.0304, 0.0301, 0.0318, 0.0291], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 23:55:42,135 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-20 23:55:44,401 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=44906.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:55:46,774 INFO [train.py:901] (0/2) Epoch 16, batch 2550, loss[loss=0.1473, simple_loss=0.2213, pruned_loss=0.03669, over 7141.00 frames. ], tot_loss[loss=0.1624, simple_loss=0.237, pruned_loss=0.04386, over 1439137.62 frames. ], batch size: 41, lr: 1.01e-02, grad_scale: 16.0 +2023-03-20 23:55:51,397 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.227e+02 2.271e+02 2.744e+02 3.382e+02 5.975e+02, threshold=5.489e+02, percent-clipped=3.0 +2023-03-20 23:56:11,325 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.91 vs. limit=2.0 +2023-03-20 23:56:12,445 INFO [train.py:901] (0/2) Epoch 16, batch 2600, loss[loss=0.1676, simple_loss=0.241, pruned_loss=0.04704, over 7341.00 frames. ], tot_loss[loss=0.1621, simple_loss=0.2368, pruned_loss=0.04368, over 1441210.14 frames. ], batch size: 61, lr: 1.01e-02, grad_scale: 16.0 +2023-03-20 23:56:38,026 INFO [train.py:901] (0/2) Epoch 16, batch 2650, loss[loss=0.1742, simple_loss=0.2419, pruned_loss=0.05322, over 7328.00 frames. ], tot_loss[loss=0.1623, simple_loss=0.2371, pruned_loss=0.04372, over 1443055.48 frames. ], batch size: 75, lr: 1.01e-02, grad_scale: 16.0 +2023-03-20 23:56:42,042 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.330e+02 2.314e+02 2.696e+02 3.336e+02 6.334e+02, threshold=5.393e+02, percent-clipped=2.0 +2023-03-20 23:57:00,338 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=45056.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:57:02,757 INFO [train.py:901] (0/2) Epoch 16, batch 2700, loss[loss=0.1641, simple_loss=0.2414, pruned_loss=0.04337, over 7283.00 frames. ], tot_loss[loss=0.1618, simple_loss=0.2368, pruned_loss=0.04339, over 1445293.24 frames. ], batch size: 68, lr: 1.01e-02, grad_scale: 16.0 +2023-03-20 23:57:03,406 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1017, 2.6861, 1.9308, 3.2127, 2.1903, 2.7000, 1.5894, 1.8171], + device='cuda:0'), covar=tensor([0.0389, 0.0618, 0.2160, 0.0486, 0.0387, 0.0666, 0.2920, 0.1789], + device='cuda:0'), in_proj_covar=tensor([0.0247, 0.0233, 0.0304, 0.0242, 0.0251, 0.0253, 0.0264, 0.0285], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0004, 0.0004], + device='cuda:0') +2023-03-20 23:57:12,726 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=45081.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:57:27,938 INFO [train.py:901] (0/2) Epoch 16, batch 2750, loss[loss=0.1592, simple_loss=0.23, pruned_loss=0.04419, over 7257.00 frames. ], tot_loss[loss=0.161, simple_loss=0.236, pruned_loss=0.04299, over 1444518.40 frames. ], batch size: 52, lr: 1.01e-02, grad_scale: 16.0 +2023-03-20 23:57:31,011 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=45117.0, num_to_drop=1, layers_to_drop={2} +2023-03-20 23:57:31,798 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.386e+02 1.985e+02 2.477e+02 3.081e+02 5.070e+02, threshold=4.954e+02, percent-clipped=0.0 +2023-03-20 23:57:32,921 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2585, 4.2243, 3.7250, 3.5975, 3.5544, 2.3487, 1.6570, 4.2374], + device='cuda:0'), covar=tensor([0.0034, 0.0046, 0.0082, 0.0064, 0.0092, 0.0407, 0.0533, 0.0041], + device='cuda:0'), in_proj_covar=tensor([0.0082, 0.0069, 0.0090, 0.0075, 0.0099, 0.0114, 0.0113, 0.0082], + device='cuda:0'), out_proj_covar=tensor([1.0260e-04, 9.7174e-05, 1.1842e-04, 1.0173e-04, 1.2678e-04, 1.4887e-04, + 1.4906e-04, 1.0531e-04], device='cuda:0') +2023-03-20 23:57:36,393 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7049, 5.1901, 5.2720, 5.2435, 5.0287, 4.7477, 5.3338, 5.0610], + device='cuda:0'), covar=tensor([0.0407, 0.0395, 0.0429, 0.0353, 0.0302, 0.0331, 0.0309, 0.0477], + device='cuda:0'), in_proj_covar=tensor([0.0180, 0.0205, 0.0152, 0.0150, 0.0126, 0.0188, 0.0159, 0.0124], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:57:43,280 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=45142.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:57:52,363 INFO [train.py:901] (0/2) Epoch 16, batch 2800, loss[loss=0.182, simple_loss=0.2467, pruned_loss=0.05869, over 7248.00 frames. ], tot_loss[loss=0.162, simple_loss=0.2368, pruned_loss=0.04362, over 1442659.23 frames. ], batch size: 47, lr: 1.01e-02, grad_scale: 8.0 +2023-03-20 23:57:55,696 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.00 vs. limit=2.0 +2023-03-20 23:58:03,109 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=45183.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:58:04,866 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-16.pt +2023-03-20 23:58:22,226 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-20 23:58:22,286 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-20 23:58:22,312 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-20 23:58:22,331 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-20 23:58:22,505 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-20 23:58:22,522 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-20 23:58:22,532 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-20 23:58:22,632 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-20 23:58:22,707 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-20 23:58:22,722 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-20 23:58:22,739 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-20 23:58:23,253 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-20 23:58:23,259 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-20 23:58:23,266 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-20 23:58:23,406 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-20 23:58:23,409 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-20 23:58:23,468 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-20 23:58:25,830 INFO [train.py:901] (0/2) Epoch 17, batch 0, loss[loss=0.1684, simple_loss=0.2437, pruned_loss=0.04652, over 7297.00 frames. ], tot_loss[loss=0.1684, simple_loss=0.2437, pruned_loss=0.04652, over 7297.00 frames. ], batch size: 68, lr: 9.83e-03, grad_scale: 8.0 +2023-03-20 23:58:25,832 INFO [train.py:926] (0/2) Computing validation loss +2023-03-20 23:58:36,381 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8299, 2.7763, 3.0348, 3.0908, 3.2184, 2.8324, 2.5647, 3.3134], + device='cuda:0'), covar=tensor([0.2996, 0.0776, 0.1828, 0.1648, 0.0954, 0.1683, 0.2926, 0.1172], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0049, 0.0038, 0.0039, 0.0039, 0.0036, 0.0054, 0.0041], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-20 23:58:52,332 INFO [train.py:935] (0/2) Epoch 17, validation: loss=0.1675, simple_loss=0.2548, pruned_loss=0.04006, over 1622729.00 frames. +2023-03-20 23:58:52,332 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-20 23:58:58,932 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-20 23:59:02,925 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4975, 3.5188, 3.3194, 3.1814, 2.5468, 2.8007, 3.4969, 2.8223], + device='cuda:0'), covar=tensor([0.0293, 0.0263, 0.0301, 0.0329, 0.0381, 0.0569, 0.0271, 0.0933], + device='cuda:0'), in_proj_covar=tensor([0.0313, 0.0315, 0.0251, 0.0334, 0.0305, 0.0303, 0.0318, 0.0292], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 23:59:03,343 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=45206.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:59:09,920 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-20 23:59:10,867 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.483e+02 2.268e+02 2.699e+02 3.291e+02 1.082e+03, threshold=5.398e+02, percent-clipped=4.0 +2023-03-20 23:59:16,388 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-20 23:59:16,421 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=45231.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:59:18,356 INFO [train.py:901] (0/2) Epoch 17, batch 50, loss[loss=0.159, simple_loss=0.2314, pruned_loss=0.04329, over 7228.00 frames. ], tot_loss[loss=0.1633, simple_loss=0.2389, pruned_loss=0.0439, over 326185.73 frames. ], batch size: 50, lr: 9.82e-03, grad_scale: 8.0 +2023-03-20 23:59:18,874 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-20 23:59:20,548 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3662, 3.3623, 3.3556, 3.2168, 2.6432, 2.8838, 3.3938, 2.8740], + device='cuda:0'), covar=tensor([0.0210, 0.0315, 0.0299, 0.0295, 0.0373, 0.0541, 0.0371, 0.0912], + device='cuda:0'), in_proj_covar=tensor([0.0312, 0.0316, 0.0250, 0.0333, 0.0305, 0.0302, 0.0317, 0.0292], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-20 23:59:21,832 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-20 23:59:23,943 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9426, 4.4808, 4.5813, 4.4946, 4.5045, 4.1753, 4.6105, 4.4135], + device='cuda:0'), covar=tensor([0.0560, 0.0489, 0.0392, 0.0437, 0.0351, 0.0354, 0.0351, 0.0570], + device='cuda:0'), in_proj_covar=tensor([0.0182, 0.0208, 0.0154, 0.0151, 0.0128, 0.0189, 0.0161, 0.0124], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-20 23:59:27,925 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=45254.0, num_to_drop=0, layers_to_drop=set() +2023-03-20 23:59:37,987 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-20 23:59:38,484 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-20 23:59:44,003 INFO [train.py:901] (0/2) Epoch 17, batch 100, loss[loss=0.1368, simple_loss=0.1943, pruned_loss=0.03969, over 6177.00 frames. ], tot_loss[loss=0.1615, simple_loss=0.2371, pruned_loss=0.04293, over 572899.95 frames. ], batch size: 27, lr: 9.82e-03, grad_scale: 8.0 +2023-03-21 00:00:00,277 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2099, 2.9764, 3.0848, 2.7359, 2.3127, 2.6798, 3.1729, 2.4514], + device='cuda:0'), covar=tensor([0.0329, 0.0236, 0.0331, 0.0339, 0.0415, 0.0564, 0.0350, 0.1114], + device='cuda:0'), in_proj_covar=tensor([0.0310, 0.0314, 0.0250, 0.0334, 0.0303, 0.0302, 0.0316, 0.0291], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 00:00:02,107 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.391e+02 2.128e+02 2.527e+02 2.960e+02 4.350e+02, threshold=5.054e+02, percent-clipped=0.0 +2023-03-21 00:00:08,369 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3466, 1.4468, 1.1909, 1.3653, 1.4271, 1.1955, 1.2903, 0.9825], + device='cuda:0'), covar=tensor([0.0110, 0.0120, 0.0171, 0.0171, 0.0224, 0.0130, 0.0155, 0.0142], + device='cuda:0'), in_proj_covar=tensor([0.0023, 0.0021, 0.0022, 0.0022, 0.0023, 0.0022, 0.0023, 0.0029], + device='cuda:0'), out_proj_covar=tensor([2.7571e-05, 2.4535e-05, 2.5579e-05, 2.5309e-05, 2.6899e-05, 2.4286e-05, + 2.6603e-05, 3.4206e-05], device='cuda:0') +2023-03-21 00:00:09,797 INFO [train.py:901] (0/2) Epoch 17, batch 150, loss[loss=0.1266, simple_loss=0.1907, pruned_loss=0.03126, over 5885.00 frames. ], tot_loss[loss=0.162, simple_loss=0.2371, pruned_loss=0.04344, over 762559.54 frames. ], batch size: 25, lr: 9.81e-03, grad_scale: 8.0 +2023-03-21 00:00:16,604 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=45348.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 00:00:19,170 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9595, 2.8321, 2.8615, 2.6341, 2.0403, 2.4981, 2.9906, 2.2291], + device='cuda:0'), covar=tensor([0.0312, 0.0344, 0.0290, 0.0388, 0.0409, 0.0563, 0.0531, 0.1257], + device='cuda:0'), in_proj_covar=tensor([0.0310, 0.0314, 0.0251, 0.0336, 0.0303, 0.0303, 0.0317, 0.0291], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 00:00:22,855 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2505, 2.5880, 2.2009, 3.5694, 1.5421, 3.3601, 1.4649, 2.9608], + device='cuda:0'), covar=tensor([0.0061, 0.0759, 0.1427, 0.0100, 0.3282, 0.0116, 0.1071, 0.0255], + device='cuda:0'), in_proj_covar=tensor([0.0164, 0.0258, 0.0290, 0.0162, 0.0276, 0.0173, 0.0260, 0.0217], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 00:00:36,397 INFO [train.py:901] (0/2) Epoch 17, batch 200, loss[loss=0.1455, simple_loss=0.2169, pruned_loss=0.03703, over 7162.00 frames. ], tot_loss[loss=0.1615, simple_loss=0.2364, pruned_loss=0.04327, over 913824.73 frames. ], batch size: 39, lr: 9.80e-03, grad_scale: 8.0 +2023-03-21 00:00:38,985 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 00:00:44,392 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 00:00:49,067 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=45409.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 00:00:50,443 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 00:00:50,502 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=45412.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 00:00:54,312 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.548e+02 2.033e+02 2.393e+02 3.029e+02 5.248e+02, threshold=4.786e+02, percent-clipped=2.0 +2023-03-21 00:01:01,785 INFO [train.py:901] (0/2) Epoch 17, batch 250, loss[loss=0.1952, simple_loss=0.2601, pruned_loss=0.06517, over 7291.00 frames. ], tot_loss[loss=0.1617, simple_loss=0.2364, pruned_loss=0.04352, over 1032960.03 frames. ], batch size: 68, lr: 9.80e-03, grad_scale: 8.0 +2023-03-21 00:01:02,826 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 00:01:02,877 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=45437.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:01:24,107 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 00:01:27,733 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4097, 4.4360, 3.9387, 3.7263, 3.9128, 2.5905, 2.2241, 4.4167], + device='cuda:0'), covar=tensor([0.0027, 0.0032, 0.0057, 0.0042, 0.0058, 0.0346, 0.0398, 0.0032], + device='cuda:0'), in_proj_covar=tensor([0.0082, 0.0070, 0.0090, 0.0076, 0.0099, 0.0115, 0.0114, 0.0082], + device='cuda:0'), out_proj_covar=tensor([1.0375e-04, 9.8215e-05, 1.1713e-04, 1.0275e-04, 1.2617e-04, 1.5021e-04, + 1.4972e-04, 1.0489e-04], device='cuda:0') +2023-03-21 00:01:28,134 INFO [train.py:901] (0/2) Epoch 17, batch 300, loss[loss=0.1722, simple_loss=0.2515, pruned_loss=0.04647, over 7244.00 frames. ], tot_loss[loss=0.1616, simple_loss=0.2366, pruned_loss=0.0433, over 1124614.86 frames. ], batch size: 45, lr: 9.79e-03, grad_scale: 8.0 +2023-03-21 00:01:32,756 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 00:01:32,937 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1814, 3.0287, 3.0037, 2.7839, 2.2684, 2.3965, 3.2896, 2.3556], + device='cuda:0'), covar=tensor([0.0387, 0.0394, 0.0300, 0.0357, 0.0396, 0.0615, 0.0458, 0.1069], + device='cuda:0'), in_proj_covar=tensor([0.0310, 0.0315, 0.0252, 0.0335, 0.0302, 0.0301, 0.0318, 0.0290], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 00:01:45,531 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.226e+02 2.225e+02 2.796e+02 3.288e+02 5.407e+02, threshold=5.591e+02, percent-clipped=4.0 +2023-03-21 00:01:53,618 INFO [train.py:901] (0/2) Epoch 17, batch 350, loss[loss=0.1455, simple_loss=0.2153, pruned_loss=0.03784, over 7142.00 frames. ], tot_loss[loss=0.1621, simple_loss=0.2374, pruned_loss=0.04345, over 1198353.46 frames. ], batch size: 41, lr: 9.79e-03, grad_scale: 8.0 +2023-03-21 00:02:06,419 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 00:02:17,172 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4928, 3.0249, 2.3492, 4.0463, 1.6261, 3.8346, 1.4365, 3.2434], + device='cuda:0'), covar=tensor([0.0051, 0.0669, 0.1527, 0.0071, 0.3955, 0.0106, 0.1083, 0.0209], + device='cuda:0'), in_proj_covar=tensor([0.0164, 0.0259, 0.0288, 0.0162, 0.0276, 0.0175, 0.0259, 0.0215], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 00:02:19,547 INFO [train.py:901] (0/2) Epoch 17, batch 400, loss[loss=0.1622, simple_loss=0.2448, pruned_loss=0.03976, over 7306.00 frames. ], tot_loss[loss=0.1604, simple_loss=0.2357, pruned_loss=0.04249, over 1252774.78 frames. ], batch size: 59, lr: 9.78e-03, grad_scale: 8.0 +2023-03-21 00:02:38,160 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.499e+02 2.037e+02 2.425e+02 2.810e+02 4.522e+02, threshold=4.850e+02, percent-clipped=0.0 +2023-03-21 00:02:45,678 INFO [train.py:901] (0/2) Epoch 17, batch 450, loss[loss=0.139, simple_loss=0.2152, pruned_loss=0.03139, over 7294.00 frames. ], tot_loss[loss=0.1598, simple_loss=0.2353, pruned_loss=0.04216, over 1295760.53 frames. ], batch size: 42, lr: 9.78e-03, grad_scale: 8.0 +2023-03-21 00:02:49,805 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 00:02:50,316 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 00:02:52,363 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8640, 3.6597, 3.6629, 3.6640, 3.4433, 3.4911, 3.8226, 3.4427], + device='cuda:0'), covar=tensor([0.0202, 0.0158, 0.0117, 0.0148, 0.0457, 0.0138, 0.0158, 0.0173], + device='cuda:0'), in_proj_covar=tensor([0.0071, 0.0076, 0.0077, 0.0066, 0.0135, 0.0088, 0.0082, 0.0083], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:03:04,401 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4585, 2.7528, 2.1668, 3.9896, 1.5232, 3.6156, 1.4182, 3.0329], + device='cuda:0'), covar=tensor([0.0075, 0.0879, 0.1771, 0.0109, 0.4565, 0.0121, 0.1200, 0.0272], + device='cuda:0'), in_proj_covar=tensor([0.0165, 0.0261, 0.0289, 0.0162, 0.0277, 0.0174, 0.0260, 0.0217], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 00:03:05,808 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0796, 3.4487, 4.1046, 3.8022, 4.0562, 4.1509, 3.9858, 3.8782], + device='cuda:0'), covar=tensor([0.0027, 0.0090, 0.0029, 0.0041, 0.0030, 0.0025, 0.0033, 0.0044], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0050, 0.0044, 0.0042, 0.0042, 0.0044, 0.0044, 0.0054], + device='cuda:0'), out_proj_covar=tensor([7.9226e-05, 1.2582e-04, 1.0600e-04, 9.4415e-05, 9.4507e-05, 9.7801e-05, + 1.0875e-04, 1.2438e-04], device='cuda:0') +2023-03-21 00:03:11,152 INFO [train.py:901] (0/2) Epoch 17, batch 500, loss[loss=0.1278, simple_loss=0.1846, pruned_loss=0.03547, over 6193.00 frames. ], tot_loss[loss=0.1595, simple_loss=0.2344, pruned_loss=0.04234, over 1324793.14 frames. ], batch size: 27, lr: 9.77e-03, grad_scale: 8.0 +2023-03-21 00:03:18,901 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0230, 2.3661, 3.0256, 2.7461, 2.9664, 2.5370, 2.4811, 2.7773], + device='cuda:0'), covar=tensor([0.0895, 0.0492, 0.0893, 0.1483, 0.0755, 0.1219, 0.2153, 0.1265], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0048, 0.0037, 0.0039, 0.0037, 0.0035, 0.0053, 0.0041], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 00:03:20,486 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=45702.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:03:21,414 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=45704.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 00:03:22,854 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 00:03:24,366 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 00:03:24,919 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 00:03:25,507 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=45712.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 00:03:27,447 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 00:03:29,418 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.410e+02 2.314e+02 2.733e+02 3.177e+02 5.495e+02, threshold=5.467e+02, percent-clipped=2.0 +2023-03-21 00:03:31,633 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 00:03:37,732 INFO [train.py:901] (0/2) Epoch 17, batch 550, loss[loss=0.1627, simple_loss=0.2281, pruned_loss=0.04867, over 7239.00 frames. ], tot_loss[loss=0.1601, simple_loss=0.2349, pruned_loss=0.04264, over 1350174.36 frames. ], batch size: 45, lr: 9.77e-03, grad_scale: 8.0 +2023-03-21 00:03:38,887 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=45737.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:03:42,334 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 00:03:50,356 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=45760.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:03:51,238 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 00:03:51,884 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=45763.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:03:54,193 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 00:03:57,242 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=45774.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:04:01,595 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 00:04:02,584 INFO [train.py:901] (0/2) Epoch 17, batch 600, loss[loss=0.1369, simple_loss=0.2042, pruned_loss=0.03484, over 7011.00 frames. ], tot_loss[loss=0.16, simple_loss=0.2343, pruned_loss=0.04288, over 1367891.00 frames. ], batch size: 35, lr: 9.76e-03, grad_scale: 8.0 +2023-03-21 00:04:02,650 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=45785.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:04:18,377 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7452, 3.0909, 2.5740, 2.6784, 2.5031, 2.4609, 2.8863, 2.8040], + device='cuda:0'), covar=tensor([0.0901, 0.0952, 0.1069, 0.2698, 0.2244, 0.0760, 0.0991, 0.0788], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0045, 0.0051, 0.0045, 0.0043, 0.0044, 0.0045, 0.0043], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:04:19,240 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 00:04:20,356 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7187, 1.8262, 2.0990, 1.8243, 2.0238, 1.7962, 1.6416, 1.4502], + device='cuda:0'), covar=tensor([0.0328, 0.0247, 0.0084, 0.0141, 0.0287, 0.0232, 0.0266, 0.0267], + device='cuda:0'), in_proj_covar=tensor([0.0024, 0.0024, 0.0022, 0.0021, 0.0022, 0.0020, 0.0024, 0.0025], + device='cuda:0'), out_proj_covar=tensor([6.0393e-05, 6.0132e-05, 5.4565e-05, 5.0577e-05, 5.7141e-05, 5.3105e-05, + 5.8454e-05, 6.2433e-05], device='cuda:0') +2023-03-21 00:04:21,685 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.428e+02 1.996e+02 2.450e+02 2.921e+02 7.497e+02, threshold=4.900e+02, percent-clipped=1.0 +2023-03-21 00:04:26,384 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1008, 3.4991, 4.1083, 4.0083, 4.0326, 3.9297, 4.0553, 3.9156], + device='cuda:0'), covar=tensor([0.0021, 0.0081, 0.0029, 0.0031, 0.0028, 0.0028, 0.0029, 0.0041], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0051, 0.0045, 0.0043, 0.0043, 0.0045, 0.0044, 0.0055], + device='cuda:0'), out_proj_covar=tensor([8.1302e-05, 1.2796e-04, 1.0853e-04, 9.4927e-05, 9.4779e-05, 1.0021e-04, + 1.0891e-04, 1.2677e-04], device='cuda:0') +2023-03-21 00:04:28,293 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 00:04:29,297 INFO [train.py:901] (0/2) Epoch 17, batch 650, loss[loss=0.1436, simple_loss=0.2171, pruned_loss=0.0351, over 7173.00 frames. ], tot_loss[loss=0.1594, simple_loss=0.2343, pruned_loss=0.04226, over 1385389.19 frames. ], batch size: 39, lr: 9.76e-03, grad_scale: 8.0 +2023-03-21 00:04:29,465 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=45835.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:04:38,239 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2466, 0.9626, 1.4786, 1.6945, 1.5118, 1.7231, 1.5472, 1.7556], + device='cuda:0'), covar=tensor([0.2707, 0.3017, 0.0737, 0.0780, 0.1598, 0.1001, 0.1393, 0.0991], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0052, 0.0036, 0.0037, 0.0040, 0.0041, 0.0054, 0.0040], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 00:04:41,678 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=45859.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:04:46,475 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 00:04:48,591 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2149, 3.8040, 3.8564, 4.2816, 4.2525, 4.1897, 3.5295, 3.6838], + device='cuda:0'), covar=tensor([0.0929, 0.2766, 0.2504, 0.1100, 0.0827, 0.1354, 0.1041, 0.1293], + device='cuda:0'), in_proj_covar=tensor([0.0202, 0.0318, 0.0251, 0.0247, 0.0185, 0.0304, 0.0179, 0.0223], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:04:55,321 INFO [train.py:901] (0/2) Epoch 17, batch 700, loss[loss=0.1281, simple_loss=0.1958, pruned_loss=0.0302, over 6997.00 frames. ], tot_loss[loss=0.1579, simple_loss=0.2328, pruned_loss=0.0415, over 1397056.25 frames. ], batch size: 35, lr: 9.75e-03, grad_scale: 8.0 +2023-03-21 00:04:55,361 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 00:05:06,540 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1261, 0.9058, 1.3446, 1.5949, 1.4816, 1.6309, 1.4790, 1.6554], + device='cuda:0'), covar=tensor([0.2452, 0.3272, 0.0909, 0.1200, 0.2611, 0.1947, 0.1352, 0.2032], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0052, 0.0036, 0.0037, 0.0039, 0.0040, 0.0054, 0.0040], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 00:05:11,786 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-03-21 00:05:13,502 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.307e+02 2.037e+02 2.465e+02 3.045e+02 9.193e+02, threshold=4.929e+02, percent-clipped=2.0 +2023-03-21 00:05:13,661 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=45920.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:05:19,565 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 00:05:19,580 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 00:05:21,061 INFO [train.py:901] (0/2) Epoch 17, batch 750, loss[loss=0.1977, simple_loss=0.266, pruned_loss=0.06467, over 6665.00 frames. ], tot_loss[loss=0.1584, simple_loss=0.2337, pruned_loss=0.04161, over 1409030.34 frames. ], batch size: 106, lr: 9.75e-03, grad_scale: 8.0 +2023-03-21 00:05:33,146 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 00:05:37,746 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 00:05:44,489 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 00:05:46,058 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 00:05:47,538 INFO [train.py:901] (0/2) Epoch 17, batch 800, loss[loss=0.1873, simple_loss=0.2601, pruned_loss=0.05724, over 7273.00 frames. ], tot_loss[loss=0.1588, simple_loss=0.2342, pruned_loss=0.0417, over 1416649.28 frames. ], batch size: 70, lr: 9.74e-03, grad_scale: 8.0 +2023-03-21 00:05:57,457 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 00:05:57,558 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=46004.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 00:06:05,490 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.317e+02 2.247e+02 2.680e+02 3.304e+02 7.121e+02, threshold=5.361e+02, percent-clipped=5.0 +2023-03-21 00:06:12,892 INFO [train.py:901] (0/2) Epoch 17, batch 850, loss[loss=0.1536, simple_loss=0.2366, pruned_loss=0.03527, over 7266.00 frames. ], tot_loss[loss=0.1589, simple_loss=0.2345, pruned_loss=0.04168, over 1424081.47 frames. ], batch size: 55, lr: 9.74e-03, grad_scale: 8.0 +2023-03-21 00:06:16,007 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 00:06:16,012 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 00:06:21,587 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 00:06:22,260 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=46052.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 00:06:25,259 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=46058.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:06:25,713 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 00:06:39,613 INFO [train.py:901] (0/2) Epoch 17, batch 900, loss[loss=0.1555, simple_loss=0.2359, pruned_loss=0.03758, over 7357.00 frames. ], tot_loss[loss=0.1586, simple_loss=0.2343, pruned_loss=0.04141, over 1429311.97 frames. ], batch size: 54, lr: 9.73e-03, grad_scale: 8.0 +2023-03-21 00:06:56,109 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.07 vs. limit=2.0 +2023-03-21 00:06:57,339 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.352e+02 2.073e+02 2.414e+02 2.873e+02 6.303e+02, threshold=4.829e+02, percent-clipped=2.0 +2023-03-21 00:06:58,651 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-21 00:07:01,461 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8769, 5.4224, 5.4672, 5.3714, 5.1157, 4.9442, 5.4819, 5.2131], + device='cuda:0'), covar=tensor([0.0361, 0.0265, 0.0232, 0.0340, 0.0290, 0.0275, 0.0227, 0.0399], + device='cuda:0'), in_proj_covar=tensor([0.0179, 0.0206, 0.0151, 0.0152, 0.0125, 0.0189, 0.0162, 0.0123], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:07:02,441 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=46130.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:07:03,911 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 00:07:04,913 INFO [train.py:901] (0/2) Epoch 17, batch 950, loss[loss=0.1648, simple_loss=0.2445, pruned_loss=0.04254, over 7207.00 frames. ], tot_loss[loss=0.1592, simple_loss=0.235, pruned_loss=0.04175, over 1432955.61 frames. ], batch size: 93, lr: 9.73e-03, grad_scale: 8.0 +2023-03-21 00:07:25,292 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6395, 3.8137, 3.6120, 3.6601, 3.5949, 3.6543, 3.9290, 3.9568], + device='cuda:0'), covar=tensor([0.0218, 0.0162, 0.0200, 0.0200, 0.0269, 0.0395, 0.0285, 0.0215], + device='cuda:0'), in_proj_covar=tensor([0.0106, 0.0106, 0.0099, 0.0107, 0.0101, 0.0088, 0.0087, 0.0083], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:07:27,374 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9994, 4.0809, 3.6256, 3.4056, 3.2591, 2.4797, 1.8234, 3.9993], + device='cuda:0'), covar=tensor([0.0036, 0.0025, 0.0055, 0.0059, 0.0086, 0.0348, 0.0467, 0.0041], + device='cuda:0'), in_proj_covar=tensor([0.0084, 0.0070, 0.0091, 0.0078, 0.0100, 0.0116, 0.0114, 0.0083], + device='cuda:0'), out_proj_covar=tensor([1.0510e-04, 9.8629e-05, 1.1884e-04, 1.0612e-04, 1.2769e-04, 1.5209e-04, + 1.4980e-04, 1.0602e-04], device='cuda:0') +2023-03-21 00:07:28,301 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 00:07:31,211 INFO [train.py:901] (0/2) Epoch 17, batch 1000, loss[loss=0.1777, simple_loss=0.2502, pruned_loss=0.05266, over 7308.00 frames. ], tot_loss[loss=0.1596, simple_loss=0.2355, pruned_loss=0.0419, over 1435883.60 frames. ], batch size: 83, lr: 9.72e-03, grad_scale: 8.0 +2023-03-21 00:07:38,402 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5782, 1.3096, 1.5729, 1.9756, 1.6824, 2.1416, 1.9891, 2.0764], + device='cuda:0'), covar=tensor([0.1601, 0.2345, 0.1277, 0.0886, 0.3007, 0.1145, 0.1465, 0.2100], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0054, 0.0038, 0.0038, 0.0042, 0.0043, 0.0058, 0.0042], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 00:07:46,768 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=46215.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:07:49,125 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.429e+02 2.224e+02 2.568e+02 3.031e+02 6.347e+02, threshold=5.135e+02, percent-clipped=2.0 +2023-03-21 00:07:49,658 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 00:07:57,328 INFO [train.py:901] (0/2) Epoch 17, batch 1050, loss[loss=0.1478, simple_loss=0.2303, pruned_loss=0.03261, over 7310.00 frames. ], tot_loss[loss=0.1595, simple_loss=0.2348, pruned_loss=0.04206, over 1437114.16 frames. ], batch size: 86, lr: 9.71e-03, grad_scale: 8.0 +2023-03-21 00:08:12,584 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 00:08:16,630 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 00:08:18,736 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=46276.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:08:19,270 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0125, 0.7045, 1.2545, 1.4310, 1.3581, 1.4254, 1.4045, 1.5028], + device='cuda:0'), covar=tensor([0.1877, 0.4799, 0.1468, 0.0921, 0.1658, 0.1864, 0.2994, 0.1532], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0054, 0.0038, 0.0038, 0.0041, 0.0043, 0.0058, 0.0043], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 00:08:23,171 INFO [train.py:901] (0/2) Epoch 17, batch 1100, loss[loss=0.1477, simple_loss=0.2241, pruned_loss=0.03565, over 7297.00 frames. ], tot_loss[loss=0.1591, simple_loss=0.2345, pruned_loss=0.04191, over 1438138.35 frames. ], batch size: 49, lr: 9.71e-03, grad_scale: 8.0 +2023-03-21 00:08:41,467 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.290e+02 2.294e+02 2.723e+02 3.256e+02 7.179e+02, threshold=5.446e+02, percent-clipped=3.0 +2023-03-21 00:08:45,037 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 00:08:45,506 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 00:08:49,631 INFO [train.py:901] (0/2) Epoch 17, batch 1150, loss[loss=0.1506, simple_loss=0.2229, pruned_loss=0.03914, over 7341.00 frames. ], tot_loss[loss=0.1595, simple_loss=0.2347, pruned_loss=0.0421, over 1440023.02 frames. ], batch size: 44, lr: 9.70e-03, grad_scale: 8.0 +2023-03-21 00:08:50,815 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=46337.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:08:57,948 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 00:08:58,434 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 00:09:01,553 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=46358.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:09:04,670 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1500, 1.2222, 1.1301, 1.1351, 1.1845, 1.1400, 1.1394, 0.8857], + device='cuda:0'), covar=tensor([0.0134, 0.0094, 0.0145, 0.0077, 0.0103, 0.0105, 0.0150, 0.0121], + device='cuda:0'), in_proj_covar=tensor([0.0024, 0.0022, 0.0022, 0.0022, 0.0023, 0.0022, 0.0023, 0.0030], + device='cuda:0'), out_proj_covar=tensor([2.8426e-05, 2.5285e-05, 2.5724e-05, 2.4987e-05, 2.7372e-05, 2.4595e-05, + 2.6904e-05, 3.4649e-05], device='cuda:0') +2023-03-21 00:09:15,025 INFO [train.py:901] (0/2) Epoch 17, batch 1200, loss[loss=0.1669, simple_loss=0.2344, pruned_loss=0.04975, over 7289.00 frames. ], tot_loss[loss=0.1599, simple_loss=0.2352, pruned_loss=0.04232, over 1440114.27 frames. ], batch size: 42, lr: 9.70e-03, grad_scale: 8.0 +2023-03-21 00:09:26,570 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=46406.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:09:32,046 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 00:09:34,159 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.336e+02 2.112e+02 2.403e+02 2.943e+02 5.370e+02, threshold=4.806e+02, percent-clipped=0.0 +2023-03-21 00:09:39,336 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=46430.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:09:41,787 INFO [train.py:901] (0/2) Epoch 17, batch 1250, loss[loss=0.1447, simple_loss=0.2211, pruned_loss=0.03411, over 7152.00 frames. ], tot_loss[loss=0.1589, simple_loss=0.2339, pruned_loss=0.04199, over 1437831.90 frames. ], batch size: 41, lr: 9.69e-03, grad_scale: 8.0 +2023-03-21 00:09:55,486 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 00:09:55,539 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5038, 4.0513, 3.9949, 4.4800, 4.4530, 4.4534, 4.0187, 4.0801], + device='cuda:0'), covar=tensor([0.0795, 0.2631, 0.2374, 0.1138, 0.0826, 0.1364, 0.0724, 0.1057], + device='cuda:0'), in_proj_covar=tensor([0.0205, 0.0326, 0.0252, 0.0253, 0.0182, 0.0312, 0.0183, 0.0227], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:09:58,986 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 00:10:00,496 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 00:10:03,544 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=46478.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:10:07,011 INFO [train.py:901] (0/2) Epoch 17, batch 1300, loss[loss=0.1567, simple_loss=0.237, pruned_loss=0.03824, over 7350.00 frames. ], tot_loss[loss=0.1584, simple_loss=0.2336, pruned_loss=0.0416, over 1437262.12 frames. ], batch size: 63, lr: 9.69e-03, grad_scale: 8.0 +2023-03-21 00:10:14,654 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.57 vs. limit=5.0 +2023-03-21 00:10:23,630 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=46515.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:10:25,023 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 00:10:26,037 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.425e+02 2.102e+02 2.512e+02 2.983e+02 5.030e+02, threshold=5.025e+02, percent-clipped=2.0 +2023-03-21 00:10:27,077 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 00:10:30,652 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 00:10:33,722 INFO [train.py:901] (0/2) Epoch 17, batch 1350, loss[loss=0.1811, simple_loss=0.2577, pruned_loss=0.05225, over 7377.00 frames. ], tot_loss[loss=0.1583, simple_loss=0.2337, pruned_loss=0.04139, over 1440017.28 frames. ], batch size: 65, lr: 9.68e-03, grad_scale: 8.0 +2023-03-21 00:10:38,616 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.51 vs. limit=5.0 +2023-03-21 00:10:40,786 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 00:10:47,917 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=46563.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:11:00,122 INFO [train.py:901] (0/2) Epoch 17, batch 1400, loss[loss=0.1303, simple_loss=0.2102, pruned_loss=0.02524, over 7334.00 frames. ], tot_loss[loss=0.1585, simple_loss=0.2341, pruned_loss=0.04148, over 1440328.17 frames. ], batch size: 44, lr: 9.68e-03, grad_scale: 8.0 +2023-03-21 00:11:05,304 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=46595.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:11:09,137 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=46602.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:11:14,080 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 00:11:17,904 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.456e+02 2.103e+02 2.396e+02 2.900e+02 8.640e+02, threshold=4.792e+02, percent-clipped=2.0 +2023-03-21 00:11:18,617 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1241, 3.1439, 2.8952, 3.0106, 2.3870, 2.5010, 3.1732, 2.3812], + device='cuda:0'), covar=tensor([0.0340, 0.0335, 0.0307, 0.0309, 0.0355, 0.0516, 0.0496, 0.0864], + device='cuda:0'), in_proj_covar=tensor([0.0315, 0.0316, 0.0255, 0.0340, 0.0306, 0.0307, 0.0322, 0.0296], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 00:11:23,939 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=46632.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:11:25,397 INFO [train.py:901] (0/2) Epoch 17, batch 1450, loss[loss=0.1841, simple_loss=0.2545, pruned_loss=0.05686, over 7238.00 frames. ], tot_loss[loss=0.1581, simple_loss=0.2336, pruned_loss=0.04132, over 1441125.03 frames. ], batch size: 93, lr: 9.67e-03, grad_scale: 8.0 +2023-03-21 00:11:35,572 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2384, 0.8688, 1.3907, 1.5704, 1.3350, 1.6710, 1.4948, 1.5993], + device='cuda:0'), covar=tensor([0.1329, 0.2482, 0.1385, 0.0727, 0.2659, 0.1319, 0.1126, 0.1885], + device='cuda:0'), in_proj_covar=tensor([0.0045, 0.0055, 0.0038, 0.0037, 0.0041, 0.0042, 0.0058, 0.0042], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 00:11:36,091 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=46656.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 00:11:38,524 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 00:11:40,175 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=46663.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:11:51,667 INFO [train.py:901] (0/2) Epoch 17, batch 1500, loss[loss=0.1563, simple_loss=0.2318, pruned_loss=0.04034, over 7282.00 frames. ], tot_loss[loss=0.1584, simple_loss=0.234, pruned_loss=0.04141, over 1444107.83 frames. ], batch size: 70, lr: 9.67e-03, grad_scale: 8.0 +2023-03-21 00:11:54,196 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 00:11:56,635 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-21 00:12:09,334 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.385e+02 2.167e+02 2.620e+02 3.081e+02 7.014e+02, threshold=5.239e+02, percent-clipped=3.0 +2023-03-21 00:12:15,637 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.76 vs. limit=2.0 +2023-03-21 00:12:16,862 INFO [train.py:901] (0/2) Epoch 17, batch 1550, loss[loss=0.1481, simple_loss=0.2238, pruned_loss=0.03615, over 7279.00 frames. ], tot_loss[loss=0.158, simple_loss=0.2337, pruned_loss=0.04113, over 1443787.62 frames. ], batch size: 70, lr: 9.66e-03, grad_scale: 8.0 +2023-03-21 00:12:18,283 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 00:12:21,333 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.1308, 4.6446, 4.6504, 5.1605, 5.1111, 5.1126, 4.4887, 4.7046], + device='cuda:0'), covar=tensor([0.0647, 0.2597, 0.2084, 0.0807, 0.0628, 0.1123, 0.0625, 0.1065], + device='cuda:0'), in_proj_covar=tensor([0.0206, 0.0324, 0.0251, 0.0251, 0.0181, 0.0311, 0.0183, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:12:27,552 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9832, 2.8844, 2.0830, 3.5482, 2.1031, 2.9121, 1.6798, 1.8193], + device='cuda:0'), covar=tensor([0.0287, 0.0575, 0.2214, 0.0450, 0.0464, 0.0601, 0.2983, 0.1877], + device='cuda:0'), in_proj_covar=tensor([0.0251, 0.0235, 0.0302, 0.0243, 0.0257, 0.0250, 0.0264, 0.0284], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 00:12:43,157 INFO [train.py:901] (0/2) Epoch 17, batch 1600, loss[loss=0.1518, simple_loss=0.2277, pruned_loss=0.038, over 7233.00 frames. ], tot_loss[loss=0.158, simple_loss=0.2336, pruned_loss=0.04118, over 1445214.01 frames. ], batch size: 45, lr: 9.66e-03, grad_scale: 8.0 +2023-03-21 00:12:51,698 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 00:12:52,746 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 00:12:55,715 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 00:12:56,305 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=46810.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:13:01,322 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.537e+02 2.076e+02 2.534e+02 3.354e+02 1.323e+03, threshold=5.067e+02, percent-clipped=3.0 +2023-03-21 00:13:01,611 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.53 vs. limit=5.0 +2023-03-21 00:13:03,047 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9203, 4.0601, 3.7503, 3.9738, 3.7741, 3.9982, 4.3277, 4.3620], + device='cuda:0'), covar=tensor([0.0207, 0.0140, 0.0207, 0.0183, 0.0313, 0.0295, 0.0242, 0.0188], + device='cuda:0'), in_proj_covar=tensor([0.0108, 0.0107, 0.0100, 0.0108, 0.0101, 0.0090, 0.0088, 0.0085], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:13:04,999 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 00:13:08,995 INFO [train.py:901] (0/2) Epoch 17, batch 1650, loss[loss=0.1484, simple_loss=0.2238, pruned_loss=0.03647, over 7320.00 frames. ], tot_loss[loss=0.158, simple_loss=0.2337, pruned_loss=0.04114, over 1445631.81 frames. ], batch size: 80, lr: 9.65e-03, grad_scale: 8.0 +2023-03-21 00:13:09,629 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 00:13:17,778 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 00:13:21,872 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1897, 3.7272, 3.8348, 3.9109, 3.6995, 3.7509, 4.0178, 3.5887], + device='cuda:0'), covar=tensor([0.0154, 0.0143, 0.0123, 0.0137, 0.0427, 0.0131, 0.0164, 0.0174], + device='cuda:0'), in_proj_covar=tensor([0.0070, 0.0076, 0.0076, 0.0067, 0.0134, 0.0088, 0.0082, 0.0084], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:13:23,137 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.75 vs. limit=5.0 +2023-03-21 00:13:28,449 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=46871.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:13:34,817 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 00:13:35,340 INFO [train.py:901] (0/2) Epoch 17, batch 1700, loss[loss=0.1067, simple_loss=0.1704, pruned_loss=0.02151, over 6366.00 frames. ], tot_loss[loss=0.158, simple_loss=0.2334, pruned_loss=0.0413, over 1442995.84 frames. ], batch size: 27, lr: 9.65e-03, grad_scale: 8.0 +2023-03-21 00:13:35,500 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3427, 1.5372, 1.4973, 1.2602, 1.4678, 1.3169, 1.3795, 0.9470], + device='cuda:0'), covar=tensor([0.0074, 0.0097, 0.0120, 0.0157, 0.0093, 0.0077, 0.0163, 0.0196], + device='cuda:0'), in_proj_covar=tensor([0.0024, 0.0022, 0.0021, 0.0022, 0.0022, 0.0021, 0.0023, 0.0029], + device='cuda:0'), out_proj_covar=tensor([2.8515e-05, 2.4627e-05, 2.4573e-05, 2.4636e-05, 2.6786e-05, 2.3615e-05, + 2.6393e-05, 3.4354e-05], device='cuda:0') +2023-03-21 00:13:39,507 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 00:13:50,346 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 00:13:52,747 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.281e+02 2.347e+02 2.673e+02 3.126e+02 7.111e+02, threshold=5.346e+02, percent-clipped=5.0 +2023-03-21 00:13:53,472 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9046, 3.0608, 2.7058, 3.0434, 2.9996, 2.6305, 3.0789, 2.7513], + device='cuda:0'), covar=tensor([0.0722, 0.0666, 0.0613, 0.0695, 0.1417, 0.0933, 0.1415, 0.0881], + device='cuda:0'), in_proj_covar=tensor([0.0046, 0.0047, 0.0052, 0.0046, 0.0046, 0.0044, 0.0046, 0.0043], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:14:00,309 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=46932.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:14:01,727 INFO [train.py:901] (0/2) Epoch 17, batch 1750, loss[loss=0.1864, simple_loss=0.2575, pruned_loss=0.05768, over 7311.00 frames. ], tot_loss[loss=0.1583, simple_loss=0.2336, pruned_loss=0.04152, over 1443984.41 frames. ], batch size: 49, lr: 9.64e-03, grad_scale: 8.0 +2023-03-21 00:14:02,882 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3237, 1.5286, 1.5308, 1.2925, 1.3343, 1.3566, 1.4477, 1.1006], + device='cuda:0'), covar=tensor([0.0160, 0.0121, 0.0204, 0.0115, 0.0104, 0.0138, 0.0134, 0.0112], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0022, 0.0021, 0.0022, 0.0023, 0.0021, 0.0023, 0.0030], + device='cuda:0'), out_proj_covar=tensor([2.9142e-05, 2.4957e-05, 2.5164e-05, 2.4875e-05, 2.6915e-05, 2.3979e-05, + 2.6494e-05, 3.4687e-05], device='cuda:0') +2023-03-21 00:14:06,175 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.82 vs. limit=5.0 +2023-03-21 00:14:09,815 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=46951.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 00:14:13,378 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=46958.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:14:15,807 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 00:14:16,815 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 00:14:24,197 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=46980.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:14:26,628 INFO [train.py:901] (0/2) Epoch 17, batch 1800, loss[loss=0.1553, simple_loss=0.2345, pruned_loss=0.03809, over 7301.00 frames. ], tot_loss[loss=0.1585, simple_loss=0.2339, pruned_loss=0.04151, over 1443216.46 frames. ], batch size: 68, lr: 9.64e-03, grad_scale: 8.0 +2023-03-21 00:14:37,563 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 00:14:45,780 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.279e+02 2.122e+02 2.479e+02 3.047e+02 5.778e+02, threshold=4.959e+02, percent-clipped=1.0 +2023-03-21 00:14:51,960 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 00:14:53,481 INFO [train.py:901] (0/2) Epoch 17, batch 1850, loss[loss=0.147, simple_loss=0.2283, pruned_loss=0.03287, over 7353.00 frames. ], tot_loss[loss=0.1583, simple_loss=0.2341, pruned_loss=0.04129, over 1445058.48 frames. ], batch size: 63, lr: 9.63e-03, grad_scale: 8.0 +2023-03-21 00:15:01,527 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 00:15:09,945 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.40 vs. limit=5.0 +2023-03-21 00:15:13,911 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-21 00:15:17,644 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 00:15:18,608 INFO [train.py:901] (0/2) Epoch 17, batch 1900, loss[loss=0.1247, simple_loss=0.186, pruned_loss=0.03169, over 6130.00 frames. ], tot_loss[loss=0.1578, simple_loss=0.234, pruned_loss=0.04082, over 1446205.15 frames. ], batch size: 27, lr: 9.63e-03, grad_scale: 8.0 +2023-03-21 00:15:22,484 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.77 vs. limit=5.0 +2023-03-21 00:15:37,426 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.393e+02 2.079e+02 2.464e+02 3.016e+02 5.174e+02, threshold=4.929e+02, percent-clipped=1.0 +2023-03-21 00:15:43,001 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.90 vs. limit=5.0 +2023-03-21 00:15:44,651 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 00:15:45,136 INFO [train.py:901] (0/2) Epoch 17, batch 1950, loss[loss=0.1614, simple_loss=0.2381, pruned_loss=0.04238, over 7311.00 frames. ], tot_loss[loss=0.1576, simple_loss=0.2336, pruned_loss=0.04082, over 1446183.63 frames. ], batch size: 83, lr: 9.62e-03, grad_scale: 16.0 +2023-03-21 00:15:54,677 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 00:15:59,268 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 00:15:59,763 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 00:16:00,843 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=47166.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:16:01,875 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4812, 4.6768, 4.3954, 4.5759, 4.3866, 4.6595, 4.9943, 4.9362], + device='cuda:0'), covar=tensor([0.0172, 0.0127, 0.0192, 0.0162, 0.0286, 0.0171, 0.0163, 0.0189], + device='cuda:0'), in_proj_covar=tensor([0.0112, 0.0110, 0.0104, 0.0111, 0.0104, 0.0093, 0.0090, 0.0089], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:16:02,145 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.85 vs. limit=2.0 +2023-03-21 00:16:10,990 INFO [train.py:901] (0/2) Epoch 17, batch 2000, loss[loss=0.1391, simple_loss=0.2192, pruned_loss=0.0295, over 7308.00 frames. ], tot_loss[loss=0.1577, simple_loss=0.2337, pruned_loss=0.04083, over 1445014.84 frames. ], batch size: 75, lr: 9.62e-03, grad_scale: 16.0 +2023-03-21 00:16:16,646 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 00:16:17,236 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1520, 3.6336, 4.0210, 3.5775, 4.0926, 3.9779, 3.9565, 3.9086], + device='cuda:0'), covar=tensor([0.0025, 0.0077, 0.0029, 0.0042, 0.0027, 0.0028, 0.0031, 0.0048], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0052, 0.0045, 0.0044, 0.0044, 0.0047, 0.0046, 0.0057], + device='cuda:0'), out_proj_covar=tensor([8.1660e-05, 1.3017e-04, 1.0774e-04, 9.7828e-05, 9.7326e-05, 1.0216e-04, + 1.1118e-04, 1.2957e-04], device='cuda:0') +2023-03-21 00:16:18,945 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.44 vs. limit=2.0 +2023-03-21 00:16:28,347 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 00:16:29,341 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.291e+02 2.033e+02 2.454e+02 2.931e+02 5.819e+02, threshold=4.907e+02, percent-clipped=3.0 +2023-03-21 00:16:30,545 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0510, 2.3477, 1.9063, 3.0554, 2.6185, 2.2884, 2.3786, 2.7105], + device='cuda:0'), covar=tensor([0.2012, 0.0726, 0.3127, 0.0482, 0.0128, 0.0067, 0.0178, 0.0249], + device='cuda:0'), in_proj_covar=tensor([0.0284, 0.0238, 0.0272, 0.0265, 0.0149, 0.0144, 0.0175, 0.0196], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:16:35,105 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0268, 4.1166, 3.9434, 4.0193, 3.8985, 4.1585, 4.4751, 4.4819], + device='cuda:0'), covar=tensor([0.0204, 0.0144, 0.0195, 0.0173, 0.0297, 0.0303, 0.0228, 0.0196], + device='cuda:0'), in_proj_covar=tensor([0.0110, 0.0109, 0.0102, 0.0109, 0.0103, 0.0092, 0.0089, 0.0087], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:16:36,566 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 00:16:37,053 INFO [train.py:901] (0/2) Epoch 17, batch 2050, loss[loss=0.1649, simple_loss=0.2372, pruned_loss=0.04631, over 7271.00 frames. ], tot_loss[loss=0.1578, simple_loss=0.2339, pruned_loss=0.04085, over 1446509.59 frames. ], batch size: 77, lr: 9.61e-03, grad_scale: 16.0 +2023-03-21 00:16:40,253 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=47241.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:16:45,220 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=47251.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:16:48,638 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=47258.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:17:03,231 INFO [train.py:901] (0/2) Epoch 17, batch 2100, loss[loss=0.1508, simple_loss=0.2309, pruned_loss=0.03541, over 7277.00 frames. ], tot_loss[loss=0.1583, simple_loss=0.2343, pruned_loss=0.04117, over 1446032.99 frames. ], batch size: 57, lr: 9.61e-03, grad_scale: 8.0 +2023-03-21 00:17:10,522 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=47299.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:17:11,466 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 00:17:12,043 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=47302.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:17:13,977 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=47306.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:17:14,453 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 00:17:19,753 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-21 00:17:21,501 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.466e+02 2.164e+02 2.514e+02 3.054e+02 5.580e+02, threshold=5.028e+02, percent-clipped=1.0 +2023-03-21 00:17:28,636 INFO [train.py:901] (0/2) Epoch 17, batch 2150, loss[loss=0.1308, simple_loss=0.1934, pruned_loss=0.03407, over 5962.00 frames. ], tot_loss[loss=0.158, simple_loss=0.2338, pruned_loss=0.0411, over 1443737.12 frames. ], batch size: 26, lr: 9.60e-03, grad_scale: 8.0 +2023-03-21 00:17:52,399 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=47379.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:17:55,210 INFO [train.py:901] (0/2) Epoch 17, batch 2200, loss[loss=0.1456, simple_loss=0.2262, pruned_loss=0.03249, over 7370.00 frames. ], tot_loss[loss=0.1583, simple_loss=0.2341, pruned_loss=0.04125, over 1442414.76 frames. ], batch size: 73, lr: 9.60e-03, grad_scale: 8.0 +2023-03-21 00:18:00,303 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 00:18:13,749 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.428e+02 2.138e+02 2.450e+02 3.196e+02 6.599e+02, threshold=4.899e+02, percent-clipped=2.0 +2023-03-21 00:18:20,806 INFO [train.py:901] (0/2) Epoch 17, batch 2250, loss[loss=0.1776, simple_loss=0.2495, pruned_loss=0.05282, over 7327.00 frames. ], tot_loss[loss=0.1586, simple_loss=0.2346, pruned_loss=0.04132, over 1443912.94 frames. ], batch size: 49, lr: 9.59e-03, grad_scale: 8.0 +2023-03-21 00:18:23,516 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=47440.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:18:35,219 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 00:18:35,701 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 00:18:37,805 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=47466.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:18:43,227 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.62 vs. limit=5.0 +2023-03-21 00:18:47,234 INFO [train.py:901] (0/2) Epoch 17, batch 2300, loss[loss=0.1618, simple_loss=0.236, pruned_loss=0.04378, over 7340.00 frames. ], tot_loss[loss=0.1584, simple_loss=0.2343, pruned_loss=0.04126, over 1443021.33 frames. ], batch size: 63, lr: 9.59e-03, grad_scale: 8.0 +2023-03-21 00:18:48,229 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 00:19:01,774 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=47514.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:19:05,250 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.279e+02 2.078e+02 2.405e+02 2.974e+02 9.251e+02, threshold=4.809e+02, percent-clipped=4.0 +2023-03-21 00:19:09,473 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0505, 2.7373, 3.0325, 2.7130, 3.1937, 2.6933, 2.4880, 3.0672], + device='cuda:0'), covar=tensor([0.1577, 0.0721, 0.1952, 0.2771, 0.1018, 0.2187, 0.2646, 0.1620], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0049, 0.0039, 0.0040, 0.0038, 0.0037, 0.0055, 0.0042], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 00:19:12,305 INFO [train.py:901] (0/2) Epoch 17, batch 2350, loss[loss=0.1621, simple_loss=0.2375, pruned_loss=0.04333, over 7251.00 frames. ], tot_loss[loss=0.1587, simple_loss=0.2346, pruned_loss=0.04143, over 1444946.80 frames. ], batch size: 89, lr: 9.58e-03, grad_scale: 8.0 +2023-03-21 00:19:19,926 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.77 vs. limit=2.0 +2023-03-21 00:19:34,548 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 00:19:38,528 INFO [train.py:901] (0/2) Epoch 17, batch 2400, loss[loss=0.1605, simple_loss=0.2408, pruned_loss=0.04008, over 7300.00 frames. ], tot_loss[loss=0.159, simple_loss=0.2349, pruned_loss=0.04155, over 1444870.19 frames. ], batch size: 80, lr: 9.58e-03, grad_scale: 8.0 +2023-03-21 00:19:41,113 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 00:19:44,639 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=47597.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:19:52,381 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 00:19:55,394 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 00:19:56,803 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.471e+02 2.109e+02 2.488e+02 3.025e+02 6.588e+02, threshold=4.976e+02, percent-clipped=2.0 +2023-03-21 00:20:05,131 INFO [train.py:901] (0/2) Epoch 17, batch 2450, loss[loss=0.1479, simple_loss=0.2197, pruned_loss=0.03805, over 7346.00 frames. ], tot_loss[loss=0.1584, simple_loss=0.2341, pruned_loss=0.04136, over 1445630.79 frames. ], batch size: 54, lr: 9.57e-03, grad_scale: 8.0 +2023-03-21 00:20:22,004 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 00:20:30,618 INFO [train.py:901] (0/2) Epoch 17, batch 2500, loss[loss=0.1607, simple_loss=0.2328, pruned_loss=0.04427, over 7343.00 frames. ], tot_loss[loss=0.1591, simple_loss=0.2348, pruned_loss=0.0417, over 1446109.68 frames. ], batch size: 51, lr: 9.57e-03, grad_scale: 8.0 +2023-03-21 00:20:34,537 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.01 vs. limit=2.0 +2023-03-21 00:20:47,595 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 00:20:50,036 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.424e+02 2.240e+02 2.665e+02 3.286e+02 7.455e+02, threshold=5.331e+02, percent-clipped=3.0 +2023-03-21 00:20:57,348 INFO [train.py:901] (0/2) Epoch 17, batch 2550, loss[loss=0.1545, simple_loss=0.2316, pruned_loss=0.03872, over 7292.00 frames. ], tot_loss[loss=0.159, simple_loss=0.2347, pruned_loss=0.04166, over 1443993.65 frames. ], batch size: 66, lr: 9.56e-03, grad_scale: 8.0 +2023-03-21 00:20:57,426 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=47735.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:20:58,992 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=47738.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:21:07,165 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2616, 4.7921, 4.8003, 4.7582, 4.6693, 4.3185, 4.8589, 4.6557], + device='cuda:0'), covar=tensor([0.0489, 0.0417, 0.0462, 0.0443, 0.0349, 0.0385, 0.0384, 0.0445], + device='cuda:0'), in_proj_covar=tensor([0.0184, 0.0211, 0.0153, 0.0155, 0.0129, 0.0194, 0.0164, 0.0124], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:21:22,594 INFO [train.py:901] (0/2) Epoch 17, batch 2600, loss[loss=0.1696, simple_loss=0.2426, pruned_loss=0.04834, over 7337.00 frames. ], tot_loss[loss=0.159, simple_loss=0.2344, pruned_loss=0.0418, over 1441844.51 frames. ], batch size: 51, lr: 9.56e-03, grad_scale: 8.0 +2023-03-21 00:21:24,214 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=47788.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:21:29,808 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=47799.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:21:32,041 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3543, 2.5689, 2.2555, 2.5979, 2.5719, 2.2648, 2.6520, 2.4566], + device='cuda:0'), covar=tensor([0.0768, 0.0477, 0.0971, 0.0625, 0.1018, 0.0390, 0.0587, 0.0600], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0043, 0.0049, 0.0044, 0.0042, 0.0042, 0.0045, 0.0041], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:21:34,939 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=47809.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 00:21:41,943 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.493e+02 2.119e+02 2.477e+02 3.299e+02 7.952e+02, threshold=4.955e+02, percent-clipped=2.0 +2023-03-21 00:21:48,754 INFO [train.py:901] (0/2) Epoch 17, batch 2650, loss[loss=0.1746, simple_loss=0.2473, pruned_loss=0.05096, over 7240.00 frames. ], tot_loss[loss=0.1593, simple_loss=0.2348, pruned_loss=0.04193, over 1442702.24 frames. ], batch size: 55, lr: 9.55e-03, grad_scale: 8.0 +2023-03-21 00:21:55,806 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=47849.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:22:00,392 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.82 vs. limit=2.0 +2023-03-21 00:22:06,148 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=47870.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 00:22:13,286 INFO [train.py:901] (0/2) Epoch 17, batch 2700, loss[loss=0.1698, simple_loss=0.2475, pruned_loss=0.04603, over 7296.00 frames. ], tot_loss[loss=0.1594, simple_loss=0.2351, pruned_loss=0.04186, over 1444086.75 frames. ], batch size: 49, lr: 9.55e-03, grad_scale: 8.0 +2023-03-21 00:22:19,222 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=47897.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:22:30,937 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.375e+02 2.095e+02 2.437e+02 3.021e+02 7.441e+02, threshold=4.873e+02, percent-clipped=1.0 +2023-03-21 00:22:37,933 INFO [train.py:901] (0/2) Epoch 17, batch 2750, loss[loss=0.1537, simple_loss=0.23, pruned_loss=0.03874, over 7341.00 frames. ], tot_loss[loss=0.1584, simple_loss=0.2344, pruned_loss=0.04123, over 1441306.88 frames. ], batch size: 54, lr: 9.54e-03, grad_scale: 8.0 +2023-03-21 00:22:43,009 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=47945.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:23:02,689 INFO [train.py:901] (0/2) Epoch 17, batch 2800, loss[loss=0.1421, simple_loss=0.2095, pruned_loss=0.03731, over 6958.00 frames. ], tot_loss[loss=0.158, simple_loss=0.2341, pruned_loss=0.04101, over 1440880.03 frames. ], batch size: 35, lr: 9.54e-03, grad_scale: 8.0 +2023-03-21 00:23:10,476 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-48000.pt +2023-03-21 00:23:19,935 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-17.pt +2023-03-21 00:23:37,088 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 00:23:37,141 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 00:23:37,166 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 00:23:37,185 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 00:23:37,677 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 00:23:37,693 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 00:23:37,704 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 00:23:37,804 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 00:23:37,879 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 00:23:37,894 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 00:23:37,911 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 00:23:38,075 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 00:23:38,080 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 00:23:38,088 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 00:23:38,228 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 00:23:38,231 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 00:23:38,290 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 00:23:40,437 INFO [train.py:901] (0/2) Epoch 18, batch 0, loss[loss=0.1516, simple_loss=0.2224, pruned_loss=0.04047, over 7329.00 frames. ], tot_loss[loss=0.1516, simple_loss=0.2224, pruned_loss=0.04047, over 7329.00 frames. ], batch size: 59, lr: 9.28e-03, grad_scale: 8.0 +2023-03-21 00:23:40,438 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 00:24:06,020 INFO [train.py:935] (0/2) Epoch 18, validation: loss=0.1665, simple_loss=0.253, pruned_loss=0.03999, over 1622729.00 frames. +2023-03-21 00:24:06,020 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 00:24:12,049 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.469e+02 1.972e+02 2.348e+02 2.792e+02 4.861e+02, threshold=4.696e+02, percent-clipped=0.0 +2023-03-21 00:24:13,089 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 00:24:19,062 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=48035.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:24:24,376 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5465, 2.1314, 2.3049, 1.7262, 2.0734, 1.7950, 1.8441, 1.6521], + device='cuda:0'), covar=tensor([0.0433, 0.0274, 0.0100, 0.0108, 0.0292, 0.0267, 0.0143, 0.0203], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0025, 0.0022, 0.0022, 0.0023, 0.0023, 0.0026, 0.0026], + device='cuda:0'), out_proj_covar=tensor([6.4190e-05, 6.3239e-05, 5.5479e-05, 5.4672e-05, 6.0533e-05, 5.8236e-05, + 6.2262e-05, 6.6691e-05], device='cuda:0') +2023-03-21 00:24:24,816 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 00:24:30,556 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2263, 0.8264, 1.3317, 1.5033, 1.3181, 1.7000, 1.2776, 1.5195], + device='cuda:0'), covar=tensor([0.1096, 0.2781, 0.2214, 0.1292, 0.1440, 0.0961, 0.1203, 0.1242], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0059, 0.0039, 0.0040, 0.0041, 0.0043, 0.0060, 0.0042], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 00:24:31,984 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 00:24:32,464 INFO [train.py:901] (0/2) Epoch 18, batch 50, loss[loss=0.1599, simple_loss=0.2377, pruned_loss=0.04107, over 7264.00 frames. ], tot_loss[loss=0.1551, simple_loss=0.2307, pruned_loss=0.03975, over 324643.87 frames. ], batch size: 47, lr: 9.28e-03, grad_scale: 8.0 +2023-03-21 00:24:34,521 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 00:24:37,163 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 00:24:39,294 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=48072.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:24:44,727 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=48083.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:24:50,370 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=48094.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:24:50,419 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9198, 3.9553, 3.3115, 3.3943, 3.3046, 2.3522, 1.7051, 4.0411], + device='cuda:0'), covar=tensor([0.0040, 0.0053, 0.0091, 0.0061, 0.0096, 0.0387, 0.0520, 0.0041], + device='cuda:0'), in_proj_covar=tensor([0.0084, 0.0073, 0.0090, 0.0078, 0.0099, 0.0114, 0.0113, 0.0083], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 00:24:53,899 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 00:24:54,393 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 00:24:57,972 INFO [train.py:901] (0/2) Epoch 18, batch 100, loss[loss=0.1549, simple_loss=0.2287, pruned_loss=0.04053, over 7235.00 frames. ], tot_loss[loss=0.1555, simple_loss=0.2321, pruned_loss=0.03945, over 574853.88 frames. ], batch size: 45, lr: 9.27e-03, grad_scale: 8.0 +2023-03-21 00:25:03,115 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8610, 2.2841, 1.8817, 2.7701, 2.3938, 2.2767, 2.3908, 2.2177], + device='cuda:0'), covar=tensor([0.1914, 0.0907, 0.2925, 0.0405, 0.0106, 0.0070, 0.0243, 0.0173], + device='cuda:0'), in_proj_covar=tensor([0.0279, 0.0236, 0.0269, 0.0261, 0.0150, 0.0145, 0.0177, 0.0194], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:25:03,906 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.597e+02 2.080e+02 2.464e+02 3.092e+02 6.621e+02, threshold=4.927e+02, percent-clipped=4.0 +2023-03-21 00:25:10,735 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=48133.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:25:12,233 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2380, 3.7371, 3.9603, 3.9644, 3.7903, 3.8699, 4.1360, 3.6632], + device='cuda:0'), covar=tensor([0.0144, 0.0159, 0.0114, 0.0131, 0.0424, 0.0117, 0.0149, 0.0155], + device='cuda:0'), in_proj_covar=tensor([0.0074, 0.0081, 0.0079, 0.0072, 0.0142, 0.0093, 0.0087, 0.0088], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:25:16,459 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0579, 1.3689, 1.0413, 1.3304, 1.4105, 1.2547, 1.4008, 0.9952], + device='cuda:0'), covar=tensor([0.0237, 0.0108, 0.0293, 0.0095, 0.0110, 0.0107, 0.0121, 0.0162], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0022, 0.0022, 0.0021, 0.0022, 0.0022, 0.0023, 0.0029], + device='cuda:0'), out_proj_covar=tensor([2.9535e-05, 2.5183e-05, 2.6206e-05, 2.4314e-05, 2.6926e-05, 2.4317e-05, + 2.6928e-05, 3.4362e-05], device='cuda:0') +2023-03-21 00:25:16,848 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=48144.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:25:20,015 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9162, 2.3793, 1.7442, 2.8583, 2.3955, 2.2139, 2.2907, 2.3056], + device='cuda:0'), covar=tensor([0.1875, 0.0844, 0.3212, 0.0622, 0.0111, 0.0080, 0.0226, 0.0212], + device='cuda:0'), in_proj_covar=tensor([0.0278, 0.0235, 0.0267, 0.0261, 0.0150, 0.0145, 0.0176, 0.0193], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:25:24,274 INFO [train.py:901] (0/2) Epoch 18, batch 150, loss[loss=0.131, simple_loss=0.2099, pruned_loss=0.02608, over 7148.00 frames. ], tot_loss[loss=0.1583, simple_loss=0.2347, pruned_loss=0.04094, over 766304.45 frames. ], batch size: 41, lr: 9.27e-03, grad_scale: 8.0 +2023-03-21 00:25:27,360 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=48165.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 00:25:47,316 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1338, 2.8046, 3.2527, 3.1469, 3.2394, 2.9312, 2.4084, 3.1648], + device='cuda:0'), covar=tensor([0.1271, 0.0650, 0.1235, 0.1105, 0.0673, 0.1391, 0.2731, 0.1263], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0050, 0.0039, 0.0039, 0.0038, 0.0036, 0.0055, 0.0040], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 00:25:49,729 INFO [train.py:901] (0/2) Epoch 18, batch 200, loss[loss=0.164, simple_loss=0.2436, pruned_loss=0.04219, over 7347.00 frames. ], tot_loss[loss=0.1569, simple_loss=0.2331, pruned_loss=0.04036, over 916703.95 frames. ], batch size: 54, lr: 9.26e-03, grad_scale: 8.0 +2023-03-21 00:25:51,877 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=48212.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 00:25:55,339 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 00:25:56,914 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.462e+02 2.110e+02 2.629e+02 3.068e+02 6.550e+02, threshold=5.258e+02, percent-clipped=1.0 +2023-03-21 00:25:59,618 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7108, 2.1063, 1.6816, 2.6387, 2.1329, 2.1943, 1.9916, 2.0670], + device='cuda:0'), covar=tensor([0.1850, 0.0895, 0.2927, 0.0541, 0.0093, 0.0094, 0.0166, 0.0155], + device='cuda:0'), in_proj_covar=tensor([0.0279, 0.0236, 0.0265, 0.0262, 0.0150, 0.0145, 0.0176, 0.0193], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:25:59,960 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 00:26:06,468 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 00:26:06,601 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7890, 3.0819, 2.4799, 4.0216, 1.7189, 3.8357, 1.6678, 3.0568], + device='cuda:0'), covar=tensor([0.0079, 0.0678, 0.1415, 0.0091, 0.3579, 0.0161, 0.1034, 0.0300], + device='cuda:0'), in_proj_covar=tensor([0.0168, 0.0269, 0.0292, 0.0172, 0.0276, 0.0185, 0.0263, 0.0221], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 00:26:15,925 INFO [train.py:901] (0/2) Epoch 18, batch 250, loss[loss=0.1232, simple_loss=0.1977, pruned_loss=0.02436, over 7180.00 frames. ], tot_loss[loss=0.157, simple_loss=0.2328, pruned_loss=0.04057, over 1033539.29 frames. ], batch size: 39, lr: 9.26e-03, grad_scale: 8.0 +2023-03-21 00:26:19,466 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 00:26:23,132 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=48273.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 00:26:28,102 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4332, 2.6682, 2.1741, 2.4048, 2.4772, 2.1132, 2.6721, 2.3920], + device='cuda:0'), covar=tensor([0.1180, 0.0655, 0.1746, 0.1315, 0.1501, 0.1168, 0.1283, 0.0995], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0044, 0.0050, 0.0044, 0.0043, 0.0044, 0.0046, 0.0042], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:26:39,759 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 00:26:42,367 INFO [train.py:901] (0/2) Epoch 18, batch 300, loss[loss=0.1774, simple_loss=0.2537, pruned_loss=0.05054, over 7135.00 frames. ], tot_loss[loss=0.157, simple_loss=0.2332, pruned_loss=0.04045, over 1126464.02 frames. ], batch size: 98, lr: 9.26e-03, grad_scale: 8.0 +2023-03-21 00:26:42,466 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7020, 3.8249, 3.6792, 3.7942, 3.4946, 3.7511, 4.0611, 4.0822], + device='cuda:0'), covar=tensor([0.0211, 0.0168, 0.0217, 0.0163, 0.0380, 0.0371, 0.0239, 0.0204], + device='cuda:0'), in_proj_covar=tensor([0.0107, 0.0108, 0.0100, 0.0106, 0.0102, 0.0091, 0.0088, 0.0084], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:26:48,312 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.449e+02 2.135e+02 2.501e+02 2.825e+02 5.059e+02, threshold=5.003e+02, percent-clipped=0.0 +2023-03-21 00:26:48,839 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 00:26:49,954 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=48324.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:26:50,517 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4639, 3.2467, 3.3532, 3.1959, 2.7368, 2.8712, 3.5153, 2.6696], + device='cuda:0'), covar=tensor([0.0310, 0.0266, 0.0322, 0.0321, 0.0399, 0.0538, 0.0396, 0.1042], + device='cuda:0'), in_proj_covar=tensor([0.0316, 0.0320, 0.0257, 0.0343, 0.0309, 0.0307, 0.0328, 0.0294], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 00:27:07,445 INFO [train.py:901] (0/2) Epoch 18, batch 350, loss[loss=0.1504, simple_loss=0.2345, pruned_loss=0.03315, over 7221.00 frames. ], tot_loss[loss=0.1579, simple_loss=0.234, pruned_loss=0.04088, over 1197552.50 frames. ], batch size: 93, lr: 9.25e-03, grad_scale: 8.0 +2023-03-21 00:27:20,866 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=48385.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:27:23,746 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 00:27:26,442 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=48394.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:27:27,912 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=48397.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:27:34,233 INFO [train.py:901] (0/2) Epoch 18, batch 400, loss[loss=0.125, simple_loss=0.2136, pruned_loss=0.01825, over 7283.00 frames. ], tot_loss[loss=0.157, simple_loss=0.2333, pruned_loss=0.04037, over 1251558.13 frames. ], batch size: 70, lr: 9.25e-03, grad_scale: 8.0 +2023-03-21 00:27:40,285 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.256e+02 1.977e+02 2.369e+02 2.933e+02 5.499e+02, threshold=4.738e+02, percent-clipped=2.0 +2023-03-21 00:27:41,471 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=48423.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:27:43,861 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=48428.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:27:45,939 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2090, 0.8347, 1.5550, 1.6695, 1.3168, 1.7846, 1.4187, 1.6860], + device='cuda:0'), covar=tensor([0.0958, 0.3906, 0.2109, 0.1030, 0.1196, 0.2219, 0.1096, 0.1178], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0060, 0.0041, 0.0040, 0.0042, 0.0044, 0.0062, 0.0043], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 00:27:50,839 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=48442.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:27:51,939 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=48444.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:27:59,004 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=48458.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:27:59,328 INFO [train.py:901] (0/2) Epoch 18, batch 450, loss[loss=0.1673, simple_loss=0.2379, pruned_loss=0.04835, over 7362.00 frames. ], tot_loss[loss=0.1564, simple_loss=0.2326, pruned_loss=0.04014, over 1292454.25 frames. ], batch size: 63, lr: 9.24e-03, grad_scale: 8.0 +2023-03-21 00:28:02,508 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=48465.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 00:28:03,421 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 00:28:04,424 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 00:28:13,359 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=48484.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:28:14,341 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2806, 3.6800, 3.9360, 3.9575, 3.7281, 3.8381, 4.1520, 3.5476], + device='cuda:0'), covar=tensor([0.0134, 0.0192, 0.0124, 0.0141, 0.0408, 0.0129, 0.0163, 0.0203], + device='cuda:0'), in_proj_covar=tensor([0.0070, 0.0076, 0.0075, 0.0068, 0.0134, 0.0088, 0.0082, 0.0084], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:28:17,230 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=48492.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:28:25,772 INFO [train.py:901] (0/2) Epoch 18, batch 500, loss[loss=0.1721, simple_loss=0.2526, pruned_loss=0.04578, over 6832.00 frames. ], tot_loss[loss=0.1554, simple_loss=0.2314, pruned_loss=0.03973, over 1324654.49 frames. ], batch size: 107, lr: 9.24e-03, grad_scale: 8.0 +2023-03-21 00:28:27,845 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=48513.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 00:28:31,848 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.337e+02 2.257e+02 2.553e+02 3.004e+02 5.113e+02, threshold=5.107e+02, percent-clipped=1.0 +2023-03-21 00:28:35,602 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4728, 2.4980, 2.2084, 3.7167, 1.5653, 3.3474, 1.5122, 3.1227], + device='cuda:0'), covar=tensor([0.0152, 0.1028, 0.1711, 0.0095, 0.3805, 0.0146, 0.1072, 0.0297], + device='cuda:0'), in_proj_covar=tensor([0.0172, 0.0265, 0.0292, 0.0172, 0.0276, 0.0185, 0.0262, 0.0222], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 00:28:37,981 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 00:28:39,494 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 00:28:40,012 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 00:28:41,984 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 00:28:46,486 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 00:28:50,925 INFO [train.py:901] (0/2) Epoch 18, batch 550, loss[loss=0.1691, simple_loss=0.2469, pruned_loss=0.04566, over 7282.00 frames. ], tot_loss[loss=0.1555, simple_loss=0.2315, pruned_loss=0.0398, over 1350387.70 frames. ], batch size: 77, lr: 9.23e-03, grad_scale: 8.0 +2023-03-21 00:28:56,917 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=48568.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 00:28:59,792 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 00:29:00,445 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9497, 3.0173, 2.4641, 3.9384, 1.4937, 3.7983, 1.4097, 3.1857], + device='cuda:0'), covar=tensor([0.0121, 0.0833, 0.1515, 0.0104, 0.4684, 0.0131, 0.1365, 0.0454], + device='cuda:0'), in_proj_covar=tensor([0.0171, 0.0264, 0.0291, 0.0172, 0.0276, 0.0185, 0.0262, 0.0222], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 00:29:07,907 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 00:29:11,402 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 00:29:17,778 INFO [train.py:901] (0/2) Epoch 18, batch 600, loss[loss=0.1738, simple_loss=0.2533, pruned_loss=0.04712, over 7279.00 frames. ], tot_loss[loss=0.1558, simple_loss=0.2318, pruned_loss=0.0399, over 1367960.64 frames. ], batch size: 70, lr: 9.23e-03, grad_scale: 8.0 +2023-03-21 00:29:18,321 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 00:29:23,766 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.651e+02 2.190e+02 2.516e+02 3.083e+02 5.159e+02, threshold=5.033e+02, percent-clipped=1.0 +2023-03-21 00:29:34,341 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 00:29:43,692 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 00:29:44,192 INFO [train.py:901] (0/2) Epoch 18, batch 650, loss[loss=0.1806, simple_loss=0.2592, pruned_loss=0.05095, over 7239.00 frames. ], tot_loss[loss=0.1554, simple_loss=0.2315, pruned_loss=0.03968, over 1384049.13 frames. ], batch size: 89, lr: 9.22e-03, grad_scale: 8.0 +2023-03-21 00:29:53,076 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.91 vs. limit=5.0 +2023-03-21 00:29:54,771 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=48680.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:29:56,247 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.3661, 5.0267, 4.9038, 5.4216, 5.3399, 5.4246, 4.6636, 5.0646], + device='cuda:0'), covar=tensor([0.0673, 0.2147, 0.1774, 0.0711, 0.0612, 0.1015, 0.0672, 0.0793], + device='cuda:0'), in_proj_covar=tensor([0.0206, 0.0321, 0.0254, 0.0251, 0.0184, 0.0312, 0.0185, 0.0227], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:29:59,687 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 00:30:07,689 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 00:30:09,159 INFO [train.py:901] (0/2) Epoch 18, batch 700, loss[loss=0.1435, simple_loss=0.2181, pruned_loss=0.03446, over 7227.00 frames. ], tot_loss[loss=0.1561, simple_loss=0.2323, pruned_loss=0.03996, over 1398334.94 frames. ], batch size: 50, lr: 9.22e-03, grad_scale: 8.0 +2023-03-21 00:30:15,212 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.344e+02 2.059e+02 2.384e+02 3.008e+02 6.092e+02, threshold=4.769e+02, percent-clipped=1.0 +2023-03-21 00:30:18,862 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=48728.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:30:32,334 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=48753.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:30:32,784 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 00:30:33,309 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 00:30:35,265 INFO [train.py:901] (0/2) Epoch 18, batch 750, loss[loss=0.1523, simple_loss=0.2343, pruned_loss=0.03518, over 7214.00 frames. ], tot_loss[loss=0.156, simple_loss=0.2324, pruned_loss=0.03978, over 1408071.18 frames. ], batch size: 93, lr: 9.21e-03, grad_scale: 8.0 +2023-03-21 00:30:43,990 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=48776.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:30:45,507 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=48779.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:30:47,436 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 00:30:51,488 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 00:30:57,820 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 00:30:59,326 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 00:31:00,842 INFO [train.py:901] (0/2) Epoch 18, batch 800, loss[loss=0.1455, simple_loss=0.2298, pruned_loss=0.03058, over 7292.00 frames. ], tot_loss[loss=0.1563, simple_loss=0.2328, pruned_loss=0.03992, over 1415402.46 frames. ], batch size: 68, lr: 9.21e-03, grad_scale: 8.0 +2023-03-21 00:31:07,235 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.247e+02 2.109e+02 2.468e+02 2.979e+02 5.290e+02, threshold=4.936e+02, percent-clipped=1.0 +2023-03-21 00:31:07,391 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=48821.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:31:10,840 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 00:31:16,448 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=48838.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:31:26,867 INFO [train.py:901] (0/2) Epoch 18, batch 850, loss[loss=0.1368, simple_loss=0.1888, pruned_loss=0.04239, over 5837.00 frames. ], tot_loss[loss=0.156, simple_loss=0.2323, pruned_loss=0.03984, over 1418658.60 frames. ], batch size: 25, lr: 9.20e-03, grad_scale: 8.0 +2023-03-21 00:31:29,918 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 00:31:30,401 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 00:31:31,438 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=48868.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 00:31:34,750 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 00:31:37,828 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 00:31:38,428 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=48882.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:31:46,853 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=48899.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:31:50,630 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.73 vs. limit=2.0 +2023-03-21 00:31:52,975 INFO [train.py:901] (0/2) Epoch 18, batch 900, loss[loss=0.1457, simple_loss=0.2229, pruned_loss=0.03427, over 7263.00 frames. ], tot_loss[loss=0.1567, simple_loss=0.2332, pruned_loss=0.04016, over 1425123.01 frames. ], batch size: 52, lr: 9.20e-03, grad_scale: 8.0 +2023-03-21 00:31:56,431 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=48916.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 00:31:58,859 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.424e+02 2.181e+02 2.751e+02 3.267e+02 5.780e+02, threshold=5.502e+02, percent-clipped=4.0 +2023-03-21 00:32:00,535 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9366, 3.0089, 2.9398, 2.8923, 2.3584, 2.4094, 3.0245, 2.3466], + device='cuda:0'), covar=tensor([0.0439, 0.0347, 0.0348, 0.0376, 0.0385, 0.0622, 0.0475, 0.1159], + device='cuda:0'), in_proj_covar=tensor([0.0318, 0.0320, 0.0255, 0.0339, 0.0303, 0.0305, 0.0326, 0.0294], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 00:32:05,588 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6382, 2.7923, 2.3680, 2.8191, 2.7120, 2.2745, 2.9133, 2.4426], + device='cuda:0'), covar=tensor([0.0593, 0.0434, 0.0782, 0.0725, 0.0837, 0.0609, 0.0547, 0.1040], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0045, 0.0051, 0.0044, 0.0044, 0.0045, 0.0046, 0.0042], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:32:15,639 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 00:32:18,143 INFO [train.py:901] (0/2) Epoch 18, batch 950, loss[loss=0.1516, simple_loss=0.2314, pruned_loss=0.03585, over 7294.00 frames. ], tot_loss[loss=0.1565, simple_loss=0.2329, pruned_loss=0.04001, over 1429180.56 frames. ], batch size: 86, lr: 9.19e-03, grad_scale: 8.0 +2023-03-21 00:32:28,866 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=48980.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:32:39,159 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 00:32:42,633 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8016, 3.0078, 2.4595, 3.0594, 2.9441, 2.3393, 3.1866, 2.6533], + device='cuda:0'), covar=tensor([0.0908, 0.1201, 0.1411, 0.0836, 0.1236, 0.0776, 0.0971, 0.1723], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0045, 0.0051, 0.0045, 0.0044, 0.0046, 0.0046, 0.0043], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:32:44,977 INFO [train.py:901] (0/2) Epoch 18, batch 1000, loss[loss=0.1527, simple_loss=0.2272, pruned_loss=0.0391, over 7319.00 frames. ], tot_loss[loss=0.1569, simple_loss=0.2332, pruned_loss=0.04032, over 1432986.74 frames. ], batch size: 59, lr: 9.19e-03, grad_scale: 8.0 +2023-03-21 00:32:51,046 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.458e+02 2.043e+02 2.421e+02 2.913e+02 5.514e+02, threshold=4.842e+02, percent-clipped=1.0 +2023-03-21 00:32:54,614 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=49028.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:33:00,087 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 00:33:07,142 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=49053.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:33:08,455 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.95 vs. limit=2.0 +2023-03-21 00:33:10,115 INFO [train.py:901] (0/2) Epoch 18, batch 1050, loss[loss=0.1599, simple_loss=0.2254, pruned_loss=0.0472, over 7268.00 frames. ], tot_loss[loss=0.1566, simple_loss=0.2327, pruned_loss=0.04025, over 1432647.72 frames. ], batch size: 52, lr: 9.18e-03, grad_scale: 8.0 +2023-03-21 00:33:11,354 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1069, 1.5627, 1.5128, 1.5026, 1.6225, 1.4148, 1.5537, 1.1543], + device='cuda:0'), covar=tensor([0.0125, 0.0141, 0.0136, 0.0061, 0.0108, 0.0086, 0.0134, 0.0136], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0024, 0.0022, 0.0023, 0.0023, 0.0023, 0.0025, 0.0031], + device='cuda:0'), out_proj_covar=tensor([2.9475e-05, 2.7456e-05, 2.6254e-05, 2.5856e-05, 2.7910e-05, 2.5714e-05, + 2.8379e-05, 3.6095e-05], device='cuda:0') +2023-03-21 00:33:16,333 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8937, 4.4550, 4.4433, 4.9476, 4.9274, 4.8983, 4.3467, 4.4877], + device='cuda:0'), covar=tensor([0.0867, 0.2494, 0.1980, 0.1123, 0.0654, 0.1260, 0.0723, 0.1045], + device='cuda:0'), in_proj_covar=tensor([0.0207, 0.0324, 0.0257, 0.0256, 0.0185, 0.0319, 0.0184, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:33:21,128 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=49079.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:33:21,732 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0512, 2.8896, 2.0539, 3.3991, 2.4706, 2.9183, 1.5066, 1.9594], + device='cuda:0'), covar=tensor([0.0403, 0.0954, 0.2072, 0.0663, 0.0534, 0.0566, 0.2917, 0.1641], + device='cuda:0'), in_proj_covar=tensor([0.0257, 0.0240, 0.0298, 0.0245, 0.0254, 0.0253, 0.0260, 0.0276], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 00:33:23,131 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 00:33:27,832 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 00:33:32,931 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=49101.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:33:35,473 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3070, 2.5984, 2.1384, 2.6259, 2.6372, 2.2029, 2.7715, 2.3533], + device='cuda:0'), covar=tensor([0.1069, 0.0555, 0.1334, 0.0722, 0.0828, 0.0784, 0.0655, 0.1114], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0044, 0.0051, 0.0044, 0.0044, 0.0045, 0.0046, 0.0042], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:33:36,868 INFO [train.py:901] (0/2) Epoch 18, batch 1100, loss[loss=0.1394, simple_loss=0.2191, pruned_loss=0.02985, over 7242.00 frames. ], tot_loss[loss=0.1558, simple_loss=0.2321, pruned_loss=0.03968, over 1436066.61 frames. ], batch size: 89, lr: 9.18e-03, grad_scale: 8.0 +2023-03-21 00:33:42,801 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.227e+02 2.061e+02 2.375e+02 2.807e+02 6.437e+02, threshold=4.750e+02, percent-clipped=1.0 +2023-03-21 00:33:45,887 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=49127.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:33:56,411 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 00:33:56,894 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 00:34:01,900 INFO [train.py:901] (0/2) Epoch 18, batch 1150, loss[loss=0.1504, simple_loss=0.2268, pruned_loss=0.03703, over 7301.00 frames. ], tot_loss[loss=0.1547, simple_loss=0.231, pruned_loss=0.03922, over 1439939.69 frames. ], batch size: 80, lr: 9.18e-03, grad_scale: 8.0 +2023-03-21 00:34:08,699 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 00:34:09,705 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 00:34:12,326 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=49177.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:34:20,953 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=49194.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:34:28,732 INFO [train.py:901] (0/2) Epoch 18, batch 1200, loss[loss=0.1789, simple_loss=0.262, pruned_loss=0.0479, over 6636.00 frames. ], tot_loss[loss=0.1559, simple_loss=0.2323, pruned_loss=0.03973, over 1438635.58 frames. ], batch size: 106, lr: 9.17e-03, grad_scale: 8.0 +2023-03-21 00:34:34,052 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-21 00:34:34,778 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.271e+02 2.292e+02 2.732e+02 3.185e+02 8.511e+02, threshold=5.463e+02, percent-clipped=4.0 +2023-03-21 00:34:42,367 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 00:34:42,499 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=49236.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:34:54,720 INFO [train.py:901] (0/2) Epoch 18, batch 1250, loss[loss=0.165, simple_loss=0.2431, pruned_loss=0.04348, over 7356.00 frames. ], tot_loss[loss=0.156, simple_loss=0.2323, pruned_loss=0.03987, over 1437693.84 frames. ], batch size: 73, lr: 9.17e-03, grad_scale: 16.0 +2023-03-21 00:35:05,263 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.91 vs. limit=2.0 +2023-03-21 00:35:06,989 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 00:35:10,188 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4368, 3.4072, 2.3849, 4.0082, 2.9638, 3.4593, 1.8873, 2.2948], + device='cuda:0'), covar=tensor([0.0350, 0.0732, 0.2083, 0.0439, 0.0344, 0.0611, 0.2908, 0.1620], + device='cuda:0'), in_proj_covar=tensor([0.0259, 0.0241, 0.0301, 0.0248, 0.0258, 0.0255, 0.0264, 0.0279], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 00:35:10,595 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9417, 4.4483, 4.4317, 4.9258, 4.8237, 4.8586, 4.3125, 4.4841], + device='cuda:0'), covar=tensor([0.0698, 0.2259, 0.1802, 0.0956, 0.0788, 0.1242, 0.0654, 0.0932], + device='cuda:0'), in_proj_covar=tensor([0.0207, 0.0329, 0.0255, 0.0257, 0.0188, 0.0321, 0.0183, 0.0230], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:35:11,012 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 00:35:11,999 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 00:35:14,634 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=49297.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:35:20,485 INFO [train.py:901] (0/2) Epoch 18, batch 1300, loss[loss=0.1559, simple_loss=0.2365, pruned_loss=0.0376, over 7377.00 frames. ], tot_loss[loss=0.1572, simple_loss=0.2333, pruned_loss=0.04051, over 1436931.63 frames. ], batch size: 65, lr: 9.16e-03, grad_scale: 16.0 +2023-03-21 00:35:26,587 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.386e+02 2.121e+02 2.615e+02 3.045e+02 6.610e+02, threshold=5.231e+02, percent-clipped=1.0 +2023-03-21 00:35:35,186 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 00:35:38,313 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 00:35:42,448 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 00:35:47,080 INFO [train.py:901] (0/2) Epoch 18, batch 1350, loss[loss=0.1561, simple_loss=0.2341, pruned_loss=0.03902, over 7284.00 frames. ], tot_loss[loss=0.1564, simple_loss=0.2325, pruned_loss=0.04012, over 1439803.42 frames. ], batch size: 77, lr: 9.16e-03, grad_scale: 16.0 +2023-03-21 00:35:52,568 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 00:35:56,124 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1878, 4.7303, 4.7934, 4.7345, 4.6418, 4.2488, 4.8024, 4.6121], + device='cuda:0'), covar=tensor([0.0447, 0.0360, 0.0366, 0.0409, 0.0331, 0.0326, 0.0300, 0.0514], + device='cuda:0'), in_proj_covar=tensor([0.0181, 0.0209, 0.0149, 0.0154, 0.0125, 0.0192, 0.0162, 0.0125], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:36:12,417 INFO [train.py:901] (0/2) Epoch 18, batch 1400, loss[loss=0.1766, simple_loss=0.2502, pruned_loss=0.05152, over 7210.00 frames. ], tot_loss[loss=0.1563, simple_loss=0.2325, pruned_loss=0.04006, over 1440475.82 frames. ], batch size: 93, lr: 9.15e-03, grad_scale: 16.0 +2023-03-21 00:36:15,265 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.72 vs. limit=5.0 +2023-03-21 00:36:18,327 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.396e+02 2.167e+02 2.498e+02 3.013e+02 7.794e+02, threshold=4.996e+02, percent-clipped=2.0 +2023-03-21 00:36:18,943 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=49422.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:36:24,383 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.87 vs. limit=5.0 +2023-03-21 00:36:24,464 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 00:36:24,589 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=49432.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 00:36:38,492 INFO [train.py:901] (0/2) Epoch 18, batch 1450, loss[loss=0.1651, simple_loss=0.2367, pruned_loss=0.04673, over 7287.00 frames. ], tot_loss[loss=0.1566, simple_loss=0.2328, pruned_loss=0.04022, over 1439692.92 frames. ], batch size: 57, lr: 9.15e-03, grad_scale: 16.0 +2023-03-21 00:36:47,834 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=49477.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:36:49,245 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 00:36:50,912 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=49483.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:36:52,400 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5724, 4.1563, 4.1038, 4.6875, 4.5458, 4.5824, 4.0082, 4.1766], + device='cuda:0'), covar=tensor([0.0789, 0.2515, 0.2253, 0.0924, 0.0801, 0.1238, 0.0764, 0.1187], + device='cuda:0'), in_proj_covar=tensor([0.0203, 0.0320, 0.0252, 0.0248, 0.0181, 0.0312, 0.0178, 0.0224], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:36:55,994 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=49493.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 00:36:56,412 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=49494.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:37:03,814 INFO [train.py:901] (0/2) Epoch 18, batch 1500, loss[loss=0.1301, simple_loss=0.204, pruned_loss=0.02807, over 7179.00 frames. ], tot_loss[loss=0.1559, simple_loss=0.2316, pruned_loss=0.04012, over 1436173.21 frames. ], batch size: 39, lr: 9.14e-03, grad_scale: 16.0 +2023-03-21 00:37:05,397 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 00:37:10,443 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.480e+02 2.142e+02 2.435e+02 2.862e+02 6.987e+02, threshold=4.869e+02, percent-clipped=2.0 +2023-03-21 00:37:12,581 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=49525.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:37:20,925 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=49542.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:37:21,520 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0343, 2.4119, 2.0205, 3.0388, 3.2180, 2.9487, 2.7454, 2.4738], + device='cuda:0'), covar=tensor([0.1794, 0.0807, 0.3308, 0.0547, 0.0121, 0.0093, 0.0236, 0.0383], + device='cuda:0'), in_proj_covar=tensor([0.0279, 0.0236, 0.0273, 0.0264, 0.0155, 0.0150, 0.0184, 0.0200], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:37:29,288 INFO [train.py:901] (0/2) Epoch 18, batch 1550, loss[loss=0.1625, simple_loss=0.2344, pruned_loss=0.04532, over 7268.00 frames. ], tot_loss[loss=0.1564, simple_loss=0.2319, pruned_loss=0.04038, over 1438024.16 frames. ], batch size: 52, lr: 9.14e-03, grad_scale: 16.0 +2023-03-21 00:37:29,826 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 00:37:32,798 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.2340, 4.8383, 4.6932, 5.2011, 5.1320, 5.2307, 4.5028, 4.8637], + device='cuda:0'), covar=tensor([0.0611, 0.2270, 0.1903, 0.0950, 0.0716, 0.1058, 0.0617, 0.0843], + device='cuda:0'), in_proj_covar=tensor([0.0202, 0.0320, 0.0251, 0.0252, 0.0181, 0.0311, 0.0177, 0.0224], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:37:45,881 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=49592.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:37:51,857 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3038, 4.2000, 3.6170, 3.5555, 3.6796, 2.3788, 1.8505, 4.2840], + device='cuda:0'), covar=tensor([0.0029, 0.0042, 0.0074, 0.0051, 0.0069, 0.0386, 0.0469, 0.0037], + device='cuda:0'), in_proj_covar=tensor([0.0085, 0.0073, 0.0091, 0.0078, 0.0101, 0.0116, 0.0117, 0.0085], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 00:37:55,996 INFO [train.py:901] (0/2) Epoch 18, batch 1600, loss[loss=0.1753, simple_loss=0.2541, pruned_loss=0.04825, over 7224.00 frames. ], tot_loss[loss=0.1564, simple_loss=0.2321, pruned_loss=0.04032, over 1439596.02 frames. ], batch size: 93, lr: 9.13e-03, grad_scale: 16.0 +2023-03-21 00:38:02,089 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.528e+02 2.138e+02 2.464e+02 3.002e+02 4.958e+02, threshold=4.928e+02, percent-clipped=1.0 +2023-03-21 00:38:02,615 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 00:38:03,628 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 00:38:06,623 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 00:38:15,500 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 00:38:19,462 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 00:38:20,990 INFO [train.py:901] (0/2) Epoch 18, batch 1650, loss[loss=0.1666, simple_loss=0.2463, pruned_loss=0.0434, over 7298.00 frames. ], tot_loss[loss=0.156, simple_loss=0.2321, pruned_loss=0.03999, over 1441066.13 frames. ], batch size: 86, lr: 9.13e-03, grad_scale: 16.0 +2023-03-21 00:38:28,032 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 00:38:45,913 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 00:38:47,427 INFO [train.py:901] (0/2) Epoch 18, batch 1700, loss[loss=0.1541, simple_loss=0.2315, pruned_loss=0.03829, over 7320.00 frames. ], tot_loss[loss=0.1562, simple_loss=0.2325, pruned_loss=0.03999, over 1442019.85 frames. ], batch size: 59, lr: 9.13e-03, grad_scale: 16.0 +2023-03-21 00:38:49,929 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 00:38:53,485 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.456e+02 2.121e+02 2.517e+02 3.238e+02 5.019e+02, threshold=5.033e+02, percent-clipped=1.0 +2023-03-21 00:38:55,620 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3692, 4.9211, 5.0092, 4.9111, 4.7667, 4.4511, 5.0468, 4.7680], + device='cuda:0'), covar=tensor([0.0466, 0.0461, 0.0450, 0.0566, 0.0385, 0.0357, 0.0372, 0.0614], + device='cuda:0'), in_proj_covar=tensor([0.0185, 0.0214, 0.0152, 0.0160, 0.0127, 0.0197, 0.0166, 0.0127], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:39:00,517 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 00:39:10,754 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8088, 2.1129, 1.6687, 2.5487, 2.4965, 2.4538, 2.5759, 1.9824], + device='cuda:0'), covar=tensor([0.1931, 0.0824, 0.3528, 0.0571, 0.0127, 0.0107, 0.0231, 0.0217], + device='cuda:0'), in_proj_covar=tensor([0.0280, 0.0238, 0.0274, 0.0264, 0.0155, 0.0150, 0.0185, 0.0200], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:39:12,620 INFO [train.py:901] (0/2) Epoch 18, batch 1750, loss[loss=0.1599, simple_loss=0.2376, pruned_loss=0.04108, over 7366.00 frames. ], tot_loss[loss=0.1565, simple_loss=0.2329, pruned_loss=0.04003, over 1443249.86 frames. ], batch size: 65, lr: 9.12e-03, grad_scale: 16.0 +2023-03-21 00:39:16,337 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=49766.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:39:18,788 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.4567, 4.9290, 4.7754, 5.3968, 5.2412, 5.4264, 4.6042, 5.1127], + device='cuda:0'), covar=tensor([0.0676, 0.2464, 0.2206, 0.0981, 0.0913, 0.1048, 0.0628, 0.0836], + device='cuda:0'), in_proj_covar=tensor([0.0203, 0.0319, 0.0251, 0.0250, 0.0185, 0.0309, 0.0179, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:39:22,285 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=49778.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:39:25,974 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 00:39:26,482 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 00:39:28,536 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=49788.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 00:39:35,977 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6794, 3.2330, 2.3074, 4.0415, 2.5576, 3.0098, 1.5942, 2.1931], + device='cuda:0'), covar=tensor([0.0396, 0.0652, 0.2045, 0.0397, 0.0397, 0.0634, 0.2758, 0.1681], + device='cuda:0'), in_proj_covar=tensor([0.0257, 0.0241, 0.0294, 0.0244, 0.0256, 0.0252, 0.0258, 0.0277], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 00:39:39,292 INFO [train.py:901] (0/2) Epoch 18, batch 1800, loss[loss=0.1326, simple_loss=0.2183, pruned_loss=0.02342, over 7286.00 frames. ], tot_loss[loss=0.1567, simple_loss=0.233, pruned_loss=0.04022, over 1441804.05 frames. ], batch size: 68, lr: 9.12e-03, grad_scale: 16.0 +2023-03-21 00:39:45,279 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.430e+02 2.060e+02 2.380e+02 2.875e+02 7.454e+02, threshold=4.761e+02, percent-clipped=2.0 +2023-03-21 00:39:48,310 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 00:39:48,452 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=49827.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:40:01,434 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 00:40:04,434 INFO [train.py:901] (0/2) Epoch 18, batch 1850, loss[loss=0.1537, simple_loss=0.2345, pruned_loss=0.03647, over 7307.00 frames. ], tot_loss[loss=0.157, simple_loss=0.233, pruned_loss=0.04048, over 1442533.56 frames. ], batch size: 80, lr: 9.11e-03, grad_scale: 16.0 +2023-03-21 00:40:12,154 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 00:40:15,285 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4800, 5.0107, 5.0833, 5.0742, 4.8722, 4.5434, 5.1109, 4.8565], + device='cuda:0'), covar=tensor([0.0426, 0.0361, 0.0334, 0.0363, 0.0309, 0.0307, 0.0295, 0.0517], + device='cuda:0'), in_proj_covar=tensor([0.0181, 0.0208, 0.0149, 0.0155, 0.0124, 0.0192, 0.0161, 0.0125], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:40:22,405 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=49892.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:40:22,934 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9736, 2.5113, 1.9107, 2.9340, 2.3862, 2.9698, 2.6856, 2.5312], + device='cuda:0'), covar=tensor([0.1945, 0.0717, 0.3362, 0.0601, 0.0134, 0.0085, 0.0238, 0.0274], + device='cuda:0'), in_proj_covar=tensor([0.0276, 0.0234, 0.0270, 0.0263, 0.0152, 0.0148, 0.0182, 0.0199], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:40:28,788 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 00:40:30,770 INFO [train.py:901] (0/2) Epoch 18, batch 1900, loss[loss=0.1484, simple_loss=0.2352, pruned_loss=0.0308, over 7328.00 frames. ], tot_loss[loss=0.1568, simple_loss=0.2331, pruned_loss=0.04029, over 1443604.10 frames. ], batch size: 75, lr: 9.11e-03, grad_scale: 16.0 +2023-03-21 00:40:36,807 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.351e+02 2.153e+02 2.525e+02 3.111e+02 5.745e+02, threshold=5.051e+02, percent-clipped=1.0 +2023-03-21 00:40:39,980 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5372, 3.7211, 3.4393, 3.7176, 3.3921, 3.5310, 3.8902, 3.9312], + device='cuda:0'), covar=tensor([0.0233, 0.0150, 0.0226, 0.0166, 0.0329, 0.0590, 0.0300, 0.0187], + device='cuda:0'), in_proj_covar=tensor([0.0111, 0.0111, 0.0104, 0.0110, 0.0104, 0.0095, 0.0091, 0.0086], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:40:46,609 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=49940.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:40:49,673 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7191, 3.4330, 3.5819, 3.5207, 2.9679, 3.3762, 3.6707, 3.1805], + device='cuda:0'), covar=tensor([0.0332, 0.0255, 0.0198, 0.0241, 0.0876, 0.0216, 0.0279, 0.0312], + device='cuda:0'), in_proj_covar=tensor([0.0072, 0.0079, 0.0077, 0.0068, 0.0137, 0.0089, 0.0082, 0.0086], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:40:54,290 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 00:40:57,228 INFO [train.py:901] (0/2) Epoch 18, batch 1950, loss[loss=0.1695, simple_loss=0.2358, pruned_loss=0.05161, over 7315.00 frames. ], tot_loss[loss=0.1561, simple_loss=0.2322, pruned_loss=0.03994, over 1443090.28 frames. ], batch size: 49, lr: 9.10e-03, grad_scale: 16.0 +2023-03-21 00:41:05,318 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 00:41:10,263 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 00:41:10,759 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 00:41:22,514 INFO [train.py:901] (0/2) Epoch 18, batch 2000, loss[loss=0.145, simple_loss=0.2295, pruned_loss=0.03026, over 7232.00 frames. ], tot_loss[loss=0.1552, simple_loss=0.2315, pruned_loss=0.03941, over 1445252.46 frames. ], batch size: 55, lr: 9.10e-03, grad_scale: 16.0 +2023-03-21 00:41:27,162 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 00:41:27,944 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.41 vs. limit=2.0 +2023-03-21 00:41:28,634 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.390e+02 2.229e+02 2.681e+02 3.198e+02 6.835e+02, threshold=5.363e+02, percent-clipped=3.0 +2023-03-21 00:41:32,289 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7565, 3.0117, 2.5845, 2.9201, 2.8908, 2.4297, 3.0097, 2.5599], + device='cuda:0'), covar=tensor([0.1012, 0.0556, 0.1142, 0.1056, 0.0873, 0.0678, 0.0886, 0.1384], + device='cuda:0'), in_proj_covar=tensor([0.0046, 0.0044, 0.0053, 0.0046, 0.0045, 0.0046, 0.0047, 0.0044], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:41:35,732 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2559, 3.4234, 2.5337, 4.0072, 1.6839, 3.4457, 1.6956, 3.1470], + device='cuda:0'), covar=tensor([0.0060, 0.0609, 0.1471, 0.0097, 0.4250, 0.0142, 0.1228, 0.0221], + device='cuda:0'), in_proj_covar=tensor([0.0170, 0.0259, 0.0283, 0.0171, 0.0270, 0.0183, 0.0257, 0.0215], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 00:41:38,825 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5093, 1.9915, 2.2848, 1.8310, 2.0983, 2.0262, 1.8469, 1.5692], + device='cuda:0'), covar=tensor([0.0499, 0.0303, 0.0103, 0.0151, 0.0235, 0.0212, 0.0200, 0.0243], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0025, 0.0022, 0.0022, 0.0024, 0.0022, 0.0026, 0.0026], + device='cuda:0'), out_proj_covar=tensor([6.4199e-05, 6.2677e-05, 5.6204e-05, 5.5532e-05, 6.0676e-05, 5.6975e-05, + 6.3194e-05, 6.5375e-05], device='cuda:0') +2023-03-21 00:41:39,220 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 00:41:43,267 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.89 vs. limit=5.0 +2023-03-21 00:41:47,442 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 00:41:48,937 INFO [train.py:901] (0/2) Epoch 18, batch 2050, loss[loss=0.1599, simple_loss=0.2434, pruned_loss=0.0382, over 7238.00 frames. ], tot_loss[loss=0.1557, simple_loss=0.232, pruned_loss=0.03973, over 1445711.31 frames. ], batch size: 93, lr: 9.09e-03, grad_scale: 16.0 +2023-03-21 00:41:58,462 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=50078.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:42:03,472 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=50088.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 00:42:13,958 INFO [train.py:901] (0/2) Epoch 18, batch 2100, loss[loss=0.1751, simple_loss=0.2446, pruned_loss=0.05285, over 7278.00 frames. ], tot_loss[loss=0.1569, simple_loss=0.2331, pruned_loss=0.04033, over 1446755.86 frames. ], batch size: 57, lr: 9.09e-03, grad_scale: 16.0 +2023-03-21 00:42:20,552 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.448e+02 1.971e+02 2.438e+02 2.876e+02 7.861e+02, threshold=4.875e+02, percent-clipped=1.0 +2023-03-21 00:42:20,591 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 00:42:21,137 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=50122.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:42:23,096 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 00:42:23,134 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=50126.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:42:24,286 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=50128.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:42:28,864 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=50136.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 00:42:33,230 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.7170, 5.3273, 5.2515, 5.7181, 5.6025, 5.6785, 5.1544, 5.3601], + device='cuda:0'), covar=tensor([0.0530, 0.1655, 0.1630, 0.0660, 0.0685, 0.0910, 0.0468, 0.0677], + device='cuda:0'), in_proj_covar=tensor([0.0208, 0.0321, 0.0252, 0.0254, 0.0186, 0.0316, 0.0181, 0.0227], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:42:35,828 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2779, 1.5153, 1.4453, 1.4704, 1.6322, 1.3741, 1.4375, 1.0751], + device='cuda:0'), covar=tensor([0.0189, 0.0187, 0.0186, 0.0232, 0.0091, 0.0092, 0.0164, 0.0188], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0024, 0.0023, 0.0024, 0.0023, 0.0023, 0.0025, 0.0031], + device='cuda:0'), out_proj_covar=tensor([2.9700e-05, 2.7647e-05, 2.6946e-05, 2.6623e-05, 2.7604e-05, 2.6203e-05, + 2.8890e-05, 3.6078e-05], device='cuda:0') +2023-03-21 00:42:40,314 INFO [train.py:901] (0/2) Epoch 18, batch 2150, loss[loss=0.1614, simple_loss=0.2367, pruned_loss=0.043, over 7297.00 frames. ], tot_loss[loss=0.1566, simple_loss=0.2328, pruned_loss=0.04021, over 1446181.69 frames. ], batch size: 86, lr: 9.08e-03, grad_scale: 16.0 +2023-03-21 00:42:41,403 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5420, 2.4070, 2.2725, 3.7215, 1.5258, 3.3526, 1.4356, 3.2593], + device='cuda:0'), covar=tensor([0.0119, 0.0940, 0.1644, 0.0105, 0.3826, 0.0158, 0.1078, 0.0250], + device='cuda:0'), in_proj_covar=tensor([0.0171, 0.0259, 0.0283, 0.0172, 0.0270, 0.0185, 0.0255, 0.0215], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 00:42:55,382 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=50189.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:42:59,281 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=50197.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 00:43:01,633 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.0031, 4.7158, 4.5033, 5.0903, 4.9650, 5.0138, 4.4510, 4.6272], + device='cuda:0'), covar=tensor([0.0775, 0.2279, 0.2361, 0.1047, 0.0898, 0.1157, 0.0720, 0.0980], + device='cuda:0'), in_proj_covar=tensor([0.0208, 0.0322, 0.0255, 0.0253, 0.0187, 0.0314, 0.0182, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:43:06,079 INFO [train.py:901] (0/2) Epoch 18, batch 2200, loss[loss=0.1658, simple_loss=0.238, pruned_loss=0.04677, over 7222.00 frames. ], tot_loss[loss=0.1564, simple_loss=0.2324, pruned_loss=0.04021, over 1445027.64 frames. ], batch size: 45, lr: 9.08e-03, grad_scale: 16.0 +2023-03-21 00:43:08,536 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 00:43:12,643 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.466e+02 2.073e+02 2.473e+02 2.866e+02 5.110e+02, threshold=4.946e+02, percent-clipped=1.0 +2023-03-21 00:43:14,774 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0693, 4.0832, 3.3499, 3.4280, 3.2757, 2.3653, 1.5941, 4.0209], + device='cuda:0'), covar=tensor([0.0033, 0.0041, 0.0101, 0.0061, 0.0095, 0.0385, 0.0557, 0.0041], + device='cuda:0'), in_proj_covar=tensor([0.0086, 0.0073, 0.0093, 0.0080, 0.0101, 0.0117, 0.0117, 0.0086], + device='cuda:0'), out_proj_covar=tensor([1.0771e-04, 9.9554e-05, 1.2063e-04, 1.0667e-04, 1.2708e-04, 1.4979e-04, + 1.5179e-04, 1.0761e-04], device='cuda:0') +2023-03-21 00:43:31,389 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=50258.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 00:43:31,733 INFO [train.py:901] (0/2) Epoch 18, batch 2250, loss[loss=0.1412, simple_loss=0.2206, pruned_loss=0.0309, over 7149.00 frames. ], tot_loss[loss=0.1564, simple_loss=0.2322, pruned_loss=0.04028, over 1442777.46 frames. ], batch size: 41, lr: 9.08e-03, grad_scale: 16.0 +2023-03-21 00:43:43,968 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 00:43:43,979 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 00:43:50,190 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8601, 3.9497, 3.1106, 3.2226, 3.1030, 2.2145, 1.6533, 3.8767], + device='cuda:0'), covar=tensor([0.0048, 0.0028, 0.0108, 0.0069, 0.0112, 0.0436, 0.0575, 0.0051], + device='cuda:0'), in_proj_covar=tensor([0.0086, 0.0073, 0.0093, 0.0079, 0.0101, 0.0117, 0.0117, 0.0086], + device='cuda:0'), out_proj_covar=tensor([1.0698e-04, 9.9592e-05, 1.1920e-04, 1.0661e-04, 1.2694e-04, 1.5006e-04, + 1.5182e-04, 1.0792e-04], device='cuda:0') +2023-03-21 00:43:58,195 INFO [train.py:901] (0/2) Epoch 18, batch 2300, loss[loss=0.1292, simple_loss=0.1857, pruned_loss=0.03636, over 6153.00 frames. ], tot_loss[loss=0.1562, simple_loss=0.2324, pruned_loss=0.03996, over 1442160.87 frames. ], batch size: 26, lr: 9.07e-03, grad_scale: 16.0 +2023-03-21 00:43:58,203 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 00:44:04,184 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.450e+02 2.045e+02 2.583e+02 3.047e+02 4.891e+02, threshold=5.165e+02, percent-clipped=0.0 +2023-03-21 00:44:23,274 INFO [train.py:901] (0/2) Epoch 18, batch 2350, loss[loss=0.1872, simple_loss=0.2626, pruned_loss=0.05585, over 6776.00 frames. ], tot_loss[loss=0.1564, simple_loss=0.2329, pruned_loss=0.04, over 1443390.88 frames. ], batch size: 106, lr: 9.07e-03, grad_scale: 8.0 +2023-03-21 00:44:33,706 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.95 vs. limit=2.0 +2023-03-21 00:44:45,991 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 00:44:50,019 INFO [train.py:901] (0/2) Epoch 18, batch 2400, loss[loss=0.1537, simple_loss=0.228, pruned_loss=0.03971, over 7318.00 frames. ], tot_loss[loss=0.1567, simple_loss=0.2331, pruned_loss=0.04017, over 1442594.67 frames. ], batch size: 59, lr: 9.06e-03, grad_scale: 8.0 +2023-03-21 00:44:52,540 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 00:44:56,603 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.401e+02 1.959e+02 2.320e+02 2.795e+02 6.835e+02, threshold=4.641e+02, percent-clipped=1.0 +2023-03-21 00:44:56,720 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=50422.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:44:58,403 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.87 vs. limit=2.0 +2023-03-21 00:45:03,296 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 00:45:05,798 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 00:45:15,322 INFO [train.py:901] (0/2) Epoch 18, batch 2450, loss[loss=0.1775, simple_loss=0.2478, pruned_loss=0.05358, over 7345.00 frames. ], tot_loss[loss=0.155, simple_loss=0.2315, pruned_loss=0.03923, over 1440069.54 frames. ], batch size: 63, lr: 9.06e-03, grad_scale: 8.0 +2023-03-21 00:45:21,415 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=50470.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:45:28,424 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=50484.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:45:32,916 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 00:45:41,448 INFO [train.py:901] (0/2) Epoch 18, batch 2500, loss[loss=0.1527, simple_loss=0.2365, pruned_loss=0.03447, over 7312.00 frames. ], tot_loss[loss=0.1557, simple_loss=0.2323, pruned_loss=0.03951, over 1443648.13 frames. ], batch size: 75, lr: 9.05e-03, grad_scale: 8.0 +2023-03-21 00:45:47,924 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.466e+02 1.929e+02 2.329e+02 2.819e+02 6.164e+02, threshold=4.658e+02, percent-clipped=3.0 +2023-03-21 00:45:48,229 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.60 vs. limit=5.0 +2023-03-21 00:45:58,174 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 00:45:59,296 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7780, 3.4597, 3.3681, 3.2000, 3.0414, 2.8917, 3.6271, 2.8743], + device='cuda:0'), covar=tensor([0.0336, 0.0279, 0.0295, 0.0320, 0.0408, 0.0616, 0.0439, 0.0961], + device='cuda:0'), in_proj_covar=tensor([0.0321, 0.0324, 0.0258, 0.0342, 0.0306, 0.0304, 0.0329, 0.0291], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 00:46:00,256 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=50546.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:46:03,694 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=50553.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 00:46:07,303 INFO [train.py:901] (0/2) Epoch 18, batch 2550, loss[loss=0.142, simple_loss=0.2215, pruned_loss=0.0312, over 7367.00 frames. ], tot_loss[loss=0.1549, simple_loss=0.2315, pruned_loss=0.03909, over 1444027.85 frames. ], batch size: 73, lr: 9.05e-03, grad_scale: 8.0 +2023-03-21 00:46:12,923 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8975, 4.0501, 3.4983, 3.3121, 3.1977, 2.1523, 1.5895, 4.0176], + device='cuda:0'), covar=tensor([0.0071, 0.0072, 0.0111, 0.0079, 0.0146, 0.0620, 0.0675, 0.0066], + device='cuda:0'), in_proj_covar=tensor([0.0085, 0.0073, 0.0091, 0.0079, 0.0099, 0.0116, 0.0116, 0.0085], + device='cuda:0'), out_proj_covar=tensor([1.0481e-04, 9.8856e-05, 1.1728e-04, 1.0610e-04, 1.2517e-04, 1.4801e-04, + 1.4950e-04, 1.0614e-04], device='cuda:0') +2023-03-21 00:46:32,566 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=50607.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:46:33,441 INFO [train.py:901] (0/2) Epoch 18, batch 2600, loss[loss=0.1565, simple_loss=0.2392, pruned_loss=0.03689, over 7343.00 frames. ], tot_loss[loss=0.1553, simple_loss=0.2318, pruned_loss=0.03941, over 1444188.73 frames. ], batch size: 73, lr: 9.04e-03, grad_scale: 8.0 +2023-03-21 00:46:39,872 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.366e+02 2.102e+02 2.428e+02 3.097e+02 8.694e+02, threshold=4.856e+02, percent-clipped=5.0 +2023-03-21 00:46:40,633 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.90 vs. limit=5.0 +2023-03-21 00:46:47,196 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.89 vs. limit=5.0 +2023-03-21 00:46:58,162 INFO [train.py:901] (0/2) Epoch 18, batch 2650, loss[loss=0.1392, simple_loss=0.2168, pruned_loss=0.03078, over 7159.00 frames. ], tot_loss[loss=0.155, simple_loss=0.2315, pruned_loss=0.03921, over 1445415.30 frames. ], batch size: 41, lr: 9.04e-03, grad_scale: 8.0 +2023-03-21 00:47:22,547 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8187, 2.6718, 1.9187, 3.1374, 2.1576, 2.6726, 1.2801, 1.7884], + device='cuda:0'), covar=tensor([0.0417, 0.1024, 0.2323, 0.0673, 0.0480, 0.0577, 0.2981, 0.1601], + device='cuda:0'), in_proj_covar=tensor([0.0261, 0.0241, 0.0294, 0.0249, 0.0261, 0.0255, 0.0260, 0.0276], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 00:47:22,872 INFO [train.py:901] (0/2) Epoch 18, batch 2700, loss[loss=0.1863, simple_loss=0.2583, pruned_loss=0.05712, over 6682.00 frames. ], tot_loss[loss=0.1552, simple_loss=0.2316, pruned_loss=0.0394, over 1442249.45 frames. ], batch size: 106, lr: 9.04e-03, grad_scale: 8.0 +2023-03-21 00:47:29,656 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.454e+02 2.009e+02 2.432e+02 3.028e+02 8.989e+02, threshold=4.865e+02, percent-clipped=2.0 +2023-03-21 00:47:32,686 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9328, 2.6510, 3.0970, 2.8789, 3.1891, 2.7410, 2.6189, 3.1590], + device='cuda:0'), covar=tensor([0.1545, 0.0650, 0.1037, 0.1955, 0.0650, 0.1915, 0.1922, 0.1170], + device='cuda:0'), in_proj_covar=tensor([0.0045, 0.0052, 0.0040, 0.0041, 0.0040, 0.0038, 0.0055, 0.0042], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 00:47:47,628 INFO [train.py:901] (0/2) Epoch 18, batch 2750, loss[loss=0.1698, simple_loss=0.2472, pruned_loss=0.04622, over 7216.00 frames. ], tot_loss[loss=0.1557, simple_loss=0.232, pruned_loss=0.03967, over 1444775.63 frames. ], batch size: 50, lr: 9.03e-03, grad_scale: 8.0 +2023-03-21 00:47:59,193 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=50782.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:48:00,559 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=50784.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:48:12,412 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=50808.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 00:48:12,740 INFO [train.py:901] (0/2) Epoch 18, batch 2800, loss[loss=0.1943, simple_loss=0.2623, pruned_loss=0.06311, over 6723.00 frames. ], tot_loss[loss=0.1562, simple_loss=0.2327, pruned_loss=0.03982, over 1445701.97 frames. ], batch size: 107, lr: 9.03e-03, grad_scale: 8.0 +2023-03-21 00:48:13,311 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3030, 3.4597, 3.1659, 3.4678, 3.3571, 3.2435, 3.7188, 3.7177], + device='cuda:0'), covar=tensor([0.0254, 0.0193, 0.0263, 0.0194, 0.0261, 0.0773, 0.0224, 0.0182], + device='cuda:0'), in_proj_covar=tensor([0.0108, 0.0107, 0.0102, 0.0105, 0.0102, 0.0090, 0.0087, 0.0081], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:48:18,919 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.308e+02 1.954e+02 2.260e+02 2.876e+02 5.616e+02, threshold=4.520e+02, percent-clipped=4.0 +2023-03-21 00:48:21,746 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-03-21 00:48:25,156 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-18.pt +2023-03-21 00:48:42,627 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 00:48:42,691 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 00:48:42,716 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 00:48:42,737 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 00:48:42,932 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 00:48:42,949 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 00:48:42,962 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 00:48:43,067 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 00:48:43,148 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 00:48:43,164 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 00:48:43,181 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 00:48:43,695 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 00:48:43,701 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 00:48:43,709 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 00:48:43,848 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 00:48:43,850 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 00:48:43,909 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 00:48:45,779 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=50832.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:48:46,226 INFO [train.py:901] (0/2) Epoch 19, batch 0, loss[loss=0.164, simple_loss=0.2475, pruned_loss=0.0402, over 7305.00 frames. ], tot_loss[loss=0.164, simple_loss=0.2475, pruned_loss=0.0402, over 7305.00 frames. ], batch size: 83, lr: 8.80e-03, grad_scale: 8.0 +2023-03-21 00:48:46,227 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 00:49:04,914 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8971, 3.5049, 3.6482, 3.6261, 3.6979, 3.3716, 3.7328, 3.1914], + device='cuda:0'), covar=tensor([0.0128, 0.0193, 0.0141, 0.0175, 0.0387, 0.0157, 0.0169, 0.0285], + device='cuda:0'), in_proj_covar=tensor([0.0071, 0.0078, 0.0078, 0.0070, 0.0139, 0.0091, 0.0081, 0.0086], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:49:09,162 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4229, 1.2099, 1.6606, 1.6497, 1.5781, 1.9034, 1.3460, 1.6696], + device='cuda:0'), covar=tensor([0.3055, 0.3869, 0.1147, 0.1746, 0.1094, 0.1680, 0.2007, 0.2649], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0057, 0.0041, 0.0040, 0.0043, 0.0043, 0.0064, 0.0043], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 00:49:12,200 INFO [train.py:935] (0/2) Epoch 19, validation: loss=0.166, simple_loss=0.2533, pruned_loss=0.03935, over 1622729.00 frames. +2023-03-21 00:49:12,201 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 00:49:17,308 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=50843.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:49:18,712 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 00:49:22,183 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=50853.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 00:49:28,700 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7663, 2.2121, 1.7304, 2.5755, 2.0964, 2.7369, 2.2259, 2.2651], + device='cuda:0'), covar=tensor([0.1814, 0.0957, 0.3157, 0.0535, 0.0100, 0.0113, 0.0260, 0.0182], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0236, 0.0271, 0.0266, 0.0157, 0.0153, 0.0187, 0.0200], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:49:29,629 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.2231, 5.6677, 5.6533, 5.6110, 5.4258, 5.3742, 5.7871, 5.5937], + device='cuda:0'), covar=tensor([0.0346, 0.0313, 0.0347, 0.0388, 0.0257, 0.0226, 0.0265, 0.0365], + device='cuda:0'), in_proj_covar=tensor([0.0183, 0.0213, 0.0153, 0.0155, 0.0128, 0.0195, 0.0165, 0.0127], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:49:30,074 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 00:49:30,203 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=50869.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 00:49:37,072 INFO [train.py:901] (0/2) Epoch 19, batch 50, loss[loss=0.1568, simple_loss=0.2368, pruned_loss=0.03839, over 7298.00 frames. ], tot_loss[loss=0.155, simple_loss=0.2322, pruned_loss=0.03893, over 327388.28 frames. ], batch size: 68, lr: 8.79e-03, grad_scale: 8.0 +2023-03-21 00:49:37,089 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 00:49:39,147 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 00:49:42,587 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 00:49:46,148 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=50901.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 00:49:46,641 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=50902.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:49:57,770 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.168e+02 2.080e+02 2.542e+02 3.231e+02 6.129e+02, threshold=5.083e+02, percent-clipped=6.0 +2023-03-21 00:50:01,304 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 00:50:02,307 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 00:50:03,348 INFO [train.py:901] (0/2) Epoch 19, batch 100, loss[loss=0.1265, simple_loss=0.2101, pruned_loss=0.02145, over 7135.00 frames. ], tot_loss[loss=0.1532, simple_loss=0.2294, pruned_loss=0.03848, over 574265.89 frames. ], batch size: 41, lr: 8.79e-03, grad_scale: 8.0 +2023-03-21 00:50:10,045 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=50946.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:50:21,007 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4294, 3.2682, 2.9592, 3.1281, 2.4856, 2.5110, 3.2369, 2.4361], + device='cuda:0'), covar=tensor([0.0356, 0.0374, 0.0323, 0.0381, 0.0592, 0.0721, 0.0504, 0.1325], + device='cuda:0'), in_proj_covar=tensor([0.0320, 0.0324, 0.0260, 0.0342, 0.0308, 0.0303, 0.0329, 0.0295], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 00:50:28,289 INFO [train.py:901] (0/2) Epoch 19, batch 150, loss[loss=0.146, simple_loss=0.2106, pruned_loss=0.04065, over 6944.00 frames. ], tot_loss[loss=0.153, simple_loss=0.2293, pruned_loss=0.03835, over 767861.04 frames. ], batch size: 35, lr: 8.78e-03, grad_scale: 8.0 +2023-03-21 00:50:42,230 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=51007.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:50:49,621 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.504e+02 2.085e+02 2.413e+02 2.711e+02 7.121e+02, threshold=4.826e+02, percent-clipped=1.0 +2023-03-21 00:50:55,265 INFO [train.py:901] (0/2) Epoch 19, batch 200, loss[loss=0.1534, simple_loss=0.2318, pruned_loss=0.03754, over 7289.00 frames. ], tot_loss[loss=0.1537, simple_loss=0.23, pruned_loss=0.03872, over 918136.71 frames. ], batch size: 77, lr: 8.78e-03, grad_scale: 8.0 +2023-03-21 00:50:58,387 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7308, 5.2232, 5.2716, 5.2324, 5.0089, 4.8005, 5.3209, 5.0930], + device='cuda:0'), covar=tensor([0.0380, 0.0370, 0.0363, 0.0381, 0.0321, 0.0263, 0.0306, 0.0430], + device='cuda:0'), in_proj_covar=tensor([0.0183, 0.0213, 0.0155, 0.0156, 0.0129, 0.0196, 0.0166, 0.0128], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:51:02,989 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 00:51:07,519 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 00:51:10,159 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=51062.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:51:12,470 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 00:51:14,809 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.42 vs. limit=2.0 +2023-03-21 00:51:21,235 INFO [train.py:901] (0/2) Epoch 19, batch 250, loss[loss=0.1387, simple_loss=0.2158, pruned_loss=0.03081, over 7275.00 frames. ], tot_loss[loss=0.1538, simple_loss=0.2301, pruned_loss=0.03881, over 1035309.59 frames. ], batch size: 66, lr: 8.78e-03, grad_scale: 8.0 +2023-03-21 00:51:27,235 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 00:51:41,223 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.398e+02 2.027e+02 2.472e+02 3.001e+02 8.354e+02, threshold=4.945e+02, percent-clipped=6.0 +2023-03-21 00:51:41,886 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=51123.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:51:46,661 INFO [train.py:901] (0/2) Epoch 19, batch 300, loss[loss=0.1516, simple_loss=0.2305, pruned_loss=0.03633, over 7330.00 frames. ], tot_loss[loss=0.1542, simple_loss=0.2303, pruned_loss=0.03905, over 1124308.02 frames. ], batch size: 42, lr: 8.77e-03, grad_scale: 8.0 +2023-03-21 00:51:48,680 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 00:51:49,187 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=51138.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:51:57,105 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 00:52:02,101 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=51164.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 00:52:05,752 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2832, 1.3424, 1.2331, 1.2729, 1.2256, 1.2497, 1.3489, 1.1326], + device='cuda:0'), covar=tensor([0.0107, 0.0082, 0.0131, 0.0088, 0.0086, 0.0077, 0.0133, 0.0094], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0023, 0.0023, 0.0023, 0.0023, 0.0023, 0.0025, 0.0030], + device='cuda:0'), out_proj_covar=tensor([2.9482e-05, 2.6851e-05, 2.6788e-05, 2.6095e-05, 2.7321e-05, 2.5748e-05, + 2.8051e-05, 3.5230e-05], device='cuda:0') +2023-03-21 00:52:06,764 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=51172.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 00:52:12,031 INFO [train.py:901] (0/2) Epoch 19, batch 350, loss[loss=0.1382, simple_loss=0.2187, pruned_loss=0.02885, over 7308.00 frames. ], tot_loss[loss=0.1536, simple_loss=0.2297, pruned_loss=0.03871, over 1194269.30 frames. ], batch size: 80, lr: 8.77e-03, grad_scale: 8.0 +2023-03-21 00:52:22,688 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=51202.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:52:26,060 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8894, 4.4265, 4.3045, 4.8422, 4.7126, 4.8076, 4.2211, 4.4539], + device='cuda:0'), covar=tensor([0.0643, 0.2275, 0.1882, 0.0847, 0.0736, 0.1041, 0.0649, 0.0902], + device='cuda:0'), in_proj_covar=tensor([0.0212, 0.0326, 0.0259, 0.0255, 0.0190, 0.0318, 0.0183, 0.0233], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:52:27,365 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-03-21 00:52:31,003 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 00:52:32,475 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.486e+02 2.221e+02 2.732e+02 3.384e+02 5.150e+02, threshold=5.463e+02, percent-clipped=2.0 +2023-03-21 00:52:38,077 INFO [train.py:901] (0/2) Epoch 19, batch 400, loss[loss=0.1716, simple_loss=0.2364, pruned_loss=0.05344, over 7250.00 frames. ], tot_loss[loss=0.1539, simple_loss=0.2305, pruned_loss=0.0387, over 1251114.79 frames. ], batch size: 47, lr: 8.76e-03, grad_scale: 8.0 +2023-03-21 00:52:38,247 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=51233.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 00:52:46,613 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=51250.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:52:46,704 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8368, 3.1827, 2.5890, 2.9133, 2.9679, 2.5643, 3.0926, 2.8528], + device='cuda:0'), covar=tensor([0.0910, 0.0936, 0.1152, 0.2222, 0.1498, 0.0723, 0.0832, 0.1596], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0046, 0.0054, 0.0047, 0.0045, 0.0047, 0.0048, 0.0044], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:52:47,196 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5332, 2.0946, 2.2218, 1.9986, 1.8619, 1.8754, 1.9357, 1.5645], + device='cuda:0'), covar=tensor([0.0440, 0.0397, 0.0120, 0.0117, 0.0362, 0.0295, 0.0164, 0.0209], + device='cuda:0'), in_proj_covar=tensor([0.0026, 0.0026, 0.0024, 0.0022, 0.0024, 0.0023, 0.0027, 0.0027], + device='cuda:0'), out_proj_covar=tensor([6.6621e-05, 6.6199e-05, 5.9280e-05, 5.6462e-05, 6.2680e-05, 6.0039e-05, + 6.5185e-05, 6.7696e-05], device='cuda:0') +2023-03-21 00:52:56,177 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.65 vs. limit=5.0 +2023-03-21 00:53:02,439 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.98 vs. limit=5.0 +2023-03-21 00:53:04,714 INFO [train.py:901] (0/2) Epoch 19, batch 450, loss[loss=0.1533, simple_loss=0.2378, pruned_loss=0.03438, over 7310.00 frames. ], tot_loss[loss=0.1536, simple_loss=0.2304, pruned_loss=0.03835, over 1293267.73 frames. ], batch size: 83, lr: 8.76e-03, grad_scale: 8.0 +2023-03-21 00:53:13,549 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 00:53:14,042 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 00:53:14,618 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=51302.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:53:24,705 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.185e+02 2.176e+02 2.553e+02 3.010e+02 4.956e+02, threshold=5.106e+02, percent-clipped=0.0 +2023-03-21 00:53:30,232 INFO [train.py:901] (0/2) Epoch 19, batch 500, loss[loss=0.1394, simple_loss=0.22, pruned_loss=0.02938, over 7293.00 frames. ], tot_loss[loss=0.1537, simple_loss=0.2301, pruned_loss=0.03867, over 1322378.92 frames. ], batch size: 77, lr: 8.75e-03, grad_scale: 8.0 +2023-03-21 00:53:46,514 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 00:53:48,009 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 00:53:48,519 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 00:53:50,522 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 00:53:55,359 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 00:53:55,999 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=51382.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:53:56,382 INFO [train.py:901] (0/2) Epoch 19, batch 550, loss[loss=0.1598, simple_loss=0.2456, pruned_loss=0.03698, over 6762.00 frames. ], tot_loss[loss=0.1539, simple_loss=0.2305, pruned_loss=0.03865, over 1350241.50 frames. ], batch size: 107, lr: 8.75e-03, grad_scale: 8.0 +2023-03-21 00:54:01,102 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=51392.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:54:06,374 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 00:54:13,787 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 00:54:14,348 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=51418.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:54:16,259 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.429e+02 2.200e+02 2.639e+02 3.099e+02 6.064e+02, threshold=5.278e+02, percent-clipped=2.0 +2023-03-21 00:54:16,310 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 00:54:22,247 INFO [train.py:901] (0/2) Epoch 19, batch 600, loss[loss=0.1567, simple_loss=0.2354, pruned_loss=0.03899, over 7320.00 frames. ], tot_loss[loss=0.1548, simple_loss=0.2316, pruned_loss=0.03902, over 1370329.22 frames. ], batch size: 61, lr: 8.75e-03, grad_scale: 8.0 +2023-03-21 00:54:23,767 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 00:54:24,822 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=51438.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:54:25,331 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6124, 2.8982, 2.3692, 2.8372, 2.6834, 2.4200, 2.7637, 2.5876], + device='cuda:0'), covar=tensor([0.0679, 0.0726, 0.1718, 0.0871, 0.1199, 0.0570, 0.0849, 0.1167], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0046, 0.0053, 0.0046, 0.0044, 0.0047, 0.0048, 0.0043], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:54:27,973 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=51443.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:54:32,945 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=51453.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:54:38,409 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=51464.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 00:54:39,798 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 00:54:44,655 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.64 vs. limit=5.0 +2023-03-21 00:54:47,681 INFO [train.py:901] (0/2) Epoch 19, batch 650, loss[loss=0.1706, simple_loss=0.2428, pruned_loss=0.04921, over 7302.00 frames. ], tot_loss[loss=0.155, simple_loss=0.2317, pruned_loss=0.03915, over 1387553.52 frames. ], batch size: 49, lr: 8.74e-03, grad_scale: 8.0 +2023-03-21 00:54:48,719 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 00:54:49,253 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=51486.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:54:53,738 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=51495.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:54:58,802 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0700, 3.2056, 2.0653, 3.8368, 2.6586, 3.1093, 1.6411, 2.0023], + device='cuda:0'), covar=tensor([0.0339, 0.0801, 0.2536, 0.0654, 0.0447, 0.0587, 0.3314, 0.2063], + device='cuda:0'), in_proj_covar=tensor([0.0257, 0.0239, 0.0295, 0.0252, 0.0262, 0.0252, 0.0258, 0.0273], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 00:55:02,345 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=51512.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 00:55:05,155 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1158, 2.5553, 1.9579, 3.0233, 2.6894, 2.4485, 2.3694, 2.4288], + device='cuda:0'), covar=tensor([0.1691, 0.0813, 0.3092, 0.0534, 0.0133, 0.0073, 0.0194, 0.0227], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0232, 0.0267, 0.0264, 0.0155, 0.0150, 0.0186, 0.0199], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 00:55:07,543 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 00:55:07,993 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.287e+02 2.099e+02 2.381e+02 2.975e+02 4.917e+02, threshold=4.762e+02, percent-clipped=0.0 +2023-03-21 00:55:11,172 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=51528.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 00:55:14,436 INFO [train.py:901] (0/2) Epoch 19, batch 700, loss[loss=0.1558, simple_loss=0.2308, pruned_loss=0.04038, over 7268.00 frames. ], tot_loss[loss=0.1543, simple_loss=0.231, pruned_loss=0.03883, over 1400899.57 frames. ], batch size: 70, lr: 8.74e-03, grad_scale: 8.0 +2023-03-21 00:55:16,493 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 00:55:26,096 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=51556.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:55:39,515 INFO [train.py:901] (0/2) Epoch 19, batch 750, loss[loss=0.1541, simple_loss=0.232, pruned_loss=0.0381, over 7335.00 frames. ], tot_loss[loss=0.1536, simple_loss=0.2304, pruned_loss=0.03844, over 1411529.60 frames. ], batch size: 61, lr: 8.73e-03, grad_scale: 8.0 +2023-03-21 00:55:41,049 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 00:55:41,549 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 00:55:49,658 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=51602.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:55:56,775 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 00:56:00,949 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.391e+02 2.035e+02 2.484e+02 2.955e+02 5.287e+02, threshold=4.968e+02, percent-clipped=1.0 +2023-03-21 00:56:01,505 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 00:56:06,520 INFO [train.py:901] (0/2) Epoch 19, batch 800, loss[loss=0.1988, simple_loss=0.2636, pruned_loss=0.06701, over 7236.00 frames. ], tot_loss[loss=0.1543, simple_loss=0.2309, pruned_loss=0.03887, over 1416986.18 frames. ], batch size: 93, lr: 8.73e-03, grad_scale: 8.0 +2023-03-21 00:56:06,537 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 00:56:07,537 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 00:56:15,112 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=51650.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:56:17,220 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8815, 2.9771, 1.9752, 3.4887, 2.3939, 2.9539, 1.5547, 1.6856], + device='cuda:0'), covar=tensor([0.0421, 0.0781, 0.3166, 0.0594, 0.0472, 0.0552, 0.3656, 0.2704], + device='cuda:0'), in_proj_covar=tensor([0.0258, 0.0238, 0.0297, 0.0251, 0.0260, 0.0252, 0.0260, 0.0274], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 00:56:17,531 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 00:56:21,101 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6190, 1.8966, 2.0958, 1.8781, 2.0098, 1.9111, 2.0090, 1.4917], + device='cuda:0'), covar=tensor([0.0416, 0.0552, 0.0205, 0.0173, 0.0475, 0.0409, 0.0289, 0.0318], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0025, 0.0023, 0.0022, 0.0024, 0.0023, 0.0027, 0.0027], + device='cuda:0'), out_proj_covar=tensor([6.4643e-05, 6.4445e-05, 5.7919e-05, 5.4961e-05, 6.1475e-05, 5.9120e-05, + 6.5162e-05, 6.7431e-05], device='cuda:0') +2023-03-21 00:56:32,138 INFO [train.py:901] (0/2) Epoch 19, batch 850, loss[loss=0.1325, simple_loss=0.2094, pruned_loss=0.0278, over 7193.00 frames. ], tot_loss[loss=0.155, simple_loss=0.2315, pruned_loss=0.03924, over 1423915.22 frames. ], batch size: 39, lr: 8.73e-03, grad_scale: 8.0 +2023-03-21 00:56:36,191 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 00:56:36,195 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 00:56:41,909 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 00:56:44,965 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 00:56:50,286 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.62 vs. limit=2.0 +2023-03-21 00:56:50,629 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=51718.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:56:52,531 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.416e+02 2.319e+02 2.686e+02 3.166e+02 5.097e+02, threshold=5.372e+02, percent-clipped=1.0 +2023-03-21 00:56:57,946 INFO [train.py:901] (0/2) Epoch 19, batch 900, loss[loss=0.1662, simple_loss=0.2456, pruned_loss=0.04341, over 7131.00 frames. ], tot_loss[loss=0.1546, simple_loss=0.2313, pruned_loss=0.03892, over 1428365.92 frames. ], batch size: 98, lr: 8.72e-03, grad_scale: 8.0 +2023-03-21 00:56:59,700 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.87 vs. limit=2.0 +2023-03-21 00:57:00,476 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=51738.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:57:05,089 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=51747.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:57:05,516 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=51748.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:57:08,097 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3384, 1.1146, 1.5976, 1.8155, 1.5605, 2.0019, 1.5866, 1.6247], + device='cuda:0'), covar=tensor([0.1989, 0.3831, 0.1010, 0.2058, 0.5745, 0.2072, 0.1858, 0.3893], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0055, 0.0042, 0.0039, 0.0042, 0.0042, 0.0061, 0.0043], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 00:57:12,630 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8618, 2.2990, 2.5323, 2.2380, 2.3300, 2.2373, 2.2678, 1.6059], + device='cuda:0'), covar=tensor([0.0373, 0.0316, 0.0195, 0.0095, 0.0342, 0.0321, 0.0221, 0.0330], + device='cuda:0'), in_proj_covar=tensor([0.0026, 0.0025, 0.0023, 0.0022, 0.0024, 0.0023, 0.0027, 0.0027], + device='cuda:0'), out_proj_covar=tensor([6.5380e-05, 6.4835e-05, 5.7929e-05, 5.5416e-05, 6.1881e-05, 5.9166e-05, + 6.5468e-05, 6.7567e-05], device='cuda:0') +2023-03-21 00:57:14,573 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=51766.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:57:23,827 INFO [train.py:901] (0/2) Epoch 19, batch 950, loss[loss=0.1633, simple_loss=0.2366, pruned_loss=0.04495, over 7315.00 frames. ], tot_loss[loss=0.1539, simple_loss=0.2305, pruned_loss=0.03869, over 1432946.15 frames. ], batch size: 83, lr: 8.72e-03, grad_scale: 8.0 +2023-03-21 00:57:24,353 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 00:57:37,515 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=51808.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:57:44,393 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.232e+02 2.043e+02 2.485e+02 2.996e+02 5.664e+02, threshold=4.969e+02, percent-clipped=1.0 +2023-03-21 00:57:47,490 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=51828.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 00:57:48,376 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 00:57:49,838 INFO [train.py:901] (0/2) Epoch 19, batch 1000, loss[loss=0.1705, simple_loss=0.2289, pruned_loss=0.05609, over 7154.00 frames. ], tot_loss[loss=0.1541, simple_loss=0.2305, pruned_loss=0.0388, over 1433253.33 frames. ], batch size: 39, lr: 8.71e-03, grad_scale: 8.0 +2023-03-21 00:57:59,012 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=51851.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:58:10,177 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 00:58:12,180 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=51876.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 00:58:15,531 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-03-21 00:58:16,246 INFO [train.py:901] (0/2) Epoch 19, batch 1050, loss[loss=0.1765, simple_loss=0.2546, pruned_loss=0.04923, over 6679.00 frames. ], tot_loss[loss=0.1539, simple_loss=0.2303, pruned_loss=0.0388, over 1435384.63 frames. ], batch size: 107, lr: 8.71e-03, grad_scale: 8.0 +2023-03-21 00:58:19,656 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.81 vs. limit=2.0 +2023-03-21 00:58:31,792 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 00:58:35,846 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.265e+02 2.062e+02 2.448e+02 3.155e+02 8.388e+02, threshold=4.897e+02, percent-clipped=3.0 +2023-03-21 00:58:35,880 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 00:58:41,364 INFO [train.py:901] (0/2) Epoch 19, batch 1100, loss[loss=0.1429, simple_loss=0.2214, pruned_loss=0.03221, over 7275.00 frames. ], tot_loss[loss=0.1538, simple_loss=0.2304, pruned_loss=0.03862, over 1437064.85 frames. ], batch size: 52, lr: 8.70e-03, grad_scale: 8.0 +2023-03-21 00:58:45,675 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.89 vs. limit=2.0 +2023-03-21 00:59:03,312 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9430, 3.1647, 3.8004, 3.6202, 3.7592, 3.8853, 3.8605, 3.7938], + device='cuda:0'), covar=tensor([0.0019, 0.0091, 0.0027, 0.0034, 0.0031, 0.0022, 0.0033, 0.0035], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0052, 0.0046, 0.0044, 0.0043, 0.0047, 0.0044, 0.0056], + device='cuda:0'), out_proj_covar=tensor([7.6315e-05, 1.2536e-04, 1.0430e-04, 9.3940e-05, 9.3787e-05, 9.8455e-05, + 1.0317e-04, 1.2253e-04], device='cuda:0') +2023-03-21 00:59:05,206 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 00:59:05,684 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 00:59:07,704 INFO [train.py:901] (0/2) Epoch 19, batch 1150, loss[loss=0.1342, simple_loss=0.2136, pruned_loss=0.02733, over 7342.00 frames. ], tot_loss[loss=0.1537, simple_loss=0.23, pruned_loss=0.03866, over 1437957.77 frames. ], batch size: 44, lr: 8.70e-03, grad_scale: 8.0 +2023-03-21 00:59:16,511 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-52000.pt +2023-03-21 00:59:21,585 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 00:59:22,066 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 00:59:31,083 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.626e+02 2.216e+02 2.556e+02 3.107e+02 7.950e+02, threshold=5.111e+02, percent-clipped=3.0 +2023-03-21 00:59:37,237 INFO [train.py:901] (0/2) Epoch 19, batch 1200, loss[loss=0.1489, simple_loss=0.2286, pruned_loss=0.03461, over 7279.00 frames. ], tot_loss[loss=0.1537, simple_loss=0.2303, pruned_loss=0.03859, over 1440156.07 frames. ], batch size: 66, lr: 8.70e-03, grad_scale: 8.0 +2023-03-21 00:59:39,933 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=52038.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:59:40,888 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=52040.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:59:44,742 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=52048.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 00:59:54,788 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 01:00:02,804 INFO [train.py:901] (0/2) Epoch 19, batch 1250, loss[loss=0.1535, simple_loss=0.2315, pruned_loss=0.03772, over 7320.00 frames. ], tot_loss[loss=0.154, simple_loss=0.2308, pruned_loss=0.03856, over 1442890.18 frames. ], batch size: 80, lr: 8.69e-03, grad_scale: 8.0 +2023-03-21 01:00:04,381 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=52086.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:00:09,381 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=52096.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:00:12,001 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=52101.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:00:12,884 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=52103.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:00:18,394 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 01:00:22,907 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.500e+02 2.050e+02 2.393e+02 2.856e+02 5.508e+02, threshold=4.786e+02, percent-clipped=1.0 +2023-03-21 01:00:22,961 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 01:00:24,392 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 01:00:25,091 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.0467, 1.3550, 1.1542, 1.2691, 1.3107, 1.1869, 1.3346, 1.0664], + device='cuda:0'), covar=tensor([0.0120, 0.0091, 0.0192, 0.0077, 0.0083, 0.0078, 0.0074, 0.0100], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0024, 0.0024, 0.0024, 0.0023, 0.0023, 0.0025, 0.0031], + device='cuda:0'), out_proj_covar=tensor([2.9247e-05, 2.7535e-05, 2.7600e-05, 2.7176e-05, 2.6742e-05, 2.6048e-05, + 2.8282e-05, 3.6199e-05], device='cuda:0') +2023-03-21 01:00:28,410 INFO [train.py:901] (0/2) Epoch 19, batch 1300, loss[loss=0.1458, simple_loss=0.229, pruned_loss=0.03133, over 7243.00 frames. ], tot_loss[loss=0.1537, simple_loss=0.2306, pruned_loss=0.03845, over 1441193.50 frames. ], batch size: 89, lr: 8.69e-03, grad_scale: 8.0 +2023-03-21 01:00:35,328 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6419, 3.0146, 2.5280, 2.8590, 2.7332, 2.5044, 2.8009, 2.7643], + device='cuda:0'), covar=tensor([0.0959, 0.0793, 0.1012, 0.1289, 0.1430, 0.0706, 0.1173, 0.1108], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0045, 0.0053, 0.0047, 0.0045, 0.0047, 0.0048, 0.0043], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:00:38,328 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=52151.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:00:47,119 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 01:00:50,091 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 01:00:53,661 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 01:00:54,148 INFO [train.py:901] (0/2) Epoch 19, batch 1350, loss[loss=0.1507, simple_loss=0.23, pruned_loss=0.03568, over 7321.00 frames. ], tot_loss[loss=0.1539, simple_loss=0.2306, pruned_loss=0.03855, over 1441668.33 frames. ], batch size: 49, lr: 8.68e-03, grad_scale: 8.0 +2023-03-21 01:00:59,283 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=52193.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:01:02,299 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=52199.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:01:04,556 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 01:01:07,689 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4990, 4.0047, 3.9822, 4.4352, 4.4119, 4.4360, 3.8120, 3.9853], + device='cuda:0'), covar=tensor([0.0807, 0.2760, 0.2662, 0.1118, 0.0839, 0.1342, 0.0855, 0.1049], + device='cuda:0'), in_proj_covar=tensor([0.0213, 0.0330, 0.0261, 0.0261, 0.0190, 0.0319, 0.0186, 0.0234], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:01:14,638 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.452e+02 2.232e+02 2.578e+02 3.253e+02 9.285e+02, threshold=5.156e+02, percent-clipped=5.0 +2023-03-21 01:01:20,782 INFO [train.py:901] (0/2) Epoch 19, batch 1400, loss[loss=0.1237, simple_loss=0.1814, pruned_loss=0.03297, over 6635.00 frames. ], tot_loss[loss=0.1532, simple_loss=0.23, pruned_loss=0.03821, over 1441613.95 frames. ], batch size: 29, lr: 8.68e-03, grad_scale: 8.0 +2023-03-21 01:01:31,259 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=52254.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:01:34,816 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2616, 1.6257, 1.4851, 1.6317, 1.5933, 1.3729, 1.4906, 1.0985], + device='cuda:0'), covar=tensor([0.0086, 0.0091, 0.0154, 0.0064, 0.0057, 0.0080, 0.0083, 0.0141], + device='cuda:0'), in_proj_covar=tensor([0.0024, 0.0024, 0.0023, 0.0024, 0.0022, 0.0023, 0.0025, 0.0031], + device='cuda:0'), out_proj_covar=tensor([2.8626e-05, 2.7055e-05, 2.6878e-05, 2.6522e-05, 2.6439e-05, 2.5673e-05, + 2.7932e-05, 3.5624e-05], device='cuda:0') +2023-03-21 01:01:37,695 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 01:01:39,717 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8102, 3.1353, 3.7132, 3.6416, 3.7821, 3.7456, 3.7352, 3.5904], + device='cuda:0'), covar=tensor([0.0024, 0.0102, 0.0037, 0.0040, 0.0026, 0.0028, 0.0047, 0.0052], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0053, 0.0047, 0.0045, 0.0045, 0.0048, 0.0046, 0.0058], + device='cuda:0'), out_proj_covar=tensor([7.9300e-05, 1.2887e-04, 1.0786e-04, 9.6027e-05, 9.5338e-05, 1.0214e-04, + 1.0766e-04, 1.2636e-04], device='cuda:0') +2023-03-21 01:01:45,626 INFO [train.py:901] (0/2) Epoch 19, batch 1450, loss[loss=0.1672, simple_loss=0.2343, pruned_loss=0.05009, over 7215.00 frames. ], tot_loss[loss=0.1534, simple_loss=0.2301, pruned_loss=0.03834, over 1441936.70 frames. ], batch size: 45, lr: 8.68e-03, grad_scale: 8.0 +2023-03-21 01:02:00,984 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=52312.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:02:01,358 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 01:02:06,382 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.471e+02 2.084e+02 2.470e+02 2.878e+02 4.955e+02, threshold=4.940e+02, percent-clipped=0.0 +2023-03-21 01:02:12,006 INFO [train.py:901] (0/2) Epoch 19, batch 1500, loss[loss=0.1514, simple_loss=0.2344, pruned_loss=0.03421, over 7305.00 frames. ], tot_loss[loss=0.1542, simple_loss=0.2308, pruned_loss=0.03881, over 1442207.97 frames. ], batch size: 80, lr: 8.67e-03, grad_scale: 8.0 +2023-03-21 01:02:17,973 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 01:02:32,293 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=52373.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:02:37,704 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.78 vs. limit=2.0 +2023-03-21 01:02:37,823 INFO [train.py:901] (0/2) Epoch 19, batch 1550, loss[loss=0.1747, simple_loss=0.2501, pruned_loss=0.04967, over 7235.00 frames. ], tot_loss[loss=0.1543, simple_loss=0.2309, pruned_loss=0.03881, over 1440584.85 frames. ], batch size: 93, lr: 8.67e-03, grad_scale: 16.0 +2023-03-21 01:02:43,368 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 01:02:44,428 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=52396.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:02:48,252 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=52403.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:02:51,270 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9043, 2.1042, 1.7121, 2.6595, 2.3007, 2.4682, 2.3109, 2.3654], + device='cuda:0'), covar=tensor([0.1899, 0.0945, 0.3158, 0.0734, 0.0112, 0.0173, 0.0175, 0.0284], + device='cuda:0'), in_proj_covar=tensor([0.0276, 0.0232, 0.0269, 0.0262, 0.0155, 0.0154, 0.0185, 0.0200], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:02:56,421 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2096, 4.7169, 4.7399, 4.7152, 4.7033, 4.2584, 4.8516, 4.6509], + device='cuda:0'), covar=tensor([0.0487, 0.0404, 0.0443, 0.0493, 0.0320, 0.0377, 0.0296, 0.0479], + device='cuda:0'), in_proj_covar=tensor([0.0187, 0.0214, 0.0160, 0.0159, 0.0130, 0.0197, 0.0167, 0.0130], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:02:58,271 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.373e+02 1.980e+02 2.348e+02 2.883e+02 5.696e+02, threshold=4.696e+02, percent-clipped=3.0 +2023-03-21 01:03:03,901 INFO [train.py:901] (0/2) Epoch 19, batch 1600, loss[loss=0.1456, simple_loss=0.2247, pruned_loss=0.0333, over 7290.00 frames. ], tot_loss[loss=0.1537, simple_loss=0.2302, pruned_loss=0.03855, over 1440266.12 frames. ], batch size: 70, lr: 8.66e-03, grad_scale: 16.0 +2023-03-21 01:03:12,990 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=52451.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:03:13,934 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 01:03:14,964 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 01:03:17,961 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 01:03:27,031 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 01:03:29,522 INFO [train.py:901] (0/2) Epoch 19, batch 1650, loss[loss=0.1451, simple_loss=0.2253, pruned_loss=0.03243, over 7349.00 frames. ], tot_loss[loss=0.1536, simple_loss=0.2302, pruned_loss=0.03848, over 1439148.29 frames. ], batch size: 51, lr: 8.66e-03, grad_scale: 16.0 +2023-03-21 01:03:31,488 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 01:03:37,474 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.04 vs. limit=5.0 +2023-03-21 01:03:39,811 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 01:03:40,575 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.33 vs. limit=5.0 +2023-03-21 01:03:49,860 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.218e+02 2.238e+02 2.612e+02 3.033e+02 5.856e+02, threshold=5.224e+02, percent-clipped=2.0 +2023-03-21 01:03:55,459 INFO [train.py:901] (0/2) Epoch 19, batch 1700, loss[loss=0.1701, simple_loss=0.2458, pruned_loss=0.04716, over 7253.00 frames. ], tot_loss[loss=0.1545, simple_loss=0.2312, pruned_loss=0.03892, over 1440651.61 frames. ], batch size: 89, lr: 8.65e-03, grad_scale: 8.0 +2023-03-21 01:03:55,469 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 01:03:59,510 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 01:04:00,891 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-21 01:04:03,863 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7429, 2.8604, 3.4234, 3.5911, 3.6619, 3.6223, 3.5704, 3.4846], + device='cuda:0'), covar=tensor([0.0023, 0.0120, 0.0041, 0.0035, 0.0029, 0.0032, 0.0056, 0.0051], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0053, 0.0048, 0.0045, 0.0044, 0.0049, 0.0046, 0.0058], + device='cuda:0'), out_proj_covar=tensor([7.8542e-05, 1.2868e-04, 1.1016e-04, 9.6436e-05, 9.4285e-05, 1.0294e-04, + 1.0786e-04, 1.2581e-04], device='cuda:0') +2023-03-21 01:04:04,312 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=52549.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:04:07,403 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=52555.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 01:04:10,817 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 01:04:18,244 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-03-21 01:04:21,988 INFO [train.py:901] (0/2) Epoch 19, batch 1750, loss[loss=0.1112, simple_loss=0.1724, pruned_loss=0.02504, over 6042.00 frames. ], tot_loss[loss=0.1541, simple_loss=0.2309, pruned_loss=0.03864, over 1439945.97 frames. ], batch size: 26, lr: 8.65e-03, grad_scale: 8.0 +2023-03-21 01:04:25,120 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0583, 3.9084, 3.3900, 3.3645, 2.9878, 2.1161, 1.9000, 4.0697], + device='cuda:0'), covar=tensor([0.0028, 0.0051, 0.0099, 0.0058, 0.0118, 0.0444, 0.0465, 0.0031], + device='cuda:0'), in_proj_covar=tensor([0.0086, 0.0073, 0.0092, 0.0080, 0.0103, 0.0117, 0.0117, 0.0086], + device='cuda:0'), out_proj_covar=tensor([1.0436e-04, 9.8899e-05, 1.1718e-04, 1.0638e-04, 1.2919e-04, 1.4957e-04, + 1.5008e-04, 1.0658e-04], device='cuda:0') +2023-03-21 01:04:33,999 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 01:04:35,026 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 01:04:39,136 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=52616.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 01:04:42,424 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.465e+02 2.001e+02 2.278e+02 2.833e+02 4.492e+02, threshold=4.555e+02, percent-clipped=0.0 +2023-03-21 01:04:47,522 INFO [train.py:901] (0/2) Epoch 19, batch 1800, loss[loss=0.1796, simple_loss=0.2468, pruned_loss=0.05625, over 7251.00 frames. ], tot_loss[loss=0.1546, simple_loss=0.2313, pruned_loss=0.03891, over 1441822.82 frames. ], batch size: 55, lr: 8.65e-03, grad_scale: 8.0 +2023-03-21 01:04:57,718 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 01:05:06,607 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=52668.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:05:11,490 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 01:05:13,962 INFO [train.py:901] (0/2) Epoch 19, batch 1850, loss[loss=0.1669, simple_loss=0.251, pruned_loss=0.04143, over 7118.00 frames. ], tot_loss[loss=0.1547, simple_loss=0.2313, pruned_loss=0.03901, over 1442330.22 frames. ], batch size: 98, lr: 8.64e-03, grad_scale: 8.0 +2023-03-21 01:05:20,387 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 01:05:20,463 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=52696.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:05:27,467 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9144, 4.1834, 3.8673, 4.1090, 3.7419, 4.1068, 4.4535, 4.4447], + device='cuda:0'), covar=tensor([0.0196, 0.0129, 0.0178, 0.0139, 0.0292, 0.0190, 0.0181, 0.0159], + device='cuda:0'), in_proj_covar=tensor([0.0106, 0.0104, 0.0099, 0.0104, 0.0098, 0.0088, 0.0085, 0.0080], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:05:34,437 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.453e+02 2.173e+02 2.520e+02 2.887e+02 6.459e+02, threshold=5.039e+02, percent-clipped=3.0 +2023-03-21 01:05:37,991 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 01:05:39,494 INFO [train.py:901] (0/2) Epoch 19, batch 1900, loss[loss=0.1724, simple_loss=0.2502, pruned_loss=0.04736, over 7272.00 frames. ], tot_loss[loss=0.1542, simple_loss=0.2308, pruned_loss=0.03882, over 1440966.07 frames. ], batch size: 47, lr: 8.64e-03, grad_scale: 8.0 +2023-03-21 01:05:45,154 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=52744.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:05:45,185 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0198, 4.5310, 4.6171, 4.5348, 4.5876, 4.2203, 4.6817, 4.5432], + device='cuda:0'), covar=tensor([0.0529, 0.0512, 0.0491, 0.0566, 0.0328, 0.0359, 0.0399, 0.0550], + device='cuda:0'), in_proj_covar=tensor([0.0186, 0.0214, 0.0157, 0.0160, 0.0128, 0.0193, 0.0168, 0.0131], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:05:49,824 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=52752.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:06:02,372 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 01:06:05,397 INFO [train.py:901] (0/2) Epoch 19, batch 1950, loss[loss=0.131, simple_loss=0.2148, pruned_loss=0.0236, over 7331.00 frames. ], tot_loss[loss=0.1542, simple_loss=0.2308, pruned_loss=0.0388, over 1443590.53 frames. ], batch size: 44, lr: 8.63e-03, grad_scale: 8.0 +2023-03-21 01:06:07,042 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0609, 3.9724, 3.3762, 3.3909, 3.0357, 2.2815, 1.7591, 4.0141], + device='cuda:0'), covar=tensor([0.0028, 0.0034, 0.0090, 0.0055, 0.0110, 0.0401, 0.0497, 0.0039], + device='cuda:0'), in_proj_covar=tensor([0.0085, 0.0073, 0.0091, 0.0079, 0.0101, 0.0116, 0.0115, 0.0085], + device='cuda:0'), out_proj_covar=tensor([1.0401e-04, 9.8715e-05, 1.1639e-04, 1.0476e-04, 1.2675e-04, 1.4764e-04, + 1.4762e-04, 1.0588e-04], device='cuda:0') +2023-03-21 01:06:13,056 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 01:06:18,133 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 01:06:18,628 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 01:06:21,274 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=52813.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:06:23,362 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8009, 2.9246, 2.5324, 2.7790, 2.7550, 2.5155, 2.8933, 2.7017], + device='cuda:0'), covar=tensor([0.0902, 0.0992, 0.1147, 0.1253, 0.1228, 0.0693, 0.1061, 0.0885], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0046, 0.0053, 0.0047, 0.0046, 0.0047, 0.0047, 0.0043], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:06:26,195 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.329e+02 2.025e+02 2.271e+02 2.764e+02 6.029e+02, threshold=4.543e+02, percent-clipped=1.0 +2023-03-21 01:06:31,318 INFO [train.py:901] (0/2) Epoch 19, batch 2000, loss[loss=0.1591, simple_loss=0.2339, pruned_loss=0.04213, over 7288.00 frames. ], tot_loss[loss=0.1532, simple_loss=0.2301, pruned_loss=0.03813, over 1443986.92 frames. ], batch size: 77, lr: 8.63e-03, grad_scale: 8.0 +2023-03-21 01:06:35,579 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 01:06:40,200 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=52849.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:06:46,634 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 01:06:50,813 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0832, 2.5942, 2.0043, 3.2440, 2.6769, 2.6527, 2.1090, 2.3636], + device='cuda:0'), covar=tensor([0.1847, 0.0784, 0.3161, 0.0503, 0.0128, 0.0092, 0.0175, 0.0251], + device='cuda:0'), in_proj_covar=tensor([0.0280, 0.0235, 0.0272, 0.0268, 0.0158, 0.0155, 0.0187, 0.0202], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:06:51,321 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9564, 2.2227, 2.1632, 3.3272, 1.4102, 3.1317, 1.3355, 2.6999], + device='cuda:0'), covar=tensor([0.0104, 0.0978, 0.1454, 0.0121, 0.3935, 0.0138, 0.1010, 0.0247], + device='cuda:0'), in_proj_covar=tensor([0.0180, 0.0261, 0.0286, 0.0176, 0.0272, 0.0190, 0.0258, 0.0223], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:06:55,224 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 01:06:57,194 INFO [train.py:901] (0/2) Epoch 19, batch 2050, loss[loss=0.1607, simple_loss=0.2389, pruned_loss=0.04125, over 7248.00 frames. ], tot_loss[loss=0.1537, simple_loss=0.2307, pruned_loss=0.03837, over 1443788.54 frames. ], batch size: 52, lr: 8.63e-03, grad_scale: 8.0 +2023-03-21 01:06:58,370 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7000, 2.8694, 2.4798, 2.7633, 2.6753, 2.5812, 2.7844, 2.6553], + device='cuda:0'), covar=tensor([0.1237, 0.0716, 0.1138, 0.1115, 0.1260, 0.0641, 0.0777, 0.0890], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0046, 0.0054, 0.0047, 0.0046, 0.0047, 0.0048, 0.0044], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:07:04,805 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=52897.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:07:05,882 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2254, 3.7478, 3.8755, 3.8054, 3.7282, 3.7363, 3.9810, 3.5977], + device='cuda:0'), covar=tensor([0.0102, 0.0154, 0.0107, 0.0152, 0.0400, 0.0113, 0.0150, 0.0156], + device='cuda:0'), in_proj_covar=tensor([0.0074, 0.0082, 0.0081, 0.0071, 0.0144, 0.0090, 0.0085, 0.0090], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:07:11,879 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=52911.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 01:07:14,500 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9749, 2.9719, 1.9867, 3.6740, 2.4405, 2.9954, 1.5319, 1.8407], + device='cuda:0'), covar=tensor([0.0355, 0.1072, 0.2408, 0.0468, 0.0469, 0.0608, 0.3097, 0.2066], + device='cuda:0'), in_proj_covar=tensor([0.0260, 0.0243, 0.0299, 0.0253, 0.0265, 0.0255, 0.0258, 0.0278], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 01:07:18,353 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.311e+02 2.118e+02 2.490e+02 3.094e+02 4.672e+02, threshold=4.979e+02, percent-clipped=1.0 +2023-03-21 01:07:23,363 INFO [train.py:901] (0/2) Epoch 19, batch 2100, loss[loss=0.1489, simple_loss=0.2306, pruned_loss=0.03358, over 7240.00 frames. ], tot_loss[loss=0.1533, simple_loss=0.2303, pruned_loss=0.0381, over 1444723.06 frames. ], batch size: 89, lr: 8.62e-03, grad_scale: 8.0 +2023-03-21 01:07:28,564 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.35 vs. limit=5.0 +2023-03-21 01:07:28,819 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 01:07:31,841 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 01:07:37,004 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.01 vs. limit=2.0 +2023-03-21 01:07:40,779 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=52968.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:07:48,871 INFO [train.py:901] (0/2) Epoch 19, batch 2150, loss[loss=0.1629, simple_loss=0.2419, pruned_loss=0.042, over 7335.00 frames. ], tot_loss[loss=0.1532, simple_loss=0.2303, pruned_loss=0.0381, over 1444258.55 frames. ], batch size: 54, lr: 8.62e-03, grad_scale: 8.0 +2023-03-21 01:08:00,649 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.85 vs. limit=5.0 +2023-03-21 01:08:06,594 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=53016.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:08:09,989 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 9.314e+01 1.872e+02 2.271e+02 2.909e+02 5.856e+02, threshold=4.542e+02, percent-clipped=2.0 +2023-03-21 01:08:12,311 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.28 vs. limit=5.0 +2023-03-21 01:08:15,055 INFO [train.py:901] (0/2) Epoch 19, batch 2200, loss[loss=0.1305, simple_loss=0.2058, pruned_loss=0.02757, over 7165.00 frames. ], tot_loss[loss=0.1528, simple_loss=0.2299, pruned_loss=0.03786, over 1445151.07 frames. ], batch size: 39, lr: 8.61e-03, grad_scale: 8.0 +2023-03-21 01:08:18,091 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 01:08:18,182 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2182, 3.7641, 3.8601, 3.8266, 3.6856, 3.7338, 4.0765, 3.5305], + device='cuda:0'), covar=tensor([0.0129, 0.0139, 0.0120, 0.0149, 0.0472, 0.0104, 0.0159, 0.0173], + device='cuda:0'), in_proj_covar=tensor([0.0073, 0.0081, 0.0080, 0.0071, 0.0142, 0.0090, 0.0084, 0.0088], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:08:26,685 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=53056.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:08:26,908 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-21 01:08:31,967 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8639, 2.3603, 1.6585, 2.6352, 2.1449, 2.0868, 2.0722, 2.0326], + device='cuda:0'), covar=tensor([0.1730, 0.0770, 0.3521, 0.0415, 0.0104, 0.0099, 0.0230, 0.0255], + device='cuda:0'), in_proj_covar=tensor([0.0276, 0.0232, 0.0272, 0.0263, 0.0156, 0.0155, 0.0187, 0.0201], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:08:33,978 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5951, 2.8322, 1.8881, 3.3465, 2.1697, 2.7391, 1.3167, 1.8113], + device='cuda:0'), covar=tensor([0.0347, 0.0664, 0.2275, 0.0553, 0.0452, 0.0657, 0.3011, 0.1730], + device='cuda:0'), in_proj_covar=tensor([0.0262, 0.0242, 0.0297, 0.0254, 0.0266, 0.0255, 0.0259, 0.0278], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 01:08:40,780 INFO [train.py:901] (0/2) Epoch 19, batch 2250, loss[loss=0.153, simple_loss=0.2276, pruned_loss=0.03923, over 7304.00 frames. ], tot_loss[loss=0.1534, simple_loss=0.2302, pruned_loss=0.03828, over 1444077.38 frames. ], batch size: 68, lr: 8.61e-03, grad_scale: 8.0 +2023-03-21 01:08:52,097 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 01:08:52,599 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 01:08:54,155 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=53108.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:08:58,789 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=53117.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:09:01,067 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.90 vs. limit=2.0 +2023-03-21 01:09:01,687 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.241e+02 1.920e+02 2.278e+02 2.763e+02 7.620e+02, threshold=4.557e+02, percent-clipped=1.0 +2023-03-21 01:09:02,047 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.96 vs. limit=2.0 +2023-03-21 01:09:02,909 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6989, 3.0051, 3.3135, 2.9919, 2.7428, 2.7178, 3.5121, 2.7711], + device='cuda:0'), covar=tensor([0.0297, 0.0275, 0.0389, 0.0378, 0.0507, 0.0644, 0.0405, 0.1282], + device='cuda:0'), in_proj_covar=tensor([0.0323, 0.0327, 0.0264, 0.0346, 0.0309, 0.0303, 0.0330, 0.0294], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:09:03,833 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4648, 5.0041, 5.0038, 4.9542, 4.7876, 4.5258, 5.0747, 4.8292], + device='cuda:0'), covar=tensor([0.0468, 0.0410, 0.0414, 0.0536, 0.0435, 0.0376, 0.0363, 0.0560], + device='cuda:0'), in_proj_covar=tensor([0.0185, 0.0217, 0.0160, 0.0159, 0.0132, 0.0194, 0.0168, 0.0130], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:09:05,281 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 01:09:06,737 INFO [train.py:901] (0/2) Epoch 19, batch 2300, loss[loss=0.1244, simple_loss=0.2006, pruned_loss=0.02405, over 7006.00 frames. ], tot_loss[loss=0.1529, simple_loss=0.23, pruned_loss=0.03791, over 1445013.12 frames. ], batch size: 35, lr: 8.61e-03, grad_scale: 8.0 +2023-03-21 01:09:08,925 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6351, 2.1559, 2.1470, 1.8992, 2.3338, 2.1059, 1.9683, 1.4012], + device='cuda:0'), covar=tensor([0.0486, 0.0278, 0.0207, 0.0202, 0.0330, 0.0194, 0.0187, 0.0259], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0026, 0.0025, 0.0025, 0.0026, 0.0025, 0.0029, 0.0028], + device='cuda:0'), out_proj_covar=tensor([7.0666e-05, 6.8719e-05, 6.3016e-05, 6.2090e-05, 6.6699e-05, 6.4221e-05, + 7.0781e-05, 7.1686e-05], device='cuda:0') +2023-03-21 01:09:24,060 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0384, 4.5227, 4.4548, 4.4860, 4.4534, 4.1160, 4.5921, 4.4434], + device='cuda:0'), covar=tensor([0.0414, 0.0370, 0.0379, 0.0393, 0.0297, 0.0349, 0.0306, 0.0445], + device='cuda:0'), in_proj_covar=tensor([0.0182, 0.0213, 0.0157, 0.0158, 0.0129, 0.0191, 0.0165, 0.0127], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:09:32,437 INFO [train.py:901] (0/2) Epoch 19, batch 2350, loss[loss=0.1396, simple_loss=0.205, pruned_loss=0.03708, over 7006.00 frames. ], tot_loss[loss=0.1532, simple_loss=0.2301, pruned_loss=0.03818, over 1444044.66 frames. ], batch size: 35, lr: 8.60e-03, grad_scale: 8.0 +2023-03-21 01:09:36,207 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0698, 3.5752, 3.7195, 3.7682, 3.6141, 3.7266, 4.0171, 3.4960], + device='cuda:0'), covar=tensor([0.0204, 0.0187, 0.0146, 0.0181, 0.0495, 0.0114, 0.0155, 0.0182], + device='cuda:0'), in_proj_covar=tensor([0.0074, 0.0081, 0.0081, 0.0072, 0.0144, 0.0091, 0.0085, 0.0090], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:09:47,703 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=53211.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 01:09:52,503 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 01:09:53,471 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.463e+02 2.164e+02 2.497e+02 3.144e+02 6.644e+02, threshold=4.994e+02, percent-clipped=6.0 +2023-03-21 01:09:59,152 INFO [train.py:901] (0/2) Epoch 19, batch 2400, loss[loss=0.1237, simple_loss=0.1955, pruned_loss=0.02594, over 6987.00 frames. ], tot_loss[loss=0.1529, simple_loss=0.2299, pruned_loss=0.03797, over 1442241.10 frames. ], batch size: 35, lr: 8.60e-03, grad_scale: 8.0 +2023-03-21 01:09:59,678 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 01:10:03,309 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=53241.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:10:03,604 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.70 vs. limit=2.0 +2023-03-21 01:10:06,937 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3139, 1.6072, 1.2426, 1.5784, 1.4459, 1.3240, 1.2853, 1.1996], + device='cuda:0'), covar=tensor([0.0130, 0.0102, 0.0261, 0.0116, 0.0077, 0.0081, 0.0138, 0.0124], + device='cuda:0'), in_proj_covar=tensor([0.0024, 0.0024, 0.0024, 0.0024, 0.0023, 0.0023, 0.0025, 0.0032], + device='cuda:0'), out_proj_covar=tensor([2.8662e-05, 2.7181e-05, 2.7527e-05, 2.6888e-05, 2.7534e-05, 2.5746e-05, + 2.8671e-05, 3.6846e-05], device='cuda:0') +2023-03-21 01:10:10,336 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 01:10:10,571 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.27 vs. limit=2.0 +2023-03-21 01:10:12,315 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=53259.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 01:10:13,230 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 01:10:25,029 INFO [train.py:901] (0/2) Epoch 19, batch 2450, loss[loss=0.1409, simple_loss=0.2182, pruned_loss=0.03178, over 7314.00 frames. ], tot_loss[loss=0.1526, simple_loss=0.2295, pruned_loss=0.03781, over 1442195.56 frames. ], batch size: 59, lr: 8.59e-03, grad_scale: 8.0 +2023-03-21 01:10:34,758 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=53302.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:10:35,775 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=53304.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:10:40,224 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 01:10:41,383 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=53315.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:10:45,884 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.369e+02 1.898e+02 2.232e+02 2.806e+02 4.781e+02, threshold=4.465e+02, percent-clipped=0.0 +2023-03-21 01:10:46,937 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.0211, 4.4734, 4.3784, 4.9061, 4.8456, 4.9188, 4.2931, 4.5705], + device='cuda:0'), covar=tensor([0.0743, 0.2430, 0.2427, 0.1129, 0.0940, 0.1106, 0.0765, 0.0848], + device='cuda:0'), in_proj_covar=tensor([0.0211, 0.0327, 0.0265, 0.0261, 0.0192, 0.0323, 0.0189, 0.0233], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:10:50,955 INFO [train.py:901] (0/2) Epoch 19, batch 2500, loss[loss=0.1524, simple_loss=0.2305, pruned_loss=0.03718, over 7267.00 frames. ], tot_loss[loss=0.1522, simple_loss=0.2293, pruned_loss=0.03759, over 1440510.94 frames. ], batch size: 77, lr: 8.59e-03, grad_scale: 8.0 +2023-03-21 01:11:05,993 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 01:11:07,702 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=53365.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:11:08,735 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=53367.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:11:13,291 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=53376.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:11:16,617 INFO [train.py:901] (0/2) Epoch 19, batch 2550, loss[loss=0.1277, simple_loss=0.1904, pruned_loss=0.03255, over 7081.00 frames. ], tot_loss[loss=0.1514, simple_loss=0.2282, pruned_loss=0.03729, over 1439235.37 frames. ], batch size: 35, lr: 8.59e-03, grad_scale: 8.0 +2023-03-21 01:11:29,597 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=53408.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:11:32,129 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=53412.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:11:37,567 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.453e+02 2.065e+02 2.474e+02 2.931e+02 4.379e+02, threshold=4.948e+02, percent-clipped=0.0 +2023-03-21 01:11:40,256 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=53428.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 01:11:42,559 INFO [train.py:901] (0/2) Epoch 19, batch 2600, loss[loss=0.1233, simple_loss=0.2008, pruned_loss=0.02292, over 7110.00 frames. ], tot_loss[loss=0.1516, simple_loss=0.2282, pruned_loss=0.03745, over 1439737.42 frames. ], batch size: 41, lr: 8.58e-03, grad_scale: 8.0 +2023-03-21 01:11:51,591 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8737, 2.4187, 1.7388, 2.8566, 2.4113, 2.7468, 1.9259, 2.6317], + device='cuda:0'), covar=tensor([0.1805, 0.0840, 0.3205, 0.0699, 0.0124, 0.0151, 0.0223, 0.0249], + device='cuda:0'), in_proj_covar=tensor([0.0271, 0.0229, 0.0266, 0.0258, 0.0151, 0.0151, 0.0184, 0.0197], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:11:53,994 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=53456.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:11:59,757 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.69 vs. limit=2.0 +2023-03-21 01:12:07,313 INFO [train.py:901] (0/2) Epoch 19, batch 2650, loss[loss=0.1624, simple_loss=0.2402, pruned_loss=0.04235, over 7303.00 frames. ], tot_loss[loss=0.1511, simple_loss=0.2277, pruned_loss=0.03724, over 1440926.32 frames. ], batch size: 49, lr: 8.58e-03, grad_scale: 8.0 +2023-03-21 01:12:27,465 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.393e+02 1.972e+02 2.315e+02 2.811e+02 6.076e+02, threshold=4.629e+02, percent-clipped=2.0 +2023-03-21 01:12:32,415 INFO [train.py:901] (0/2) Epoch 19, batch 2700, loss[loss=0.146, simple_loss=0.2304, pruned_loss=0.03086, over 7313.00 frames. ], tot_loss[loss=0.1512, simple_loss=0.228, pruned_loss=0.03718, over 1442909.32 frames. ], batch size: 83, lr: 8.57e-03, grad_scale: 8.0 +2023-03-21 01:12:44,221 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1251, 4.5924, 4.5848, 4.5160, 4.5675, 4.2230, 4.6672, 4.5040], + device='cuda:0'), covar=tensor([0.0418, 0.0395, 0.0375, 0.0454, 0.0270, 0.0365, 0.0288, 0.0397], + device='cuda:0'), in_proj_covar=tensor([0.0187, 0.0219, 0.0160, 0.0162, 0.0130, 0.0195, 0.0170, 0.0129], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:12:56,805 INFO [train.py:901] (0/2) Epoch 19, batch 2750, loss[loss=0.1594, simple_loss=0.2382, pruned_loss=0.04032, over 7275.00 frames. ], tot_loss[loss=0.1521, simple_loss=0.229, pruned_loss=0.03766, over 1444447.67 frames. ], batch size: 77, lr: 8.57e-03, grad_scale: 8.0 +2023-03-21 01:13:03,810 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=53597.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:13:07,342 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.80 vs. limit=2.0 +2023-03-21 01:13:16,768 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.549e+02 2.108e+02 2.376e+02 2.842e+02 5.465e+02, threshold=4.753e+02, percent-clipped=1.0 +2023-03-21 01:13:21,591 INFO [train.py:901] (0/2) Epoch 19, batch 2800, loss[loss=0.1638, simple_loss=0.233, pruned_loss=0.04734, over 7348.00 frames. ], tot_loss[loss=0.1527, simple_loss=0.2297, pruned_loss=0.03789, over 1444548.90 frames. ], batch size: 73, lr: 8.57e-03, grad_scale: 8.0 +2023-03-21 01:13:34,417 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-19.pt +2023-03-21 01:13:51,701 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 01:13:51,761 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 01:13:51,787 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 01:13:51,807 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 01:13:52,002 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 01:13:52,018 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 01:13:52,030 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 01:13:52,134 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 01:13:52,566 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 01:13:52,582 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 01:13:52,600 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 01:13:52,782 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 01:13:52,788 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 01:13:52,797 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 01:13:52,948 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 01:13:52,951 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 01:13:53,021 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 01:13:55,254 INFO [train.py:901] (0/2) Epoch 20, batch 0, loss[loss=0.1635, simple_loss=0.2377, pruned_loss=0.04469, over 7333.00 frames. ], tot_loss[loss=0.1635, simple_loss=0.2377, pruned_loss=0.04469, over 7333.00 frames. ], batch size: 51, lr: 8.36e-03, grad_scale: 8.0 +2023-03-21 01:13:55,256 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 01:14:07,393 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9273, 3.6980, 3.4384, 3.5225, 3.6400, 3.4764, 3.5633, 3.3472], + device='cuda:0'), covar=tensor([0.0116, 0.0148, 0.0178, 0.0202, 0.0393, 0.0133, 0.0228, 0.0223], + device='cuda:0'), in_proj_covar=tensor([0.0076, 0.0082, 0.0082, 0.0072, 0.0147, 0.0092, 0.0087, 0.0091], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:14:09,615 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4857, 2.4874, 1.9965, 2.9055, 1.9362, 2.3779, 1.4164, 2.0341], + device='cuda:0'), covar=tensor([0.0352, 0.0590, 0.2319, 0.0575, 0.0386, 0.0410, 0.3071, 0.1351], + device='cuda:0'), in_proj_covar=tensor([0.0263, 0.0245, 0.0299, 0.0254, 0.0266, 0.0257, 0.0261, 0.0275], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 01:14:14,060 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3205, 4.5525, 4.5400, 4.5257, 4.4448, 4.2542, 4.6330, 4.3623], + device='cuda:0'), covar=tensor([0.0350, 0.0351, 0.0362, 0.0413, 0.0286, 0.0285, 0.0242, 0.0468], + device='cuda:0'), in_proj_covar=tensor([0.0188, 0.0220, 0.0160, 0.0162, 0.0131, 0.0196, 0.0169, 0.0130], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:14:15,730 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5297, 3.1408, 3.3322, 3.5291, 3.4857, 3.5573, 3.7562, 3.5525], + device='cuda:0'), covar=tensor([0.0027, 0.0095, 0.0038, 0.0033, 0.0038, 0.0031, 0.0024, 0.0048], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0055, 0.0047, 0.0046, 0.0046, 0.0049, 0.0047, 0.0059], + device='cuda:0'), out_proj_covar=tensor([8.1165e-05, 1.3087e-04, 1.0709e-04, 9.7233e-05, 9.7141e-05, 1.0362e-04, + 1.0929e-04, 1.2865e-04], device='cuda:0') +2023-03-21 01:14:19,520 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3671, 1.5797, 1.4058, 1.4190, 1.3978, 1.4582, 1.4504, 1.1706], + device='cuda:0'), covar=tensor([0.0077, 0.0084, 0.0192, 0.0117, 0.0080, 0.0056, 0.0079, 0.0092], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0024, 0.0024, 0.0024, 0.0024, 0.0023, 0.0025, 0.0032], + device='cuda:0'), out_proj_covar=tensor([2.8948e-05, 2.7764e-05, 2.8005e-05, 2.6843e-05, 2.7822e-05, 2.6139e-05, + 2.8681e-05, 3.7216e-05], device='cuda:0') +2023-03-21 01:14:20,041 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4630, 1.9370, 2.1114, 1.8758, 2.2085, 1.8593, 1.8316, 1.3287], + device='cuda:0'), covar=tensor([0.0231, 0.0365, 0.0201, 0.0207, 0.0302, 0.0300, 0.0188, 0.0231], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0027, 0.0025, 0.0024, 0.0026, 0.0025, 0.0028, 0.0028], + device='cuda:0'), out_proj_covar=tensor([7.0782e-05, 6.9267e-05, 6.2981e-05, 6.1127e-05, 6.6113e-05, 6.4225e-05, + 6.9364e-05, 7.0859e-05], device='cuda:0') +2023-03-21 01:14:21,488 INFO [train.py:935] (0/2) Epoch 20, validation: loss=0.166, simple_loss=0.2535, pruned_loss=0.03929, over 1622729.00 frames. +2023-03-21 01:14:21,489 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 01:14:23,062 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=53660.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:14:27,905 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 01:14:28,478 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=53671.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:14:38,520 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 01:14:45,590 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 01:14:46,585 INFO [train.py:901] (0/2) Epoch 20, batch 50, loss[loss=0.145, simple_loss=0.2277, pruned_loss=0.03117, over 7272.00 frames. ], tot_loss[loss=0.1532, simple_loss=0.2306, pruned_loss=0.03785, over 328201.60 frames. ], batch size: 77, lr: 8.35e-03, grad_scale: 8.0 +2023-03-21 01:14:47,668 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 01:14:49,296 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=53712.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:14:50,694 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 01:14:52,288 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4543, 1.2360, 1.6152, 1.9979, 1.8347, 2.0755, 1.5968, 1.8500], + device='cuda:0'), covar=tensor([0.1932, 0.3495, 0.1000, 0.0576, 0.1503, 0.1483, 0.2838, 0.2271], + device='cuda:0'), in_proj_covar=tensor([0.0052, 0.0060, 0.0047, 0.0042, 0.0046, 0.0046, 0.0068, 0.0046], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0002, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 01:14:55,336 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.419e+02 2.061e+02 2.323e+02 2.773e+02 4.898e+02, threshold=4.647e+02, percent-clipped=1.0 +2023-03-21 01:14:55,433 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=53723.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 01:15:08,188 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 01:15:08,203 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 01:15:11,798 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2080, 1.4751, 1.2765, 1.2788, 1.1616, 1.3974, 1.3195, 1.1600], + device='cuda:0'), covar=tensor([0.0148, 0.0089, 0.0237, 0.0110, 0.0122, 0.0096, 0.0107, 0.0094], + device='cuda:0'), in_proj_covar=tensor([0.0024, 0.0024, 0.0024, 0.0024, 0.0024, 0.0023, 0.0025, 0.0032], + device='cuda:0'), out_proj_covar=tensor([2.8683e-05, 2.7663e-05, 2.7782e-05, 2.6868e-05, 2.7691e-05, 2.6149e-05, + 2.8633e-05, 3.6799e-05], device='cuda:0') +2023-03-21 01:15:13,130 INFO [train.py:901] (0/2) Epoch 20, batch 100, loss[loss=0.1445, simple_loss=0.217, pruned_loss=0.03598, over 7208.00 frames. ], tot_loss[loss=0.1523, simple_loss=0.23, pruned_loss=0.03733, over 575305.79 frames. ], batch size: 39, lr: 8.35e-03, grad_scale: 8.0 +2023-03-21 01:15:14,657 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=53760.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:15:20,983 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0047, 3.2920, 2.5567, 2.9578, 2.9344, 2.6700, 3.0527, 2.7543], + device='cuda:0'), covar=tensor([0.0763, 0.0515, 0.2292, 0.1478, 0.1905, 0.0523, 0.1411, 0.2009], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0047, 0.0053, 0.0048, 0.0047, 0.0047, 0.0048, 0.0044], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:15:38,024 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.91 vs. limit=2.0 +2023-03-21 01:15:38,713 INFO [train.py:901] (0/2) Epoch 20, batch 150, loss[loss=0.1553, simple_loss=0.2356, pruned_loss=0.03751, over 7322.00 frames. ], tot_loss[loss=0.1528, simple_loss=0.2305, pruned_loss=0.0376, over 769907.01 frames. ], batch size: 75, lr: 8.35e-03, grad_scale: 8.0 +2023-03-21 01:15:39,678 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.00 vs. limit=2.0 +2023-03-21 01:15:47,349 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.409e+02 1.894e+02 2.222e+02 2.716e+02 9.009e+02, threshold=4.445e+02, percent-clipped=3.0 +2023-03-21 01:15:59,403 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.99 vs. limit=5.0 +2023-03-21 01:16:05,169 INFO [train.py:901] (0/2) Epoch 20, batch 200, loss[loss=0.1235, simple_loss=0.2046, pruned_loss=0.02122, over 7219.00 frames. ], tot_loss[loss=0.1522, simple_loss=0.2299, pruned_loss=0.0372, over 921789.16 frames. ], batch size: 39, lr: 8.34e-03, grad_scale: 8.0 +2023-03-21 01:16:08,191 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 01:16:12,358 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 01:16:16,739 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-03-21 01:16:18,299 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 01:16:26,201 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=53897.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:16:31,782 INFO [train.py:901] (0/2) Epoch 20, batch 250, loss[loss=0.1505, simple_loss=0.2317, pruned_loss=0.03459, over 7253.00 frames. ], tot_loss[loss=0.1516, simple_loss=0.2294, pruned_loss=0.03693, over 1038514.23 frames. ], batch size: 55, lr: 8.34e-03, grad_scale: 8.0 +2023-03-21 01:16:31,818 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 01:16:38,059 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.34 vs. limit=2.0 +2023-03-21 01:16:39,787 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.349e+02 2.005e+02 2.500e+02 2.996e+02 9.929e+02, threshold=5.000e+02, percent-clipped=6.0 +2023-03-21 01:16:50,911 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=53945.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:16:52,854 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 01:16:56,807 INFO [train.py:901] (0/2) Epoch 20, batch 300, loss[loss=0.1684, simple_loss=0.2477, pruned_loss=0.04455, over 7285.00 frames. ], tot_loss[loss=0.1524, simple_loss=0.2299, pruned_loss=0.03747, over 1128569.16 frames. ], batch size: 57, lr: 8.33e-03, grad_scale: 8.0 +2023-03-21 01:16:58,427 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=53960.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:17:01,457 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 01:17:02,606 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5303, 2.5321, 2.3485, 3.7124, 1.5108, 3.4818, 1.4620, 3.1004], + device='cuda:0'), covar=tensor([0.0102, 0.0889, 0.1537, 0.0121, 0.3685, 0.0126, 0.1033, 0.0206], + device='cuda:0'), in_proj_covar=tensor([0.0182, 0.0263, 0.0286, 0.0176, 0.0271, 0.0192, 0.0257, 0.0224], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:17:04,045 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=53971.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:17:21,306 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0349, 3.5398, 3.7887, 3.9647, 3.9526, 4.0373, 4.0647, 3.9092], + device='cuda:0'), covar=tensor([0.0027, 0.0086, 0.0031, 0.0034, 0.0028, 0.0028, 0.0028, 0.0048], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0055, 0.0048, 0.0046, 0.0046, 0.0050, 0.0046, 0.0060], + device='cuda:0'), out_proj_covar=tensor([8.1557e-05, 1.3153e-04, 1.0807e-04, 9.7553e-05, 9.5738e-05, 1.0428e-04, + 1.0749e-04, 1.2920e-04], device='cuda:0') +2023-03-21 01:17:23,227 INFO [train.py:901] (0/2) Epoch 20, batch 350, loss[loss=0.1458, simple_loss=0.2259, pruned_loss=0.03288, over 7324.00 frames. ], tot_loss[loss=0.1529, simple_loss=0.2302, pruned_loss=0.03781, over 1200220.05 frames. ], batch size: 75, lr: 8.33e-03, grad_scale: 8.0 +2023-03-21 01:17:23,801 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=54008.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:17:27,884 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0238, 4.1719, 3.8439, 4.2097, 3.8542, 4.2013, 4.4557, 4.5027], + device='cuda:0'), covar=tensor([0.0189, 0.0152, 0.0221, 0.0165, 0.0343, 0.0292, 0.0236, 0.0165], + device='cuda:0'), in_proj_covar=tensor([0.0109, 0.0108, 0.0102, 0.0108, 0.0101, 0.0089, 0.0087, 0.0083], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:17:29,357 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=54019.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:17:31,242 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.231e+02 1.892e+02 2.279e+02 2.732e+02 6.693e+02, threshold=4.558e+02, percent-clipped=2.0 +2023-03-21 01:17:31,396 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=54023.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 01:17:37,239 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 01:17:48,243 INFO [train.py:901] (0/2) Epoch 20, batch 400, loss[loss=0.1646, simple_loss=0.2424, pruned_loss=0.04342, over 7304.00 frames. ], tot_loss[loss=0.1521, simple_loss=0.2292, pruned_loss=0.03754, over 1253384.56 frames. ], batch size: 83, lr: 8.33e-03, grad_scale: 8.0 +2023-03-21 01:17:55,876 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=54071.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:18:02,097 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7475, 2.8227, 2.5038, 2.7390, 2.7604, 2.5571, 2.8155, 2.5831], + device='cuda:0'), covar=tensor([0.0781, 0.0820, 0.0986, 0.0874, 0.0965, 0.0762, 0.1040, 0.0975], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0047, 0.0053, 0.0047, 0.0047, 0.0047, 0.0048, 0.0043], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:18:14,574 INFO [train.py:901] (0/2) Epoch 20, batch 450, loss[loss=0.1553, simple_loss=0.2328, pruned_loss=0.03897, over 7323.00 frames. ], tot_loss[loss=0.1528, simple_loss=0.2298, pruned_loss=0.03788, over 1296876.15 frames. ], batch size: 59, lr: 8.32e-03, grad_scale: 8.0 +2023-03-21 01:18:20,138 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 01:18:20,624 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 01:18:22,522 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.438e+02 2.041e+02 2.494e+02 3.021e+02 4.369e+02, threshold=4.988e+02, percent-clipped=0.0 +2023-03-21 01:18:40,471 INFO [train.py:901] (0/2) Epoch 20, batch 500, loss[loss=0.1487, simple_loss=0.2212, pruned_loss=0.03807, over 7254.00 frames. ], tot_loss[loss=0.1523, simple_loss=0.2293, pruned_loss=0.03762, over 1329440.78 frames. ], batch size: 47, lr: 8.32e-03, grad_scale: 8.0 +2023-03-21 01:18:41,119 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9901, 2.7606, 3.1716, 3.0457, 3.2866, 3.0429, 2.6613, 3.1107], + device='cuda:0'), covar=tensor([0.2291, 0.0775, 0.1299, 0.2400, 0.0753, 0.0870, 0.2280, 0.2034], + device='cuda:0'), in_proj_covar=tensor([0.0045, 0.0053, 0.0041, 0.0040, 0.0040, 0.0037, 0.0056, 0.0043], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:18:46,028 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2293, 4.7469, 4.8336, 4.7491, 4.6691, 4.3477, 4.8836, 4.7102], + device='cuda:0'), covar=tensor([0.0487, 0.0396, 0.0352, 0.0390, 0.0319, 0.0373, 0.0295, 0.0477], + device='cuda:0'), in_proj_covar=tensor([0.0188, 0.0218, 0.0164, 0.0160, 0.0130, 0.0197, 0.0170, 0.0130], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:18:52,568 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 01:18:54,076 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 01:18:54,608 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 01:18:57,168 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 01:19:02,206 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 01:19:06,575 INFO [train.py:901] (0/2) Epoch 20, batch 550, loss[loss=0.1448, simple_loss=0.2264, pruned_loss=0.03161, over 7276.00 frames. ], tot_loss[loss=0.152, simple_loss=0.2292, pruned_loss=0.0374, over 1354798.40 frames. ], batch size: 47, lr: 8.32e-03, grad_scale: 8.0 +2023-03-21 01:19:08,206 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4420, 2.3310, 2.2075, 3.5154, 1.4920, 3.4431, 1.3327, 3.1439], + device='cuda:0'), covar=tensor([0.0129, 0.0996, 0.1664, 0.0123, 0.4189, 0.0129, 0.1220, 0.0422], + device='cuda:0'), in_proj_covar=tensor([0.0183, 0.0263, 0.0282, 0.0175, 0.0269, 0.0191, 0.0254, 0.0224], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:19:10,702 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2790, 1.0064, 1.3759, 1.7461, 1.5101, 1.7030, 1.1777, 1.7040], + device='cuda:0'), covar=tensor([0.1921, 0.3406, 0.1015, 0.0600, 0.1283, 0.1231, 0.1337, 0.1361], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0060, 0.0047, 0.0042, 0.0047, 0.0047, 0.0068, 0.0048], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 01:19:14,141 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 01:19:14,606 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.132e+02 1.926e+02 2.263e+02 2.730e+02 5.297e+02, threshold=4.525e+02, percent-clipped=2.0 +2023-03-21 01:19:22,868 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9005, 3.4926, 3.6097, 3.5215, 3.3941, 3.4502, 3.7169, 3.2897], + device='cuda:0'), covar=tensor([0.0134, 0.0195, 0.0140, 0.0205, 0.0501, 0.0136, 0.0172, 0.0189], + device='cuda:0'), in_proj_covar=tensor([0.0077, 0.0084, 0.0082, 0.0073, 0.0147, 0.0093, 0.0086, 0.0090], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:19:23,262 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 01:19:26,666 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 01:19:32,392 INFO [train.py:901] (0/2) Epoch 20, batch 600, loss[loss=0.1528, simple_loss=0.2335, pruned_loss=0.03606, over 7351.00 frames. ], tot_loss[loss=0.1512, simple_loss=0.2285, pruned_loss=0.037, over 1372975.12 frames. ], batch size: 54, lr: 8.31e-03, grad_scale: 8.0 +2023-03-21 01:19:32,526 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5729, 1.8061, 2.1219, 1.8175, 1.9668, 1.9134, 1.8250, 1.4849], + device='cuda:0'), covar=tensor([0.0381, 0.0591, 0.0219, 0.0260, 0.0540, 0.0450, 0.0209, 0.0263], + device='cuda:0'), in_proj_covar=tensor([0.0029, 0.0028, 0.0025, 0.0025, 0.0027, 0.0025, 0.0028, 0.0028], + device='cuda:0'), out_proj_covar=tensor([7.2660e-05, 7.1505e-05, 6.2822e-05, 6.2702e-05, 6.7835e-05, 6.5312e-05, + 7.0361e-05, 7.2254e-05], device='cuda:0') +2023-03-21 01:19:33,915 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 01:19:50,697 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 01:19:58,396 INFO [train.py:901] (0/2) Epoch 20, batch 650, loss[loss=0.1523, simple_loss=0.2371, pruned_loss=0.03378, over 7299.00 frames. ], tot_loss[loss=0.1524, simple_loss=0.2295, pruned_loss=0.03764, over 1388890.59 frames. ], batch size: 68, lr: 8.31e-03, grad_scale: 8.0 +2023-03-21 01:19:59,896 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 01:20:06,396 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.393e+02 2.135e+02 2.443e+02 2.990e+02 6.250e+02, threshold=4.886e+02, percent-clipped=3.0 +2023-03-21 01:20:08,163 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.31 vs. limit=2.0 +2023-03-21 01:20:16,743 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 01:20:24,907 INFO [train.py:901] (0/2) Epoch 20, batch 700, loss[loss=0.111, simple_loss=0.1754, pruned_loss=0.0233, over 6605.00 frames. ], tot_loss[loss=0.152, simple_loss=0.2289, pruned_loss=0.03757, over 1400619.89 frames. ], batch size: 29, lr: 8.30e-03, grad_scale: 8.0 +2023-03-21 01:20:25,452 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 01:20:49,378 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 01:20:49,843 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 01:20:50,346 INFO [train.py:901] (0/2) Epoch 20, batch 750, loss[loss=0.1427, simple_loss=0.2286, pruned_loss=0.02844, over 7267.00 frames. ], tot_loss[loss=0.1515, simple_loss=0.2287, pruned_loss=0.03714, over 1409368.82 frames. ], batch size: 52, lr: 8.30e-03, grad_scale: 8.0 +2023-03-21 01:20:58,465 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.383e+02 1.864e+02 2.237e+02 2.756e+02 4.734e+02, threshold=4.474e+02, percent-clipped=0.0 +2023-03-21 01:21:00,244 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8138, 3.1063, 2.5995, 2.8652, 2.9155, 2.4483, 2.6989, 2.3455], + device='cuda:0'), covar=tensor([0.1101, 0.0925, 0.1432, 0.1394, 0.1418, 0.0986, 0.2157, 0.2213], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0045, 0.0053, 0.0047, 0.0047, 0.0047, 0.0047, 0.0044], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:21:04,596 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 01:21:09,646 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 01:21:16,093 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 01:21:16,595 INFO [train.py:901] (0/2) Epoch 20, batch 800, loss[loss=0.1324, simple_loss=0.215, pruned_loss=0.02489, over 7273.00 frames. ], tot_loss[loss=0.1513, simple_loss=0.2281, pruned_loss=0.03722, over 1411998.75 frames. ], batch size: 66, lr: 8.30e-03, grad_scale: 8.0 +2023-03-21 01:21:17,611 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 01:21:28,188 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 01:21:31,790 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6879, 5.1696, 5.2053, 5.1293, 4.9836, 4.7404, 5.2761, 5.0662], + device='cuda:0'), covar=tensor([0.0430, 0.0409, 0.0405, 0.0486, 0.0307, 0.0305, 0.0303, 0.0490], + device='cuda:0'), in_proj_covar=tensor([0.0187, 0.0219, 0.0164, 0.0161, 0.0128, 0.0197, 0.0171, 0.0131], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:21:41,946 INFO [train.py:901] (0/2) Epoch 20, batch 850, loss[loss=0.1496, simple_loss=0.2271, pruned_loss=0.03609, over 7337.00 frames. ], tot_loss[loss=0.151, simple_loss=0.2278, pruned_loss=0.03712, over 1419234.77 frames. ], batch size: 61, lr: 8.29e-03, grad_scale: 8.0 +2023-03-21 01:21:47,563 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 01:21:48,092 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 01:21:50,595 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.337e+02 1.930e+02 2.357e+02 2.805e+02 7.911e+02, threshold=4.713e+02, percent-clipped=4.0 +2023-03-21 01:21:53,614 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 01:21:57,784 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 01:22:02,345 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.95 vs. limit=2.0 +2023-03-21 01:22:08,504 INFO [train.py:901] (0/2) Epoch 20, batch 900, loss[loss=0.1503, simple_loss=0.2258, pruned_loss=0.03743, over 7269.00 frames. ], tot_loss[loss=0.1516, simple_loss=0.2286, pruned_loss=0.03736, over 1424947.02 frames. ], batch size: 52, lr: 8.29e-03, grad_scale: 16.0 +2023-03-21 01:22:32,022 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.02 vs. limit=5.0 +2023-03-21 01:22:34,244 INFO [train.py:901] (0/2) Epoch 20, batch 950, loss[loss=0.1581, simple_loss=0.2368, pruned_loss=0.03969, over 7357.00 frames. ], tot_loss[loss=0.1518, simple_loss=0.2285, pruned_loss=0.0375, over 1428066.54 frames. ], batch size: 63, lr: 8.29e-03, grad_scale: 16.0 +2023-03-21 01:22:37,930 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 01:22:38,679 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.87 vs. limit=5.0 +2023-03-21 01:22:43,567 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.344e+02 2.000e+02 2.324e+02 2.740e+02 5.725e+02, threshold=4.649e+02, percent-clipped=1.0 +2023-03-21 01:23:00,956 INFO [train.py:901] (0/2) Epoch 20, batch 1000, loss[loss=0.1412, simple_loss=0.2224, pruned_loss=0.03, over 7279.00 frames. ], tot_loss[loss=0.1523, simple_loss=0.229, pruned_loss=0.0378, over 1428936.16 frames. ], batch size: 47, lr: 8.28e-03, grad_scale: 16.0 +2023-03-21 01:23:01,482 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 01:23:04,585 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=54664.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 01:23:22,676 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 01:23:27,288 INFO [train.py:901] (0/2) Epoch 20, batch 1050, loss[loss=0.1385, simple_loss=0.2249, pruned_loss=0.02603, over 7293.00 frames. ], tot_loss[loss=0.1519, simple_loss=0.229, pruned_loss=0.03741, over 1433533.17 frames. ], batch size: 86, lr: 8.28e-03, grad_scale: 16.0 +2023-03-21 01:23:35,312 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.346e+02 2.128e+02 2.433e+02 2.952e+02 5.627e+02, threshold=4.867e+02, percent-clipped=1.0 +2023-03-21 01:23:36,399 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=54725.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 01:23:43,288 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 01:23:47,821 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 01:23:50,349 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.6464, 5.2103, 5.1026, 5.6551, 5.5155, 5.5822, 5.1805, 5.0695], + device='cuda:0'), covar=tensor([0.0759, 0.2096, 0.1905, 0.0828, 0.0713, 0.1025, 0.0552, 0.0939], + device='cuda:0'), in_proj_covar=tensor([0.0212, 0.0322, 0.0260, 0.0258, 0.0187, 0.0318, 0.0187, 0.0223], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:23:50,912 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8198, 2.1603, 2.2970, 2.0107, 2.3952, 2.0449, 1.9163, 1.4701], + device='cuda:0'), covar=tensor([0.0346, 0.0384, 0.0157, 0.0241, 0.0328, 0.0339, 0.0245, 0.0263], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0027, 0.0024, 0.0025, 0.0026, 0.0025, 0.0028, 0.0028], + device='cuda:0'), out_proj_covar=tensor([7.1019e-05, 6.9226e-05, 6.1356e-05, 6.1906e-05, 6.5982e-05, 6.3479e-05, + 6.9762e-05, 7.1624e-05], device='cuda:0') +2023-03-21 01:23:52,341 INFO [train.py:901] (0/2) Epoch 20, batch 1100, loss[loss=0.1487, simple_loss=0.2186, pruned_loss=0.03944, over 7316.00 frames. ], tot_loss[loss=0.1519, simple_loss=0.2289, pruned_loss=0.03748, over 1434532.40 frames. ], batch size: 49, lr: 8.27e-03, grad_scale: 16.0 +2023-03-21 01:24:16,956 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 01:24:16,968 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 01:24:18,964 INFO [train.py:901] (0/2) Epoch 20, batch 1150, loss[loss=0.1588, simple_loss=0.2311, pruned_loss=0.04325, over 7251.00 frames. ], tot_loss[loss=0.1522, simple_loss=0.2292, pruned_loss=0.0376, over 1438330.52 frames. ], batch size: 47, lr: 8.27e-03, grad_scale: 16.0 +2023-03-21 01:24:23,130 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0112, 2.9447, 1.8839, 3.3222, 2.2338, 2.7947, 1.4173, 1.8281], + device='cuda:0'), covar=tensor([0.0424, 0.0961, 0.2468, 0.0617, 0.0428, 0.0537, 0.3316, 0.1831], + device='cuda:0'), in_proj_covar=tensor([0.0266, 0.0253, 0.0303, 0.0259, 0.0266, 0.0259, 0.0268, 0.0280], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 01:24:26,908 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.285e+02 2.011e+02 2.444e+02 2.954e+02 6.699e+02, threshold=4.887e+02, percent-clipped=3.0 +2023-03-21 01:24:28,976 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 01:24:29,981 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 01:24:35,726 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3592, 2.7296, 2.2795, 3.8802, 1.6827, 3.3492, 1.5145, 2.9746], + device='cuda:0'), covar=tensor([0.0086, 0.0807, 0.1567, 0.0080, 0.3791, 0.0126, 0.1011, 0.0302], + device='cuda:0'), in_proj_covar=tensor([0.0183, 0.0263, 0.0284, 0.0175, 0.0269, 0.0190, 0.0257, 0.0225], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:24:36,671 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7168, 3.1530, 3.5557, 3.4417, 3.6304, 3.8264, 3.6151, 3.6126], + device='cuda:0'), covar=tensor([0.0026, 0.0086, 0.0033, 0.0047, 0.0035, 0.0023, 0.0045, 0.0047], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0055, 0.0047, 0.0047, 0.0045, 0.0049, 0.0045, 0.0060], + device='cuda:0'), out_proj_covar=tensor([8.0580e-05, 1.3000e-04, 1.0623e-04, 9.8566e-05, 9.4938e-05, 1.0344e-04, + 1.0454e-04, 1.2940e-04], device='cuda:0') +2023-03-21 01:24:44,026 INFO [train.py:901] (0/2) Epoch 20, batch 1200, loss[loss=0.1596, simple_loss=0.2387, pruned_loss=0.04025, over 7267.00 frames. ], tot_loss[loss=0.152, simple_loss=0.2291, pruned_loss=0.03742, over 1440369.57 frames. ], batch size: 52, lr: 8.27e-03, grad_scale: 16.0 +2023-03-21 01:24:46,641 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=54862.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:25:02,825 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 01:25:10,461 INFO [train.py:901] (0/2) Epoch 20, batch 1250, loss[loss=0.1535, simple_loss=0.2368, pruned_loss=0.03515, over 7309.00 frames. ], tot_loss[loss=0.1513, simple_loss=0.2285, pruned_loss=0.03704, over 1440094.71 frames. ], batch size: 80, lr: 8.26e-03, grad_scale: 16.0 +2023-03-21 01:25:18,343 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.491e+02 2.111e+02 2.421e+02 2.802e+02 5.340e+02, threshold=4.843e+02, percent-clipped=1.0 +2023-03-21 01:25:18,516 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=54923.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 01:25:25,964 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3409, 1.5926, 1.3035, 1.5952, 1.4547, 1.2928, 1.4064, 1.0712], + device='cuda:0'), covar=tensor([0.0093, 0.0117, 0.0210, 0.0100, 0.0090, 0.0096, 0.0068, 0.0111], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0025, 0.0024, 0.0024, 0.0023, 0.0024, 0.0025, 0.0032], + device='cuda:0'), out_proj_covar=tensor([2.9231e-05, 2.8589e-05, 2.7992e-05, 2.6782e-05, 2.7238e-05, 2.7275e-05, + 2.8552e-05, 3.7172e-05], device='cuda:0') +2023-03-21 01:25:27,304 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 01:25:30,832 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 01:25:32,358 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 01:25:35,327 INFO [train.py:901] (0/2) Epoch 20, batch 1300, loss[loss=0.1455, simple_loss=0.2243, pruned_loss=0.03335, over 7319.00 frames. ], tot_loss[loss=0.1515, simple_loss=0.2286, pruned_loss=0.03721, over 1438178.06 frames. ], batch size: 51, lr: 8.26e-03, grad_scale: 16.0 +2023-03-21 01:25:45,779 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2972, 2.8398, 3.0637, 3.0375, 2.5754, 2.3951, 3.0969, 2.3959], + device='cuda:0'), covar=tensor([0.0323, 0.0389, 0.0436, 0.0435, 0.0495, 0.0615, 0.0531, 0.1096], + device='cuda:0'), in_proj_covar=tensor([0.0326, 0.0331, 0.0266, 0.0353, 0.0305, 0.0303, 0.0338, 0.0287], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:25:55,472 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 01:25:58,313 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 01:26:01,812 INFO [train.py:901] (0/2) Epoch 20, batch 1350, loss[loss=0.1497, simple_loss=0.2319, pruned_loss=0.03376, over 7313.00 frames. ], tot_loss[loss=0.1511, simple_loss=0.2281, pruned_loss=0.03699, over 1437990.34 frames. ], batch size: 80, lr: 8.26e-03, grad_scale: 16.0 +2023-03-21 01:26:01,832 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 01:26:08,429 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=55020.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 01:26:09,865 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.235e+02 1.994e+02 2.329e+02 2.781e+02 4.639e+02, threshold=4.658e+02, percent-clipped=0.0 +2023-03-21 01:26:12,869 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 01:26:28,032 INFO [train.py:901] (0/2) Epoch 20, batch 1400, loss[loss=0.1652, simple_loss=0.2401, pruned_loss=0.04519, over 7337.00 frames. ], tot_loss[loss=0.1519, simple_loss=0.2288, pruned_loss=0.03751, over 1437732.61 frames. ], batch size: 61, lr: 8.25e-03, grad_scale: 16.0 +2023-03-21 01:26:36,672 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1251, 2.3856, 2.0738, 3.4586, 1.4744, 3.1453, 1.2763, 2.9382], + device='cuda:0'), covar=tensor([0.0082, 0.0938, 0.1579, 0.0088, 0.3584, 0.0112, 0.1044, 0.0254], + device='cuda:0'), in_proj_covar=tensor([0.0184, 0.0262, 0.0285, 0.0176, 0.0269, 0.0191, 0.0256, 0.0222], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:26:40,592 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1187, 3.7104, 3.7908, 3.7270, 3.6678, 3.6316, 3.9547, 3.5009], + device='cuda:0'), covar=tensor([0.0100, 0.0128, 0.0102, 0.0135, 0.0354, 0.0096, 0.0117, 0.0156], + device='cuda:0'), in_proj_covar=tensor([0.0074, 0.0081, 0.0079, 0.0072, 0.0143, 0.0090, 0.0084, 0.0089], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:26:46,538 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 01:26:53,078 INFO [train.py:901] (0/2) Epoch 20, batch 1450, loss[loss=0.1587, simple_loss=0.2424, pruned_loss=0.03752, over 7355.00 frames. ], tot_loss[loss=0.1515, simple_loss=0.2285, pruned_loss=0.03723, over 1438954.43 frames. ], batch size: 63, lr: 8.25e-03, grad_scale: 16.0 +2023-03-21 01:26:54,660 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1858, 3.4972, 2.7168, 3.2382, 3.1855, 2.3907, 3.1850, 2.9029], + device='cuda:0'), covar=tensor([0.0598, 0.0852, 0.1864, 0.1778, 0.2215, 0.1033, 0.1130, 0.1526], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0046, 0.0054, 0.0047, 0.0046, 0.0047, 0.0048, 0.0043], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:27:00,975 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.329e+02 1.941e+02 2.292e+02 2.919e+02 8.042e+02, threshold=4.584e+02, percent-clipped=3.0 +2023-03-21 01:27:10,127 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 01:27:14,244 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0179, 2.7542, 1.8916, 3.1277, 3.0811, 3.1996, 2.6174, 2.6882], + device='cuda:0'), covar=tensor([0.1957, 0.0882, 0.3211, 0.0620, 0.0183, 0.0152, 0.0280, 0.0272], + device='cuda:0'), in_proj_covar=tensor([0.0273, 0.0231, 0.0265, 0.0263, 0.0155, 0.0157, 0.0189, 0.0200], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:27:14,653 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=55148.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:27:19,149 INFO [train.py:901] (0/2) Epoch 20, batch 1500, loss[loss=0.1681, simple_loss=0.2555, pruned_loss=0.04037, over 6773.00 frames. ], tot_loss[loss=0.1508, simple_loss=0.228, pruned_loss=0.03681, over 1437706.74 frames. ], batch size: 107, lr: 8.24e-03, grad_scale: 16.0 +2023-03-21 01:27:26,850 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 01:27:44,732 INFO [train.py:901] (0/2) Epoch 20, batch 1550, loss[loss=0.1671, simple_loss=0.2376, pruned_loss=0.04826, over 7371.00 frames. ], tot_loss[loss=0.1518, simple_loss=0.2291, pruned_loss=0.03729, over 1439817.88 frames. ], batch size: 51, lr: 8.24e-03, grad_scale: 16.0 +2023-03-21 01:27:45,847 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=55209.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:27:46,837 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=55211.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:27:49,749 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 01:27:50,288 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=55218.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 01:27:53,352 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.410e+02 2.045e+02 2.381e+02 2.824e+02 4.578e+02, threshold=4.763e+02, percent-clipped=0.0 +2023-03-21 01:27:54,675 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.88 vs. limit=2.0 +2023-03-21 01:28:00,943 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8851, 2.5993, 1.8070, 2.8827, 2.9625, 2.5837, 2.2727, 2.2776], + device='cuda:0'), covar=tensor([0.1667, 0.0736, 0.3279, 0.0626, 0.0142, 0.0090, 0.0246, 0.0234], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0233, 0.0269, 0.0264, 0.0156, 0.0158, 0.0190, 0.0202], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:28:09,900 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5322, 2.6028, 2.3853, 2.5600, 2.6431, 2.4333, 2.4086, 2.3692], + device='cuda:0'), covar=tensor([0.0459, 0.0707, 0.1104, 0.0829, 0.0529, 0.0571, 0.0767, 0.1422], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0047, 0.0056, 0.0048, 0.0048, 0.0049, 0.0050, 0.0044], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:28:10,706 INFO [train.py:901] (0/2) Epoch 20, batch 1600, loss[loss=0.166, simple_loss=0.2355, pruned_loss=0.04826, over 7294.00 frames. ], tot_loss[loss=0.1523, simple_loss=0.2294, pruned_loss=0.03759, over 1439768.64 frames. ], batch size: 68, lr: 8.24e-03, grad_scale: 16.0 +2023-03-21 01:28:18,409 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=55272.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:28:21,828 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 01:28:22,346 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 01:28:25,338 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 01:28:33,208 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.66 vs. limit=5.0 +2023-03-21 01:28:34,458 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 01:28:35,912 INFO [train.py:901] (0/2) Epoch 20, batch 1650, loss[loss=0.1367, simple_loss=0.2129, pruned_loss=0.03028, over 7111.00 frames. ], tot_loss[loss=0.1519, simple_loss=0.2289, pruned_loss=0.03744, over 1441659.05 frames. ], batch size: 41, lr: 8.23e-03, grad_scale: 16.0 +2023-03-21 01:28:39,038 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 01:28:43,783 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=55320.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 01:28:45,232 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.261e+02 2.014e+02 2.351e+02 2.724e+02 5.713e+02, threshold=4.703e+02, percent-clipped=3.0 +2023-03-21 01:28:47,333 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 01:29:02,360 INFO [train.py:901] (0/2) Epoch 20, batch 1700, loss[loss=0.1365, simple_loss=0.2138, pruned_loss=0.02964, over 7282.00 frames. ], tot_loss[loss=0.1521, simple_loss=0.2294, pruned_loss=0.03737, over 1441700.35 frames. ], batch size: 66, lr: 8.23e-03, grad_scale: 16.0 +2023-03-21 01:29:03,912 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 01:29:07,895 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 01:29:07,941 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=55368.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 01:29:17,616 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 01:29:29,224 INFO [train.py:901] (0/2) Epoch 20, batch 1750, loss[loss=0.1626, simple_loss=0.2381, pruned_loss=0.04356, over 7277.00 frames. ], tot_loss[loss=0.1518, simple_loss=0.2293, pruned_loss=0.03713, over 1442920.74 frames. ], batch size: 57, lr: 8.23e-03, grad_scale: 16.0 +2023-03-21 01:29:35,810 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.62 vs. limit=2.0 +2023-03-21 01:29:37,408 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.350e+02 1.989e+02 2.275e+02 2.798e+02 4.613e+02, threshold=4.551e+02, percent-clipped=0.0 +2023-03-21 01:29:42,514 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 01:29:43,531 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 01:29:54,413 INFO [train.py:901] (0/2) Epoch 20, batch 1800, loss[loss=0.1379, simple_loss=0.2076, pruned_loss=0.03412, over 7122.00 frames. ], tot_loss[loss=0.1525, simple_loss=0.2301, pruned_loss=0.03746, over 1442842.22 frames. ], batch size: 41, lr: 8.22e-03, grad_scale: 16.0 +2023-03-21 01:30:04,279 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 01:30:19,151 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 01:30:19,210 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=55504.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:30:20,720 INFO [train.py:901] (0/2) Epoch 20, batch 1850, loss[loss=0.1355, simple_loss=0.2155, pruned_loss=0.02769, over 7320.00 frames. ], tot_loss[loss=0.1523, simple_loss=0.2296, pruned_loss=0.03748, over 1440950.97 frames. ], batch size: 49, lr: 8.22e-03, grad_scale: 16.0 +2023-03-21 01:30:26,495 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=55518.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 01:30:28,889 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.452e+02 1.968e+02 2.509e+02 2.876e+02 6.441e+02, threshold=5.018e+02, percent-clipped=4.0 +2023-03-21 01:30:29,051 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6345, 2.0399, 2.2407, 1.7499, 2.0967, 1.9054, 1.5706, 1.2978], + device='cuda:0'), covar=tensor([0.0345, 0.0309, 0.0076, 0.0235, 0.0301, 0.0255, 0.0305, 0.0298], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0026, 0.0024, 0.0025, 0.0026, 0.0024, 0.0027, 0.0028], + device='cuda:0'), out_proj_covar=tensor([6.9604e-05, 6.7493e-05, 6.0235e-05, 6.1553e-05, 6.5692e-05, 6.2941e-05, + 6.7912e-05, 7.0906e-05], device='cuda:0') +2023-03-21 01:30:29,407 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 01:30:42,023 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7975, 3.2864, 2.4745, 3.9200, 1.7999, 3.8163, 1.5693, 2.9790], + device='cuda:0'), covar=tensor([0.0120, 0.0629, 0.1686, 0.0091, 0.3903, 0.0173, 0.1132, 0.0328], + device='cuda:0'), in_proj_covar=tensor([0.0187, 0.0265, 0.0284, 0.0178, 0.0267, 0.0193, 0.0256, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:30:45,488 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 01:30:45,985 INFO [train.py:901] (0/2) Epoch 20, batch 1900, loss[loss=0.133, simple_loss=0.2102, pruned_loss=0.02783, over 7351.00 frames. ], tot_loss[loss=0.1529, simple_loss=0.2301, pruned_loss=0.03781, over 1440393.80 frames. ], batch size: 63, lr: 8.21e-03, grad_scale: 16.0 +2023-03-21 01:30:50,587 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=55566.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:30:51,101 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=55567.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:31:02,219 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9536, 2.8044, 2.1410, 3.7433, 2.4032, 2.8271, 1.6173, 2.2190], + device='cuda:0'), covar=tensor([0.0359, 0.0723, 0.2240, 0.0512, 0.0394, 0.0377, 0.3106, 0.1592], + device='cuda:0'), in_proj_covar=tensor([0.0266, 0.0249, 0.0296, 0.0254, 0.0269, 0.0259, 0.0260, 0.0276], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 01:31:11,848 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 01:31:12,340 INFO [train.py:901] (0/2) Epoch 20, batch 1950, loss[loss=0.1559, simple_loss=0.2298, pruned_loss=0.04106, over 7260.00 frames. ], tot_loss[loss=0.1525, simple_loss=0.2296, pruned_loss=0.03769, over 1439195.23 frames. ], batch size: 55, lr: 8.21e-03, grad_scale: 16.0 +2023-03-21 01:31:20,368 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.368e+02 2.150e+02 2.553e+02 3.108e+02 5.927e+02, threshold=5.105e+02, percent-clipped=4.0 +2023-03-21 01:31:20,696 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-21 01:31:21,459 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 01:31:25,917 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 01:31:26,419 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 01:31:38,085 INFO [train.py:901] (0/2) Epoch 20, batch 2000, loss[loss=0.158, simple_loss=0.2398, pruned_loss=0.03808, over 7238.00 frames. ], tot_loss[loss=0.1516, simple_loss=0.2286, pruned_loss=0.0373, over 1437401.57 frames. ], batch size: 55, lr: 8.21e-03, grad_scale: 16.0 +2023-03-21 01:31:44,809 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 01:31:54,871 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 01:31:56,045 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2207, 3.0179, 2.8396, 2.9483, 2.5860, 2.5047, 3.2263, 2.2110], + device='cuda:0'), covar=tensor([0.0512, 0.0515, 0.0531, 0.0546, 0.0732, 0.1050, 0.0727, 0.2007], + device='cuda:0'), in_proj_covar=tensor([0.0328, 0.0332, 0.0265, 0.0354, 0.0305, 0.0304, 0.0338, 0.0289], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:31:58,032 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1612, 1.3559, 1.1725, 1.3026, 1.3264, 1.2060, 1.1787, 0.9664], + device='cuda:0'), covar=tensor([0.0168, 0.0121, 0.0197, 0.0112, 0.0068, 0.0131, 0.0116, 0.0132], + device='cuda:0'), in_proj_covar=tensor([0.0025, 0.0025, 0.0024, 0.0024, 0.0023, 0.0024, 0.0025, 0.0032], + device='cuda:0'), out_proj_covar=tensor([2.9391e-05, 2.8115e-05, 2.7823e-05, 2.6834e-05, 2.6412e-05, 2.6427e-05, + 2.8577e-05, 3.6540e-05], device='cuda:0') +2023-03-21 01:32:03,494 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 01:32:03,970 INFO [train.py:901] (0/2) Epoch 20, batch 2050, loss[loss=0.1579, simple_loss=0.2289, pruned_loss=0.04346, over 7217.00 frames. ], tot_loss[loss=0.1515, simple_loss=0.2287, pruned_loss=0.03716, over 1440217.73 frames. ], batch size: 50, lr: 8.20e-03, grad_scale: 16.0 +2023-03-21 01:32:11,987 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.451e+02 1.903e+02 2.186e+02 2.766e+02 4.460e+02, threshold=4.371e+02, percent-clipped=0.0 +2023-03-21 01:32:26,043 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=55748.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:32:30,261 INFO [train.py:901] (0/2) Epoch 20, batch 2100, loss[loss=0.1525, simple_loss=0.2425, pruned_loss=0.03124, over 7262.00 frames. ], tot_loss[loss=0.1517, simple_loss=0.2291, pruned_loss=0.03717, over 1441505.44 frames. ], batch size: 47, lr: 8.20e-03, grad_scale: 16.0 +2023-03-21 01:32:30,399 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7809, 2.1358, 2.3327, 1.9730, 2.3971, 2.0955, 1.9055, 1.4540], + device='cuda:0'), covar=tensor([0.0388, 0.0472, 0.0236, 0.0189, 0.0402, 0.0240, 0.0215, 0.0413], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0026, 0.0024, 0.0025, 0.0025, 0.0024, 0.0027, 0.0028], + device='cuda:0'), out_proj_covar=tensor([6.9770e-05, 6.7623e-05, 6.0165e-05, 6.1572e-05, 6.5297e-05, 6.3179e-05, + 6.7080e-05, 7.0482e-05], device='cuda:0') +2023-03-21 01:32:36,784 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 01:32:39,750 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 01:32:54,200 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=55804.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:32:55,646 INFO [train.py:901] (0/2) Epoch 20, batch 2150, loss[loss=0.15, simple_loss=0.2339, pruned_loss=0.03306, over 7288.00 frames. ], tot_loss[loss=0.151, simple_loss=0.2285, pruned_loss=0.03677, over 1441032.76 frames. ], batch size: 68, lr: 8.20e-03, grad_scale: 16.0 +2023-03-21 01:32:56,802 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=55809.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:33:02,301 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.3070, 4.7492, 4.7175, 5.3243, 5.1388, 5.2007, 4.6807, 4.7657], + device='cuda:0'), covar=tensor([0.0638, 0.2300, 0.2328, 0.0882, 0.0886, 0.1209, 0.0713, 0.0892], + device='cuda:0'), in_proj_covar=tensor([0.0213, 0.0330, 0.0262, 0.0264, 0.0192, 0.0318, 0.0190, 0.0231], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:33:03,698 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.263e+02 2.025e+02 2.413e+02 3.099e+02 7.099e+02, threshold=4.826e+02, percent-clipped=3.0 +2023-03-21 01:33:17,513 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=55848.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:33:19,356 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=55852.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:33:21,764 INFO [train.py:901] (0/2) Epoch 20, batch 2200, loss[loss=0.1471, simple_loss=0.2253, pruned_loss=0.03444, over 7291.00 frames. ], tot_loss[loss=0.1512, simple_loss=0.2287, pruned_loss=0.03689, over 1440550.38 frames. ], batch size: 70, lr: 8.19e-03, grad_scale: 8.0 +2023-03-21 01:33:24,157 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.25 vs. limit=5.0 +2023-03-21 01:33:25,294 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 01:33:26,875 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=55867.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:33:42,635 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2098, 1.0952, 1.5413, 1.6340, 1.4779, 1.6017, 1.2725, 1.5187], + device='cuda:0'), covar=tensor([0.1787, 0.3861, 0.1579, 0.2007, 0.1807, 0.2637, 0.1349, 0.2663], + device='cuda:0'), in_proj_covar=tensor([0.0052, 0.0061, 0.0045, 0.0043, 0.0045, 0.0047, 0.0066, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 01:33:47,069 INFO [train.py:901] (0/2) Epoch 20, batch 2250, loss[loss=0.1074, simple_loss=0.1685, pruned_loss=0.02318, over 5978.00 frames. ], tot_loss[loss=0.1511, simple_loss=0.2286, pruned_loss=0.03681, over 1439180.29 frames. ], batch size: 25, lr: 8.19e-03, grad_scale: 8.0 +2023-03-21 01:33:48,179 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=55909.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:33:51,812 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=55915.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:33:56,854 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.213e+02 2.068e+02 2.360e+02 2.879e+02 4.084e+02, threshold=4.720e+02, percent-clipped=0.0 +2023-03-21 01:33:59,966 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 01:34:00,442 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 01:34:12,291 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 01:34:13,749 INFO [train.py:901] (0/2) Epoch 20, batch 2300, loss[loss=0.1729, simple_loss=0.2475, pruned_loss=0.04911, over 7290.00 frames. ], tot_loss[loss=0.1509, simple_loss=0.2285, pruned_loss=0.03662, over 1441568.94 frames. ], batch size: 66, lr: 8.19e-03, grad_scale: 8.0 +2023-03-21 01:34:22,210 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.36 vs. limit=5.0 +2023-03-21 01:34:35,131 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.89 vs. limit=2.0 +2023-03-21 01:34:36,611 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-56000.pt +2023-03-21 01:34:44,527 INFO [train.py:901] (0/2) Epoch 20, batch 2350, loss[loss=0.1507, simple_loss=0.2301, pruned_loss=0.03565, over 7216.00 frames. ], tot_loss[loss=0.1511, simple_loss=0.2289, pruned_loss=0.03663, over 1440237.61 frames. ], batch size: 45, lr: 8.18e-03, grad_scale: 8.0 +2023-03-21 01:34:53,106 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.335e+02 2.001e+02 2.385e+02 2.972e+02 5.246e+02, threshold=4.769e+02, percent-clipped=3.0 +2023-03-21 01:34:53,245 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7103, 4.5807, 4.4269, 3.9429, 4.1460, 2.9736, 2.3466, 4.6967], + device='cuda:0'), covar=tensor([0.0027, 0.0051, 0.0042, 0.0046, 0.0052, 0.0339, 0.0449, 0.0031], + device='cuda:0'), in_proj_covar=tensor([0.0090, 0.0076, 0.0096, 0.0082, 0.0106, 0.0119, 0.0120, 0.0088], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 01:35:03,217 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 01:35:09,365 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 01:35:09,843 INFO [train.py:901] (0/2) Epoch 20, batch 2400, loss[loss=0.1543, simple_loss=0.2346, pruned_loss=0.03701, over 6688.00 frames. ], tot_loss[loss=0.1516, simple_loss=0.2293, pruned_loss=0.03691, over 1439322.83 frames. ], batch size: 107, lr: 8.18e-03, grad_scale: 8.0 +2023-03-21 01:35:20,532 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 01:35:23,679 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 01:35:34,741 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=56104.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:35:36,186 INFO [train.py:901] (0/2) Epoch 20, batch 2450, loss[loss=0.1698, simple_loss=0.2491, pruned_loss=0.04522, over 7143.00 frames. ], tot_loss[loss=0.1512, simple_loss=0.2289, pruned_loss=0.03671, over 1441571.27 frames. ], batch size: 98, lr: 8.17e-03, grad_scale: 8.0 +2023-03-21 01:35:44,668 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.431e+02 2.178e+02 2.434e+02 2.805e+02 5.134e+02, threshold=4.868e+02, percent-clipped=1.0 +2023-03-21 01:35:47,427 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.85 vs. limit=2.0 +2023-03-21 01:35:49,733 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 01:35:55,286 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=56145.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:35:59,351 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9662, 2.8163, 1.9374, 3.0327, 2.7760, 3.0333, 2.2247, 2.5543], + device='cuda:0'), covar=tensor([0.1877, 0.0689, 0.2933, 0.0517, 0.0094, 0.0092, 0.0200, 0.0220], + device='cuda:0'), in_proj_covar=tensor([0.0265, 0.0226, 0.0260, 0.0253, 0.0152, 0.0154, 0.0189, 0.0201], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:36:01,182 INFO [train.py:901] (0/2) Epoch 20, batch 2500, loss[loss=0.1354, simple_loss=0.2081, pruned_loss=0.03135, over 7341.00 frames. ], tot_loss[loss=0.1518, simple_loss=0.2294, pruned_loss=0.03711, over 1442700.26 frames. ], batch size: 44, lr: 8.17e-03, grad_scale: 8.0 +2023-03-21 01:36:15,819 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 01:36:26,478 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=56204.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:36:27,572 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=56206.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:36:27,970 INFO [train.py:901] (0/2) Epoch 20, batch 2550, loss[loss=0.1289, simple_loss=0.2066, pruned_loss=0.0256, over 7146.00 frames. ], tot_loss[loss=0.1512, simple_loss=0.2287, pruned_loss=0.03688, over 1439364.59 frames. ], batch size: 41, lr: 8.17e-03, grad_scale: 8.0 +2023-03-21 01:36:33,038 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1859, 4.6810, 4.7394, 4.6739, 4.6443, 4.1939, 4.7338, 4.6121], + device='cuda:0'), covar=tensor([0.0482, 0.0419, 0.0402, 0.0506, 0.0316, 0.0385, 0.0414, 0.0499], + device='cuda:0'), in_proj_covar=tensor([0.0190, 0.0220, 0.0164, 0.0163, 0.0130, 0.0199, 0.0173, 0.0130], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:36:35,058 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0243, 4.5315, 4.5684, 4.4848, 4.5309, 4.0781, 4.5688, 4.4780], + device='cuda:0'), covar=tensor([0.0517, 0.0439, 0.0437, 0.0567, 0.0301, 0.0407, 0.0423, 0.0499], + device='cuda:0'), in_proj_covar=tensor([0.0190, 0.0220, 0.0164, 0.0163, 0.0130, 0.0199, 0.0174, 0.0130], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:36:36,412 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.460e+02 2.002e+02 2.403e+02 2.852e+02 5.182e+02, threshold=4.807e+02, percent-clipped=1.0 +2023-03-21 01:36:53,747 INFO [train.py:901] (0/2) Epoch 20, batch 2600, loss[loss=0.1542, simple_loss=0.2327, pruned_loss=0.0379, over 7339.00 frames. ], tot_loss[loss=0.1509, simple_loss=0.2281, pruned_loss=0.03683, over 1439961.07 frames. ], batch size: 54, lr: 8.16e-03, grad_scale: 8.0 +2023-03-21 01:37:18,980 INFO [train.py:901] (0/2) Epoch 20, batch 2650, loss[loss=0.1354, simple_loss=0.2141, pruned_loss=0.02833, over 7285.00 frames. ], tot_loss[loss=0.1503, simple_loss=0.2274, pruned_loss=0.03658, over 1438784.31 frames. ], batch size: 70, lr: 8.16e-03, grad_scale: 8.0 +2023-03-21 01:37:27,399 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.210e+02 2.152e+02 2.521e+02 3.004e+02 6.613e+02, threshold=5.042e+02, percent-clipped=2.0 +2023-03-21 01:37:33,815 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.84 vs. limit=2.0 +2023-03-21 01:37:34,998 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8476, 2.9968, 2.6443, 2.9497, 2.9702, 2.6634, 3.0687, 2.7045], + device='cuda:0'), covar=tensor([0.0554, 0.0919, 0.1057, 0.1144, 0.1266, 0.0618, 0.1120, 0.1354], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0047, 0.0054, 0.0048, 0.0048, 0.0048, 0.0049, 0.0044], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:37:43,276 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8913, 5.4156, 5.4751, 5.4670, 5.2356, 4.9746, 5.5331, 5.3591], + device='cuda:0'), covar=tensor([0.0461, 0.0407, 0.0395, 0.0448, 0.0295, 0.0318, 0.0320, 0.0421], + device='cuda:0'), in_proj_covar=tensor([0.0191, 0.0220, 0.0165, 0.0163, 0.0132, 0.0199, 0.0174, 0.0130], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:37:43,701 INFO [train.py:901] (0/2) Epoch 20, batch 2700, loss[loss=0.1292, simple_loss=0.202, pruned_loss=0.02823, over 6996.00 frames. ], tot_loss[loss=0.1511, simple_loss=0.2284, pruned_loss=0.03687, over 1441179.27 frames. ], batch size: 35, lr: 8.16e-03, grad_scale: 8.0 +2023-03-21 01:37:55,057 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4365, 3.1620, 3.2694, 3.2329, 2.7471, 2.7153, 3.4909, 2.4905], + device='cuda:0'), covar=tensor([0.0356, 0.0301, 0.0356, 0.0371, 0.0492, 0.0607, 0.0398, 0.1207], + device='cuda:0'), in_proj_covar=tensor([0.0324, 0.0327, 0.0263, 0.0349, 0.0303, 0.0301, 0.0330, 0.0287], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:38:02,201 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5352, 4.9122, 4.7639, 4.8861, 4.7080, 4.5313, 4.9839, 4.6764], + device='cuda:0'), covar=tensor([0.0811, 0.0956, 0.1052, 0.1144, 0.0627, 0.0687, 0.0842, 0.0907], + device='cuda:0'), in_proj_covar=tensor([0.0191, 0.0220, 0.0165, 0.0163, 0.0132, 0.0199, 0.0174, 0.0129], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:38:07,002 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=56404.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:38:08,416 INFO [train.py:901] (0/2) Epoch 20, batch 2750, loss[loss=0.1378, simple_loss=0.2123, pruned_loss=0.03165, over 7178.00 frames. ], tot_loss[loss=0.1513, simple_loss=0.2286, pruned_loss=0.03696, over 1442266.80 frames. ], batch size: 39, lr: 8.15e-03, grad_scale: 8.0 +2023-03-21 01:38:16,487 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.333e+02 2.245e+02 2.531e+02 3.013e+02 6.042e+02, threshold=5.062e+02, percent-clipped=2.0 +2023-03-21 01:38:30,014 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=56452.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:38:31,025 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8803, 2.7348, 3.0071, 3.0719, 3.0905, 2.7459, 2.4013, 3.3446], + device='cuda:0'), covar=tensor([0.2094, 0.0877, 0.1457, 0.1616, 0.1315, 0.1629, 0.3377, 0.1073], + device='cuda:0'), in_proj_covar=tensor([0.0046, 0.0055, 0.0042, 0.0042, 0.0042, 0.0038, 0.0056, 0.0044], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:38:31,520 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9996, 2.7160, 2.8844, 3.0808, 3.1527, 2.8889, 2.5966, 3.3903], + device='cuda:0'), covar=tensor([0.2267, 0.0796, 0.1597, 0.1805, 0.1216, 0.1024, 0.2503, 0.1112], + device='cuda:0'), in_proj_covar=tensor([0.0046, 0.0055, 0.0042, 0.0042, 0.0042, 0.0038, 0.0056, 0.0044], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:38:32,367 INFO [train.py:901] (0/2) Epoch 20, batch 2800, loss[loss=0.1418, simple_loss=0.2283, pruned_loss=0.02767, over 7293.00 frames. ], tot_loss[loss=0.1516, simple_loss=0.229, pruned_loss=0.0371, over 1442272.66 frames. ], batch size: 66, lr: 8.15e-03, grad_scale: 8.0 +2023-03-21 01:38:44,925 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-20.pt +2023-03-21 01:38:59,702 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 01:38:59,761 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 01:38:59,786 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 01:38:59,806 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 01:38:59,995 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 01:39:00,011 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 01:39:00,362 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 01:39:00,468 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 01:39:00,551 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 01:39:00,567 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 01:39:00,585 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 01:39:00,760 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 01:39:00,766 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 01:39:00,774 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 01:39:00,928 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 01:39:00,931 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 01:39:00,994 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 01:39:03,152 INFO [train.py:901] (0/2) Epoch 21, batch 0, loss[loss=0.118, simple_loss=0.1859, pruned_loss=0.02505, over 7038.00 frames. ], tot_loss[loss=0.118, simple_loss=0.1859, pruned_loss=0.02505, over 7038.00 frames. ], batch size: 35, lr: 7.96e-03, grad_scale: 8.0 +2023-03-21 01:39:03,154 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 01:39:28,649 INFO [train.py:935] (0/2) Epoch 21, validation: loss=0.1655, simple_loss=0.2529, pruned_loss=0.03902, over 1622729.00 frames. +2023-03-21 01:39:28,650 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 01:39:35,687 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 01:39:39,363 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=56501.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:39:40,911 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=56504.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:39:45,814 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 01:39:51,143 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=56523.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 01:39:51,458 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.209e+02 1.941e+02 2.187e+02 2.550e+02 5.143e+02, threshold=4.375e+02, percent-clipped=1.0 +2023-03-21 01:39:54,006 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 01:39:55,008 INFO [train.py:901] (0/2) Epoch 21, batch 50, loss[loss=0.1086, simple_loss=0.1767, pruned_loss=0.02025, over 6096.00 frames. ], tot_loss[loss=0.1533, simple_loss=0.2305, pruned_loss=0.03804, over 324750.81 frames. ], batch size: 26, lr: 7.96e-03, grad_scale: 8.0 +2023-03-21 01:39:56,549 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 01:39:59,175 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 01:40:05,805 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=56552.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:40:13,085 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.46 vs. limit=2.0 +2023-03-21 01:40:15,293 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 01:40:15,792 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 01:40:16,929 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9457, 2.5194, 1.8478, 2.7061, 2.7385, 3.1863, 2.2624, 2.5589], + device='cuda:0'), covar=tensor([0.1799, 0.0772, 0.3015, 0.0738, 0.0150, 0.0170, 0.0242, 0.0263], + device='cuda:0'), in_proj_covar=tensor([0.0269, 0.0233, 0.0268, 0.0259, 0.0156, 0.0158, 0.0194, 0.0204], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:40:20,323 INFO [train.py:901] (0/2) Epoch 21, batch 100, loss[loss=0.1636, simple_loss=0.2421, pruned_loss=0.04257, over 7220.00 frames. ], tot_loss[loss=0.15, simple_loss=0.2273, pruned_loss=0.03638, over 571547.02 frames. ], batch size: 93, lr: 7.95e-03, grad_scale: 8.0 +2023-03-21 01:40:21,961 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=56584.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 01:40:39,587 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8327, 4.4273, 4.2791, 4.8904, 4.7189, 4.8534, 4.1465, 4.3871], + device='cuda:0'), covar=tensor([0.0898, 0.2334, 0.2450, 0.1100, 0.0913, 0.1149, 0.0805, 0.0968], + device='cuda:0'), in_proj_covar=tensor([0.0214, 0.0330, 0.0263, 0.0261, 0.0193, 0.0323, 0.0192, 0.0232], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:40:43,549 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.343e+02 2.050e+02 2.363e+02 2.734e+02 5.404e+02, threshold=4.726e+02, percent-clipped=1.0 +2023-03-21 01:40:47,095 INFO [train.py:901] (0/2) Epoch 21, batch 150, loss[loss=0.1498, simple_loss=0.2312, pruned_loss=0.03424, over 7269.00 frames. ], tot_loss[loss=0.1494, simple_loss=0.2267, pruned_loss=0.03609, over 764581.62 frames. ], batch size: 70, lr: 7.95e-03, grad_scale: 8.0 +2023-03-21 01:41:12,634 INFO [train.py:901] (0/2) Epoch 21, batch 200, loss[loss=0.1376, simple_loss=0.2312, pruned_loss=0.02198, over 7347.00 frames. ], tot_loss[loss=0.1492, simple_loss=0.2264, pruned_loss=0.03596, over 914027.06 frames. ], batch size: 54, lr: 7.95e-03, grad_scale: 8.0 +2023-03-21 01:41:16,121 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 01:41:21,742 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 01:41:27,714 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 01:41:34,704 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.304e+02 1.965e+02 2.278e+02 2.575e+02 4.496e+02, threshold=4.557e+02, percent-clipped=0.0 +2023-03-21 01:41:38,258 INFO [train.py:901] (0/2) Epoch 21, batch 250, loss[loss=0.1528, simple_loss=0.2344, pruned_loss=0.03559, over 7285.00 frames. ], tot_loss[loss=0.149, simple_loss=0.2263, pruned_loss=0.03586, over 1030575.81 frames. ], batch size: 70, lr: 7.94e-03, grad_scale: 8.0 +2023-03-21 01:41:40,758 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 01:41:45,812 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5049, 1.8841, 2.0872, 1.8312, 2.0022, 2.0712, 1.6581, 1.4489], + device='cuda:0'), covar=tensor([0.0547, 0.0402, 0.0311, 0.0296, 0.0393, 0.0234, 0.0302, 0.0299], + device='cuda:0'), in_proj_covar=tensor([0.0029, 0.0028, 0.0026, 0.0027, 0.0028, 0.0025, 0.0029, 0.0029], + device='cuda:0'), out_proj_covar=tensor([7.4027e-05, 7.2339e-05, 6.5480e-05, 6.6553e-05, 7.0204e-05, 6.6510e-05, + 7.1868e-05, 7.4388e-05], device='cuda:0') +2023-03-21 01:42:01,811 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 01:42:03,827 INFO [train.py:901] (0/2) Epoch 21, batch 300, loss[loss=0.1318, simple_loss=0.2124, pruned_loss=0.02561, over 7209.00 frames. ], tot_loss[loss=0.1501, simple_loss=0.2274, pruned_loss=0.03643, over 1121141.36 frames. ], batch size: 39, lr: 7.94e-03, grad_scale: 8.0 +2023-03-21 01:42:10,998 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 01:42:14,915 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=56801.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:42:15,950 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=56803.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:42:26,232 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.478e+02 1.973e+02 2.320e+02 2.778e+02 6.179e+02, threshold=4.640e+02, percent-clipped=2.0 +2023-03-21 01:42:26,578 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.92 vs. limit=5.0 +2023-03-21 01:42:29,718 INFO [train.py:901] (0/2) Epoch 21, batch 350, loss[loss=0.1463, simple_loss=0.2305, pruned_loss=0.03103, over 7318.00 frames. ], tot_loss[loss=0.1507, simple_loss=0.228, pruned_loss=0.03668, over 1191255.58 frames. ], batch size: 59, lr: 7.93e-03, grad_scale: 8.0 +2023-03-21 01:42:39,444 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=56849.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:42:44,901 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 01:42:47,023 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=56864.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:42:55,136 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=56879.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 01:42:56,057 INFO [train.py:901] (0/2) Epoch 21, batch 400, loss[loss=0.149, simple_loss=0.2331, pruned_loss=0.03246, over 7300.00 frames. ], tot_loss[loss=0.1515, simple_loss=0.229, pruned_loss=0.03699, over 1248177.97 frames. ], batch size: 77, lr: 7.93e-03, grad_scale: 8.0 +2023-03-21 01:42:57,154 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8695, 5.3811, 5.4212, 5.3715, 5.0263, 4.9125, 5.4790, 5.1472], + device='cuda:0'), covar=tensor([0.0409, 0.0363, 0.0330, 0.0440, 0.0369, 0.0342, 0.0280, 0.0583], + device='cuda:0'), in_proj_covar=tensor([0.0189, 0.0218, 0.0165, 0.0161, 0.0131, 0.0198, 0.0171, 0.0129], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:43:09,786 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=56908.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:43:17,668 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.336e+02 2.058e+02 2.428e+02 2.833e+02 5.139e+02, threshold=4.856e+02, percent-clipped=2.0 +2023-03-21 01:43:21,707 INFO [train.py:901] (0/2) Epoch 21, batch 450, loss[loss=0.1501, simple_loss=0.2319, pruned_loss=0.03419, over 7244.00 frames. ], tot_loss[loss=0.1509, simple_loss=0.2285, pruned_loss=0.03663, over 1291037.84 frames. ], batch size: 89, lr: 7.93e-03, grad_scale: 8.0 +2023-03-21 01:43:22,841 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4663, 1.0123, 1.8837, 1.9938, 1.6642, 1.9950, 1.6193, 1.8222], + device='cuda:0'), covar=tensor([0.2519, 0.5578, 0.0795, 0.1649, 0.6232, 0.3782, 0.2146, 0.1344], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0063, 0.0047, 0.0044, 0.0047, 0.0050, 0.0069, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 01:43:25,246 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 01:43:25,776 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 01:43:41,399 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=56969.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:43:42,121 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.91 vs. limit=5.0 +2023-03-21 01:43:47,251 INFO [train.py:901] (0/2) Epoch 21, batch 500, loss[loss=0.1368, simple_loss=0.2148, pruned_loss=0.02942, over 7256.00 frames. ], tot_loss[loss=0.1504, simple_loss=0.228, pruned_loss=0.03637, over 1325249.64 frames. ], batch size: 47, lr: 7.92e-03, grad_scale: 8.0 +2023-03-21 01:43:53,324 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=56993.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:43:54,347 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=56995.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:43:57,518 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 01:43:58,999 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 01:43:59,462 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 01:44:01,992 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 01:44:06,377 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 01:44:09,359 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.406e+02 1.969e+02 2.416e+02 3.009e+02 5.673e+02, threshold=4.831e+02, percent-clipped=3.0 +2023-03-21 01:44:12,848 INFO [train.py:901] (0/2) Epoch 21, batch 550, loss[loss=0.1532, simple_loss=0.2313, pruned_loss=0.03752, over 7325.00 frames. ], tot_loss[loss=0.1509, simple_loss=0.2287, pruned_loss=0.03656, over 1352347.12 frames. ], batch size: 59, lr: 7.92e-03, grad_scale: 8.0 +2023-03-21 01:44:18,423 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 01:44:25,088 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=57054.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:44:26,091 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=57056.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:44:26,956 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 01:44:29,891 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 01:44:37,319 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 01:44:38,303 INFO [train.py:901] (0/2) Epoch 21, batch 600, loss[loss=0.1683, simple_loss=0.2443, pruned_loss=0.04617, over 7287.00 frames. ], tot_loss[loss=0.1503, simple_loss=0.2279, pruned_loss=0.03632, over 1373593.31 frames. ], batch size: 70, lr: 7.92e-03, grad_scale: 8.0 +2023-03-21 01:44:55,081 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 01:45:00,649 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.412e+02 2.016e+02 2.506e+02 2.951e+02 4.891e+02, threshold=5.012e+02, percent-clipped=1.0 +2023-03-21 01:45:04,270 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 01:45:04,764 INFO [train.py:901] (0/2) Epoch 21, batch 650, loss[loss=0.1564, simple_loss=0.2311, pruned_loss=0.04083, over 7303.00 frames. ], tot_loss[loss=0.15, simple_loss=0.2277, pruned_loss=0.03616, over 1389357.03 frames. ], batch size: 83, lr: 7.91e-03, grad_scale: 8.0 +2023-03-21 01:45:07,942 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0366, 2.7134, 3.0889, 2.9999, 3.2431, 2.8970, 2.5327, 3.1582], + device='cuda:0'), covar=tensor([0.1599, 0.0883, 0.1288, 0.1165, 0.0595, 0.0977, 0.1988, 0.1396], + device='cuda:0'), in_proj_covar=tensor([0.0045, 0.0054, 0.0041, 0.0042, 0.0040, 0.0037, 0.0056, 0.0043], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:45:18,857 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=57159.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:45:20,721 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 01:45:28,916 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=57179.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 01:45:29,325 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 01:45:29,797 INFO [train.py:901] (0/2) Epoch 21, batch 700, loss[loss=0.1575, simple_loss=0.2406, pruned_loss=0.03726, over 7347.00 frames. ], tot_loss[loss=0.1507, simple_loss=0.2283, pruned_loss=0.0366, over 1400883.99 frames. ], batch size: 73, lr: 7.91e-03, grad_scale: 8.0 +2023-03-21 01:45:33,066 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=57186.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:45:38,034 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7409, 3.9174, 3.7414, 3.9526, 3.5210, 3.8712, 4.1597, 4.1683], + device='cuda:0'), covar=tensor([0.0246, 0.0174, 0.0206, 0.0176, 0.0403, 0.0303, 0.0231, 0.0209], + device='cuda:0'), in_proj_covar=tensor([0.0112, 0.0113, 0.0104, 0.0111, 0.0103, 0.0094, 0.0092, 0.0088], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:45:38,511 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=57197.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:45:47,445 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.80 vs. limit=2.0 +2023-03-21 01:45:53,099 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.286e+02 1.965e+02 2.359e+02 2.948e+02 5.669e+02, threshold=4.718e+02, percent-clipped=2.0 +2023-03-21 01:45:54,636 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 01:45:54,675 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=57227.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 01:45:55,579 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 01:45:56,528 INFO [train.py:901] (0/2) Epoch 21, batch 750, loss[loss=0.1813, simple_loss=0.2474, pruned_loss=0.05763, over 7340.00 frames. ], tot_loss[loss=0.151, simple_loss=0.2286, pruned_loss=0.03672, over 1409043.19 frames. ], batch size: 73, lr: 7.91e-03, grad_scale: 8.0 +2023-03-21 01:45:57,319 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.75 vs. limit=2.0 +2023-03-21 01:46:02,443 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.00 vs. limit=2.0 +2023-03-21 01:46:03,217 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=57244.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:46:04,694 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=57247.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:46:09,634 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 01:46:10,338 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=57258.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 01:46:13,278 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=57264.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:46:13,724 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 01:46:20,816 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 01:46:21,806 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 01:46:22,282 INFO [train.py:901] (0/2) Epoch 21, batch 800, loss[loss=0.1499, simple_loss=0.2289, pruned_loss=0.0355, over 7349.00 frames. ], tot_loss[loss=0.1504, simple_loss=0.228, pruned_loss=0.03642, over 1416931.79 frames. ], batch size: 63, lr: 7.90e-03, grad_scale: 8.0 +2023-03-21 01:46:32,296 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 01:46:34,937 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=57305.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:46:42,812 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1240, 4.2254, 3.9569, 4.1967, 3.7739, 4.2084, 4.5486, 4.5184], + device='cuda:0'), covar=tensor([0.0205, 0.0157, 0.0228, 0.0196, 0.0386, 0.0245, 0.0221, 0.0193], + device='cuda:0'), in_proj_covar=tensor([0.0114, 0.0115, 0.0106, 0.0115, 0.0105, 0.0096, 0.0094, 0.0090], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:46:44,210 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.475e+02 2.042e+02 2.374e+02 2.907e+02 5.577e+02, threshold=4.749e+02, percent-clipped=3.0 +2023-03-21 01:46:45,176 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.66 vs. limit=2.0 +2023-03-21 01:46:47,728 INFO [train.py:901] (0/2) Epoch 21, batch 850, loss[loss=0.1411, simple_loss=0.2182, pruned_loss=0.03201, over 7280.00 frames. ], tot_loss[loss=0.1498, simple_loss=0.2273, pruned_loss=0.03618, over 1421737.77 frames. ], batch size: 57, lr: 7.90e-03, grad_scale: 8.0 +2023-03-21 01:46:50,766 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 01:46:50,770 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 01:46:51,426 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4887, 3.4021, 2.5120, 3.9945, 2.8008, 3.2616, 1.9890, 2.2902], + device='cuda:0'), covar=tensor([0.0407, 0.0731, 0.2133, 0.0389, 0.0298, 0.0922, 0.3009, 0.1768], + device='cuda:0'), in_proj_covar=tensor([0.0265, 0.0250, 0.0292, 0.0251, 0.0262, 0.0256, 0.0253, 0.0273], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:46:54,141 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.94 vs. limit=2.0 +2023-03-21 01:46:56,303 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 01:46:56,876 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=57349.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:46:57,869 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=57351.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:46:59,787 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 01:47:13,392 INFO [train.py:901] (0/2) Epoch 21, batch 900, loss[loss=0.1503, simple_loss=0.2181, pruned_loss=0.04126, over 7215.00 frames. ], tot_loss[loss=0.1498, simple_loss=0.2273, pruned_loss=0.03615, over 1428506.98 frames. ], batch size: 39, lr: 7.90e-03, grad_scale: 8.0 +2023-03-21 01:47:22,715 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=57398.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:47:25,022 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8890, 3.4201, 3.8552, 3.7237, 3.9775, 3.8912, 3.8788, 3.6622], + device='cuda:0'), covar=tensor([0.0027, 0.0084, 0.0032, 0.0038, 0.0026, 0.0033, 0.0035, 0.0053], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0055, 0.0048, 0.0046, 0.0045, 0.0050, 0.0044, 0.0060], + device='cuda:0'), out_proj_covar=tensor([7.7171e-05, 1.3074e-04, 1.0615e-04, 9.5376e-05, 9.1835e-05, 1.0351e-04, + 1.0064e-04, 1.2819e-04], device='cuda:0') +2023-03-21 01:47:35,957 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.448e+02 2.097e+02 2.421e+02 2.780e+02 5.960e+02, threshold=4.843e+02, percent-clipped=2.0 +2023-03-21 01:47:39,042 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 01:47:39,538 INFO [train.py:901] (0/2) Epoch 21, batch 950, loss[loss=0.1455, simple_loss=0.2211, pruned_loss=0.03493, over 7244.00 frames. ], tot_loss[loss=0.1498, simple_loss=0.2275, pruned_loss=0.03602, over 1432116.41 frames. ], batch size: 45, lr: 7.89e-03, grad_scale: 8.0 +2023-03-21 01:47:54,294 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=57459.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:47:54,330 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=57459.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:47:55,819 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=57462.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:48:03,875 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 01:48:05,876 INFO [train.py:901] (0/2) Epoch 21, batch 1000, loss[loss=0.1306, simple_loss=0.2104, pruned_loss=0.0254, over 7159.00 frames. ], tot_loss[loss=0.1497, simple_loss=0.2277, pruned_loss=0.03589, over 1433793.52 frames. ], batch size: 41, lr: 7.89e-03, grad_scale: 8.0 +2023-03-21 01:48:12,779 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9241, 4.3870, 4.3161, 4.8903, 4.7702, 4.8258, 4.1876, 4.3491], + device='cuda:0'), covar=tensor([0.0836, 0.2542, 0.2415, 0.1066, 0.0868, 0.1270, 0.0850, 0.1082], + device='cuda:0'), in_proj_covar=tensor([0.0220, 0.0338, 0.0270, 0.0266, 0.0194, 0.0330, 0.0194, 0.0238], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:48:12,861 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=57495.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:48:18,861 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=57507.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:48:23,882 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 01:48:26,041 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2943, 2.3076, 2.1763, 3.5176, 1.7396, 3.2748, 1.3958, 3.0156], + device='cuda:0'), covar=tensor([0.0107, 0.1097, 0.1542, 0.0116, 0.3425, 0.0177, 0.1144, 0.0454], + device='cuda:0'), in_proj_covar=tensor([0.0191, 0.0266, 0.0284, 0.0182, 0.0268, 0.0196, 0.0254, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:48:27,051 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=57523.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:48:27,398 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.303e+02 1.945e+02 2.283e+02 2.747e+02 4.323e+02, threshold=4.566e+02, percent-clipped=0.0 +2023-03-21 01:48:30,922 INFO [train.py:901] (0/2) Epoch 21, batch 1050, loss[loss=0.1239, simple_loss=0.2093, pruned_loss=0.01924, over 7327.00 frames. ], tot_loss[loss=0.1493, simple_loss=0.2272, pruned_loss=0.03574, over 1435033.09 frames. ], batch size: 75, lr: 7.89e-03, grad_scale: 8.0 +2023-03-21 01:48:33,527 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3937, 3.0395, 3.3813, 2.8950, 2.8190, 2.6632, 3.1799, 2.5995], + device='cuda:0'), covar=tensor([0.0267, 0.0337, 0.0429, 0.0421, 0.0427, 0.0651, 0.0350, 0.1236], + device='cuda:0'), in_proj_covar=tensor([0.0319, 0.0330, 0.0265, 0.0347, 0.0305, 0.0300, 0.0332, 0.0283], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:48:36,944 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=57542.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:48:42,428 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=57553.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 01:48:44,018 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=57556.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:48:44,405 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 01:48:48,712 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=57564.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:48:49,606 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 01:48:56,971 INFO [train.py:901] (0/2) Epoch 21, batch 1100, loss[loss=0.177, simple_loss=0.2526, pruned_loss=0.05072, over 7120.00 frames. ], tot_loss[loss=0.1497, simple_loss=0.2274, pruned_loss=0.03599, over 1437524.16 frames. ], batch size: 98, lr: 7.88e-03, grad_scale: 8.0 +2023-03-21 01:49:04,313 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0070, 2.5007, 1.7110, 2.7221, 2.7502, 3.0487, 2.5651, 2.3556], + device='cuda:0'), covar=tensor([0.1986, 0.0674, 0.3337, 0.0525, 0.0160, 0.0141, 0.0306, 0.0243], + device='cuda:0'), in_proj_covar=tensor([0.0270, 0.0230, 0.0266, 0.0260, 0.0160, 0.0159, 0.0190, 0.0204], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:49:07,087 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=57600.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:49:13,263 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=57612.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:49:18,839 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 01:49:19,886 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.249e+02 1.979e+02 2.394e+02 2.879e+02 9.847e+02, threshold=4.789e+02, percent-clipped=4.0 +2023-03-21 01:49:19,937 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 01:49:23,553 INFO [train.py:901] (0/2) Epoch 21, batch 1150, loss[loss=0.1553, simple_loss=0.2386, pruned_loss=0.03599, over 7348.00 frames. ], tot_loss[loss=0.1482, simple_loss=0.2261, pruned_loss=0.03514, over 1436492.47 frames. ], batch size: 54, lr: 7.88e-03, grad_scale: 8.0 +2023-03-21 01:49:29,187 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9533, 4.1920, 3.9928, 4.2134, 3.8680, 4.1662, 4.5442, 4.4967], + device='cuda:0'), covar=tensor([0.0225, 0.0139, 0.0192, 0.0164, 0.0284, 0.0221, 0.0182, 0.0171], + device='cuda:0'), in_proj_covar=tensor([0.0110, 0.0109, 0.0101, 0.0108, 0.0100, 0.0092, 0.0089, 0.0086], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:49:32,114 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 01:49:32,682 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=57649.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:49:33,078 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 01:49:33,633 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=57651.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:49:35,018 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.52 vs. limit=5.0 +2023-03-21 01:49:49,211 INFO [train.py:901] (0/2) Epoch 21, batch 1200, loss[loss=0.1581, simple_loss=0.2354, pruned_loss=0.04042, over 7304.00 frames. ], tot_loss[loss=0.1492, simple_loss=0.2273, pruned_loss=0.03551, over 1437011.32 frames. ], batch size: 68, lr: 7.88e-03, grad_scale: 8.0 +2023-03-21 01:49:57,357 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=57697.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:49:58,369 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=57699.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:50:03,938 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.1482, 4.6822, 4.6554, 5.1323, 5.0417, 5.0655, 4.3876, 4.5961], + device='cuda:0'), covar=tensor([0.0688, 0.2159, 0.2035, 0.0960, 0.0746, 0.1205, 0.0826, 0.1123], + device='cuda:0'), in_proj_covar=tensor([0.0217, 0.0332, 0.0267, 0.0264, 0.0195, 0.0326, 0.0191, 0.0235], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:50:04,869 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 01:50:11,323 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.286e+02 1.984e+02 2.400e+02 2.943e+02 4.404e+02, threshold=4.799e+02, percent-clipped=0.0 +2023-03-21 01:50:14,860 INFO [train.py:901] (0/2) Epoch 21, batch 1250, loss[loss=0.1356, simple_loss=0.2194, pruned_loss=0.02596, over 7292.00 frames. ], tot_loss[loss=0.1491, simple_loss=0.2272, pruned_loss=0.03553, over 1436107.79 frames. ], batch size: 77, lr: 7.87e-03, grad_scale: 8.0 +2023-03-21 01:50:16,019 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=57733.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:50:27,115 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=57754.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:50:29,230 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 01:50:32,835 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 01:50:33,816 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 01:50:40,371 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8336, 2.0208, 1.5955, 2.5450, 2.5456, 2.7696, 2.0294, 2.3796], + device='cuda:0'), covar=tensor([0.1880, 0.0831, 0.3326, 0.0775, 0.0196, 0.0203, 0.0243, 0.0359], + device='cuda:0'), in_proj_covar=tensor([0.0270, 0.0233, 0.0268, 0.0262, 0.0162, 0.0160, 0.0191, 0.0208], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:50:40,695 INFO [train.py:901] (0/2) Epoch 21, batch 1300, loss[loss=0.1886, simple_loss=0.2639, pruned_loss=0.05668, over 6741.00 frames. ], tot_loss[loss=0.1487, simple_loss=0.2268, pruned_loss=0.03524, over 1437326.27 frames. ], batch size: 107, lr: 7.87e-03, grad_scale: 8.0 +2023-03-21 01:50:42,003 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.98 vs. limit=2.0 +2023-03-21 01:50:44,176 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2682, 4.3886, 4.1888, 4.4492, 3.8877, 4.3940, 4.6982, 4.7437], + device='cuda:0'), covar=tensor([0.0160, 0.0120, 0.0162, 0.0150, 0.0315, 0.0188, 0.0183, 0.0141], + device='cuda:0'), in_proj_covar=tensor([0.0112, 0.0111, 0.0104, 0.0110, 0.0102, 0.0094, 0.0091, 0.0088], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:50:47,156 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=57794.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:50:56,679 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 01:50:58,721 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 01:51:00,279 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=57818.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:51:03,388 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 01:51:03,840 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.320e+02 1.899e+02 2.250e+02 2.724e+02 9.280e+02, threshold=4.499e+02, percent-clipped=3.0 +2023-03-21 01:51:07,388 INFO [train.py:901] (0/2) Epoch 21, batch 1350, loss[loss=0.1284, simple_loss=0.2041, pruned_loss=0.02632, over 7226.00 frames. ], tot_loss[loss=0.1488, simple_loss=0.227, pruned_loss=0.03532, over 1438616.21 frames. ], batch size: 45, lr: 7.87e-03, grad_scale: 8.0 +2023-03-21 01:51:12,723 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 01:51:12,817 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=57842.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:51:17,257 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=57851.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:51:18,272 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=57853.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 01:51:32,193 INFO [train.py:901] (0/2) Epoch 21, batch 1400, loss[loss=0.1428, simple_loss=0.2253, pruned_loss=0.03018, over 7293.00 frames. ], tot_loss[loss=0.1505, simple_loss=0.2282, pruned_loss=0.03643, over 1438788.49 frames. ], batch size: 86, lr: 7.86e-03, grad_scale: 16.0 +2023-03-21 01:51:37,324 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=57890.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:51:42,476 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=57900.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:51:42,898 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=57901.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:51:45,505 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 01:51:54,952 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.290e+02 2.029e+02 2.540e+02 2.891e+02 4.630e+02, threshold=5.079e+02, percent-clipped=1.0 +2023-03-21 01:51:57,163 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9850, 2.9814, 2.1056, 3.7270, 2.3318, 2.8423, 1.5706, 2.1357], + device='cuda:0'), covar=tensor([0.0319, 0.0648, 0.2377, 0.0506, 0.0399, 0.0498, 0.2819, 0.1579], + device='cuda:0'), in_proj_covar=tensor([0.0267, 0.0253, 0.0295, 0.0259, 0.0266, 0.0258, 0.0255, 0.0278], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 01:51:58,497 INFO [train.py:901] (0/2) Epoch 21, batch 1450, loss[loss=0.1329, simple_loss=0.212, pruned_loss=0.02692, over 7210.00 frames. ], tot_loss[loss=0.1495, simple_loss=0.227, pruned_loss=0.03598, over 1438600.27 frames. ], batch size: 39, lr: 7.86e-03, grad_scale: 16.0 +2023-03-21 01:52:06,950 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=57948.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:52:08,601 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7457, 2.7759, 1.9204, 3.3481, 1.9196, 2.7262, 1.3681, 1.8364], + device='cuda:0'), covar=tensor([0.0399, 0.0776, 0.2437, 0.0625, 0.0346, 0.0725, 0.3227, 0.1737], + device='cuda:0'), in_proj_covar=tensor([0.0268, 0.0253, 0.0296, 0.0260, 0.0267, 0.0260, 0.0256, 0.0279], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 01:52:08,923 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 01:52:21,658 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6426, 2.8333, 2.5864, 2.7818, 2.8307, 2.6064, 2.8389, 2.7370], + device='cuda:0'), covar=tensor([0.0724, 0.1054, 0.0914, 0.1200, 0.1028, 0.0691, 0.1034, 0.1098], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0047, 0.0056, 0.0049, 0.0048, 0.0049, 0.0050, 0.0045], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:52:23,981 INFO [train.py:901] (0/2) Epoch 21, batch 1500, loss[loss=0.1605, simple_loss=0.2365, pruned_loss=0.04229, over 7124.00 frames. ], tot_loss[loss=0.1494, simple_loss=0.2269, pruned_loss=0.036, over 1438408.04 frames. ], batch size: 98, lr: 7.86e-03, grad_scale: 16.0 +2023-03-21 01:52:25,498 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 01:52:46,378 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.254e+02 1.930e+02 2.338e+02 2.743e+02 1.049e+03, threshold=4.675e+02, percent-clipped=2.0 +2023-03-21 01:52:46,683 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.64 vs. limit=2.0 +2023-03-21 01:52:49,909 INFO [train.py:901] (0/2) Epoch 21, batch 1550, loss[loss=0.1442, simple_loss=0.229, pruned_loss=0.02967, over 7139.00 frames. ], tot_loss[loss=0.1496, simple_loss=0.2274, pruned_loss=0.03592, over 1438342.71 frames. ], batch size: 98, lr: 7.85e-03, grad_scale: 16.0 +2023-03-21 01:52:49,919 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 01:53:01,636 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=58054.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:53:07,804 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4252, 3.6119, 3.4440, 3.6259, 3.2377, 3.5888, 3.9791, 3.9278], + device='cuda:0'), covar=tensor([0.0309, 0.0199, 0.0268, 0.0233, 0.0389, 0.0340, 0.0203, 0.0190], + device='cuda:0'), in_proj_covar=tensor([0.0113, 0.0113, 0.0103, 0.0110, 0.0103, 0.0094, 0.0090, 0.0089], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:53:15,770 INFO [train.py:901] (0/2) Epoch 21, batch 1600, loss[loss=0.1603, simple_loss=0.2423, pruned_loss=0.03919, over 6746.00 frames. ], tot_loss[loss=0.1502, simple_loss=0.228, pruned_loss=0.03622, over 1438989.36 frames. ], batch size: 107, lr: 7.85e-03, grad_scale: 16.0 +2023-03-21 01:53:20,334 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 01:53:20,392 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=58089.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:53:20,855 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 01:53:24,353 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 01:53:26,870 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=58102.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:53:33,378 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 01:53:35,005 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=58118.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:53:37,462 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 01:53:37,912 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.450e+02 2.083e+02 2.533e+02 2.946e+02 4.318e+02, threshold=5.065e+02, percent-clipped=0.0 +2023-03-21 01:53:38,624 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8542, 2.1697, 1.6194, 2.6134, 2.6497, 2.8577, 2.1932, 2.5257], + device='cuda:0'), covar=tensor([0.2078, 0.0964, 0.3476, 0.0656, 0.0156, 0.0154, 0.0217, 0.0313], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0236, 0.0271, 0.0264, 0.0164, 0.0162, 0.0192, 0.0210], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:53:41,401 INFO [train.py:901] (0/2) Epoch 21, batch 1650, loss[loss=0.1566, simple_loss=0.2316, pruned_loss=0.04075, over 7311.00 frames. ], tot_loss[loss=0.15, simple_loss=0.228, pruned_loss=0.03598, over 1440725.76 frames. ], batch size: 83, lr: 7.85e-03, grad_scale: 16.0 +2023-03-21 01:53:45,364 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 01:53:46,438 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=58141.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:53:47,940 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=58144.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:53:52,028 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=58151.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:53:55,511 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5740, 4.1205, 3.9922, 4.5580, 4.3647, 4.5198, 3.9263, 4.0528], + device='cuda:0'), covar=tensor([0.0976, 0.2489, 0.2677, 0.1071, 0.1064, 0.1242, 0.0897, 0.1206], + device='cuda:0'), in_proj_covar=tensor([0.0224, 0.0339, 0.0271, 0.0271, 0.0199, 0.0332, 0.0196, 0.0240], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:53:59,525 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=58166.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:54:02,204 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0755, 3.9478, 3.4696, 3.4671, 3.3971, 2.2526, 1.7633, 4.1057], + device='cuda:0'), covar=tensor([0.0041, 0.0048, 0.0093, 0.0061, 0.0090, 0.0428, 0.0559, 0.0046], + device='cuda:0'), in_proj_covar=tensor([0.0091, 0.0076, 0.0096, 0.0081, 0.0106, 0.0119, 0.0120, 0.0087], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 01:54:04,091 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 01:54:07,592 INFO [train.py:901] (0/2) Epoch 21, batch 1700, loss[loss=0.1301, simple_loss=0.2116, pruned_loss=0.02428, over 7319.00 frames. ], tot_loss[loss=0.1503, simple_loss=0.2283, pruned_loss=0.0361, over 1440956.28 frames. ], batch size: 42, lr: 7.84e-03, grad_scale: 16.0 +2023-03-21 01:54:08,131 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 01:54:13,769 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3408, 0.8607, 1.5687, 1.6967, 1.5272, 1.6464, 1.2752, 1.5926], + device='cuda:0'), covar=tensor([0.3088, 0.3476, 0.0929, 0.2119, 0.8037, 0.1275, 0.1552, 0.2292], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0061, 0.0047, 0.0043, 0.0046, 0.0048, 0.0069, 0.0048], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 01:54:16,685 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=58199.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:54:18,616 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=58202.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:54:19,507 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 01:54:20,157 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=58205.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:54:29,470 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.229e+02 2.007e+02 2.325e+02 2.996e+02 5.829e+02, threshold=4.650e+02, percent-clipped=1.0 +2023-03-21 01:54:32,996 INFO [train.py:901] (0/2) Epoch 21, batch 1750, loss[loss=0.1481, simple_loss=0.2346, pruned_loss=0.03085, over 7303.00 frames. ], tot_loss[loss=0.1504, simple_loss=0.2283, pruned_loss=0.03624, over 1442647.92 frames. ], batch size: 59, lr: 7.84e-03, grad_scale: 16.0 +2023-03-21 01:54:40,881 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7519, 2.9145, 2.6517, 3.0478, 2.8634, 2.6271, 2.9828, 2.7197], + device='cuda:0'), covar=tensor([0.0679, 0.1004, 0.1434, 0.0812, 0.0964, 0.1022, 0.0827, 0.1775], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0048, 0.0056, 0.0049, 0.0049, 0.0049, 0.0049, 0.0045], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:54:42,855 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8266, 3.2571, 3.7233, 3.5413, 3.9436, 3.8685, 3.6943, 3.6644], + device='cuda:0'), covar=tensor([0.0027, 0.0086, 0.0026, 0.0038, 0.0023, 0.0023, 0.0042, 0.0043], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0054, 0.0047, 0.0046, 0.0044, 0.0049, 0.0043, 0.0059], + device='cuda:0'), out_proj_covar=tensor([7.4843e-05, 1.2902e-04, 1.0261e-04, 9.5244e-05, 8.9532e-05, 1.0108e-04, + 9.7727e-05, 1.2306e-04], device='cuda:0') +2023-03-21 01:54:43,268 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 01:54:44,241 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 01:54:46,841 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2524, 4.3495, 4.1672, 4.3565, 4.0408, 4.3077, 4.5613, 4.6181], + device='cuda:0'), covar=tensor([0.0151, 0.0142, 0.0175, 0.0134, 0.0241, 0.0215, 0.0204, 0.0144], + device='cuda:0'), in_proj_covar=tensor([0.0111, 0.0112, 0.0102, 0.0109, 0.0102, 0.0092, 0.0089, 0.0088], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:54:59,450 INFO [train.py:901] (0/2) Epoch 21, batch 1800, loss[loss=0.171, simple_loss=0.2528, pruned_loss=0.04464, over 6702.00 frames. ], tot_loss[loss=0.1504, simple_loss=0.2285, pruned_loss=0.03611, over 1443304.58 frames. ], batch size: 106, lr: 7.84e-03, grad_scale: 16.0 +2023-03-21 01:55:06,589 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 01:55:18,596 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7869, 4.3776, 4.3027, 4.8684, 4.6820, 4.7662, 4.0273, 4.3352], + device='cuda:0'), covar=tensor([0.0872, 0.2490, 0.2101, 0.0861, 0.0783, 0.1152, 0.0875, 0.0997], + device='cuda:0'), in_proj_covar=tensor([0.0217, 0.0327, 0.0260, 0.0261, 0.0192, 0.0320, 0.0189, 0.0232], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:55:19,628 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 01:55:21,702 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.259e+02 2.040e+02 2.436e+02 2.888e+02 6.459e+02, threshold=4.872e+02, percent-clipped=2.0 +2023-03-21 01:55:25,270 INFO [train.py:901] (0/2) Epoch 21, batch 1850, loss[loss=0.155, simple_loss=0.2359, pruned_loss=0.03702, over 7334.00 frames. ], tot_loss[loss=0.1499, simple_loss=0.2279, pruned_loss=0.03591, over 1442007.35 frames. ], batch size: 61, lr: 7.83e-03, grad_scale: 16.0 +2023-03-21 01:55:29,927 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 01:55:33,296 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1348, 3.8331, 3.8396, 3.8750, 3.3793, 3.7296, 3.9975, 3.5905], + device='cuda:0'), covar=tensor([0.0164, 0.0173, 0.0137, 0.0169, 0.0583, 0.0154, 0.0197, 0.0193], + device='cuda:0'), in_proj_covar=tensor([0.0077, 0.0084, 0.0082, 0.0075, 0.0148, 0.0094, 0.0090, 0.0093], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:55:42,851 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8861, 4.5640, 4.4106, 4.9012, 4.7161, 4.8894, 4.0927, 4.4896], + device='cuda:0'), covar=tensor([0.0752, 0.1869, 0.1780, 0.0916, 0.0834, 0.1012, 0.0783, 0.0985], + device='cuda:0'), in_proj_covar=tensor([0.0218, 0.0326, 0.0261, 0.0264, 0.0193, 0.0322, 0.0189, 0.0233], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:55:46,254 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 01:55:47,860 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9483, 2.7718, 3.3203, 2.9478, 3.1819, 2.6942, 2.7112, 3.0132], + device='cuda:0'), covar=tensor([0.2106, 0.0745, 0.1415, 0.1769, 0.0888, 0.1772, 0.2462, 0.2224], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0058, 0.0045, 0.0043, 0.0043, 0.0041, 0.0059, 0.0048], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:55:51,233 INFO [train.py:901] (0/2) Epoch 21, batch 1900, loss[loss=0.1663, simple_loss=0.2434, pruned_loss=0.04463, over 7307.00 frames. ], tot_loss[loss=0.1502, simple_loss=0.2284, pruned_loss=0.03601, over 1443392.83 frames. ], batch size: 59, lr: 7.83e-03, grad_scale: 16.0 +2023-03-21 01:55:55,315 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=58389.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:55:58,442 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3857, 2.4475, 2.2505, 3.5022, 1.5527, 3.2166, 1.2979, 3.0592], + device='cuda:0'), covar=tensor([0.0108, 0.0999, 0.1507, 0.0153, 0.3652, 0.0141, 0.1139, 0.0274], + device='cuda:0'), in_proj_covar=tensor([0.0188, 0.0264, 0.0278, 0.0183, 0.0269, 0.0193, 0.0252, 0.0224], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:55:59,866 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.0622, 4.6702, 4.5843, 5.0474, 4.9119, 5.0480, 4.2396, 4.7128], + device='cuda:0'), covar=tensor([0.0800, 0.1930, 0.2063, 0.0853, 0.0814, 0.0924, 0.0761, 0.0872], + device='cuda:0'), in_proj_covar=tensor([0.0220, 0.0328, 0.0263, 0.0267, 0.0196, 0.0324, 0.0192, 0.0235], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:56:11,951 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 01:56:13,959 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.177e+02 1.959e+02 2.260e+02 2.693e+02 5.838e+02, threshold=4.521e+02, percent-clipped=2.0 +2023-03-21 01:56:17,582 INFO [train.py:901] (0/2) Epoch 21, batch 1950, loss[loss=0.1529, simple_loss=0.234, pruned_loss=0.03594, over 7346.00 frames. ], tot_loss[loss=0.1503, simple_loss=0.2284, pruned_loss=0.03607, over 1443813.20 frames. ], batch size: 61, lr: 7.83e-03, grad_scale: 16.0 +2023-03-21 01:56:21,167 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=58437.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:56:24,120 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 01:56:28,155 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 01:56:28,663 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 01:56:43,124 INFO [train.py:901] (0/2) Epoch 21, batch 2000, loss[loss=0.1654, simple_loss=0.2406, pruned_loss=0.04506, over 7284.00 frames. ], tot_loss[loss=0.1496, simple_loss=0.2279, pruned_loss=0.03565, over 1445735.21 frames. ], batch size: 66, lr: 7.82e-03, grad_scale: 16.0 +2023-03-21 01:56:45,620 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 01:56:51,766 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=58497.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:56:53,294 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=58500.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:56:56,784 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 01:56:57,735 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-03-21 01:57:04,194 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 01:57:05,206 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.357e+02 1.942e+02 2.312e+02 2.807e+02 4.995e+02, threshold=4.624e+02, percent-clipped=1.0 +2023-03-21 01:57:09,400 INFO [train.py:901] (0/2) Epoch 21, batch 2050, loss[loss=0.105, simple_loss=0.1809, pruned_loss=0.01453, over 7053.00 frames. ], tot_loss[loss=0.1495, simple_loss=0.2276, pruned_loss=0.03565, over 1444257.04 frames. ], batch size: 35, lr: 7.82e-03, grad_scale: 16.0 +2023-03-21 01:57:10,319 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.82 vs. limit=5.0 +2023-03-21 01:57:35,027 INFO [train.py:901] (0/2) Epoch 21, batch 2100, loss[loss=0.1553, simple_loss=0.2382, pruned_loss=0.03623, over 7281.00 frames. ], tot_loss[loss=0.1492, simple_loss=0.2273, pruned_loss=0.03552, over 1444170.12 frames. ], batch size: 66, lr: 7.82e-03, grad_scale: 16.0 +2023-03-21 01:57:38,910 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 01:57:41,406 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 01:57:46,401 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3029, 3.9747, 3.8604, 3.7513, 3.4732, 2.5160, 1.7606, 4.4439], + device='cuda:0'), covar=tensor([0.0033, 0.0066, 0.0066, 0.0060, 0.0088, 0.0404, 0.0538, 0.0033], + device='cuda:0'), in_proj_covar=tensor([0.0092, 0.0077, 0.0095, 0.0081, 0.0108, 0.0121, 0.0119, 0.0088], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 01:57:57,408 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.377e+02 1.997e+02 2.319e+02 3.007e+02 6.323e+02, threshold=4.638e+02, percent-clipped=6.0 +2023-03-21 01:58:00,912 INFO [train.py:901] (0/2) Epoch 21, batch 2150, loss[loss=0.1584, simple_loss=0.24, pruned_loss=0.0384, over 7262.00 frames. ], tot_loss[loss=0.1496, simple_loss=0.2273, pruned_loss=0.03588, over 1442982.83 frames. ], batch size: 52, lr: 7.81e-03, grad_scale: 16.0 +2023-03-21 01:58:08,604 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6432, 3.4387, 3.7439, 3.4824, 3.1652, 3.1163, 3.6788, 2.8700], + device='cuda:0'), covar=tensor([0.0250, 0.0392, 0.0311, 0.0421, 0.0504, 0.0642, 0.0382, 0.1158], + device='cuda:0'), in_proj_covar=tensor([0.0325, 0.0332, 0.0266, 0.0352, 0.0305, 0.0302, 0.0335, 0.0284], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 01:58:20,701 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7808, 3.0912, 3.6935, 3.8524, 3.6767, 3.8302, 3.6535, 3.5816], + device='cuda:0'), covar=tensor([0.0028, 0.0117, 0.0041, 0.0036, 0.0046, 0.0032, 0.0054, 0.0057], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0056, 0.0048, 0.0047, 0.0046, 0.0050, 0.0045, 0.0060], + device='cuda:0'), out_proj_covar=tensor([7.7231e-05, 1.3211e-04, 1.0513e-04, 9.6996e-05, 9.3991e-05, 1.0201e-04, + 1.0093e-04, 1.2670e-04], device='cuda:0') +2023-03-21 01:58:21,156 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8819, 3.7223, 3.1526, 3.3947, 2.8550, 2.0518, 1.6320, 3.9919], + device='cuda:0'), covar=tensor([0.0066, 0.0078, 0.0158, 0.0099, 0.0194, 0.0617, 0.0692, 0.0059], + device='cuda:0'), in_proj_covar=tensor([0.0093, 0.0078, 0.0097, 0.0082, 0.0110, 0.0122, 0.0121, 0.0089], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 01:58:26,534 INFO [train.py:901] (0/2) Epoch 21, batch 2200, loss[loss=0.1437, simple_loss=0.2216, pruned_loss=0.03291, over 7155.00 frames. ], tot_loss[loss=0.1491, simple_loss=0.2271, pruned_loss=0.03556, over 1444060.84 frames. ], batch size: 41, lr: 7.81e-03, grad_scale: 16.0 +2023-03-21 01:58:26,644 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=58681.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:58:27,525 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 01:58:33,608 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.82 vs. limit=2.0 +2023-03-21 01:58:48,513 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.237e+02 1.897e+02 2.320e+02 2.979e+02 4.171e+02, threshold=4.640e+02, percent-clipped=0.0 +2023-03-21 01:58:52,125 INFO [train.py:901] (0/2) Epoch 21, batch 2250, loss[loss=0.1335, simple_loss=0.2181, pruned_loss=0.02449, over 7146.00 frames. ], tot_loss[loss=0.1492, simple_loss=0.2272, pruned_loss=0.03563, over 1442452.83 frames. ], batch size: 41, lr: 7.81e-03, grad_scale: 16.0 +2023-03-21 01:58:57,797 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=58742.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:59:01,713 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 01:59:02,200 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 01:59:14,735 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 01:59:14,868 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=58775.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:59:17,735 INFO [train.py:901] (0/2) Epoch 21, batch 2300, loss[loss=0.1483, simple_loss=0.2254, pruned_loss=0.03561, over 7272.00 frames. ], tot_loss[loss=0.1495, simple_loss=0.2276, pruned_loss=0.03572, over 1444271.40 frames. ], batch size: 70, lr: 7.80e-03, grad_scale: 16.0 +2023-03-21 01:59:26,491 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=58797.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:59:28,412 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5096, 3.6469, 3.4374, 3.7133, 3.4058, 3.6174, 3.9127, 3.9020], + device='cuda:0'), covar=tensor([0.0248, 0.0189, 0.0254, 0.0215, 0.0354, 0.0546, 0.0269, 0.0227], + device='cuda:0'), in_proj_covar=tensor([0.0113, 0.0113, 0.0104, 0.0110, 0.0103, 0.0094, 0.0089, 0.0088], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 01:59:28,437 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=58800.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:59:40,447 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.260e+02 1.931e+02 2.266e+02 2.658e+02 4.867e+02, threshold=4.532e+02, percent-clipped=2.0 +2023-03-21 01:59:43,943 INFO [train.py:901] (0/2) Epoch 21, batch 2350, loss[loss=0.1648, simple_loss=0.2443, pruned_loss=0.0426, over 7286.00 frames. ], tot_loss[loss=0.1494, simple_loss=0.2276, pruned_loss=0.03563, over 1444332.51 frames. ], batch size: 68, lr: 7.80e-03, grad_scale: 16.0 +2023-03-21 01:59:46,581 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=58836.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:59:51,620 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=58845.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 01:59:53,106 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=58848.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:00:01,582 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 02:00:08,762 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 02:00:10,246 INFO [train.py:901] (0/2) Epoch 21, batch 2400, loss[loss=0.1619, simple_loss=0.2433, pruned_loss=0.0402, over 7341.00 frames. ], tot_loss[loss=0.1498, simple_loss=0.2281, pruned_loss=0.03577, over 1443222.68 frames. ], batch size: 63, lr: 7.80e-03, grad_scale: 16.0 +2023-03-21 02:00:10,871 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3476, 1.4708, 1.3634, 1.3149, 1.6803, 1.2930, 1.4058, 1.0962], + device='cuda:0'), covar=tensor([0.0148, 0.0140, 0.0187, 0.0127, 0.0097, 0.0111, 0.0136, 0.0165], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0026, 0.0026, 0.0026, 0.0024, 0.0025, 0.0026, 0.0033], + device='cuda:0'), out_proj_covar=tensor([3.0982e-05, 2.9220e-05, 2.9426e-05, 2.9099e-05, 2.7902e-05, 2.7892e-05, + 3.0226e-05, 3.8571e-05], device='cuda:0') +2023-03-21 02:00:16,376 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2451, 1.0620, 1.7300, 1.6828, 1.6670, 1.8001, 1.2004, 1.6804], + device='cuda:0'), covar=tensor([0.2332, 0.2296, 0.1457, 0.1780, 0.0997, 0.1319, 0.1804, 0.1951], + device='cuda:0'), in_proj_covar=tensor([0.0055, 0.0061, 0.0048, 0.0045, 0.0047, 0.0050, 0.0071, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:00:18,763 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 02:00:21,778 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 02:00:22,852 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=58906.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:00:31,680 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.319e+02 1.956e+02 2.269e+02 2.844e+02 5.396e+02, threshold=4.539e+02, percent-clipped=1.0 +2023-03-21 02:00:35,225 INFO [train.py:901] (0/2) Epoch 21, batch 2450, loss[loss=0.1546, simple_loss=0.2379, pruned_loss=0.03563, over 7296.00 frames. ], tot_loss[loss=0.1495, simple_loss=0.2279, pruned_loss=0.03557, over 1444652.26 frames. ], batch size: 86, lr: 7.79e-03, grad_scale: 16.0 +2023-03-21 02:00:47,947 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 02:00:50,110 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1350, 2.7377, 3.3952, 2.8592, 3.2794, 2.9742, 2.6341, 3.1327], + device='cuda:0'), covar=tensor([0.1326, 0.0690, 0.1052, 0.1722, 0.0667, 0.1130, 0.2235, 0.1623], + device='cuda:0'), in_proj_covar=tensor([0.0046, 0.0056, 0.0044, 0.0043, 0.0042, 0.0040, 0.0058, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:00:54,780 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=58967.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:01:01,651 INFO [train.py:901] (0/2) Epoch 21, batch 2500, loss[loss=0.1587, simple_loss=0.2364, pruned_loss=0.04053, over 7346.00 frames. ], tot_loss[loss=0.1497, simple_loss=0.2282, pruned_loss=0.0356, over 1444146.22 frames. ], batch size: 54, lr: 7.79e-03, grad_scale: 16.0 +2023-03-21 02:01:14,069 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 02:01:24,729 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.296e+02 1.969e+02 2.261e+02 3.016e+02 6.580e+02, threshold=4.523e+02, percent-clipped=4.0 +2023-03-21 02:01:27,649 INFO [train.py:901] (0/2) Epoch 21, batch 2550, loss[loss=0.1444, simple_loss=0.2246, pruned_loss=0.03209, over 7317.00 frames. ], tot_loss[loss=0.15, simple_loss=0.228, pruned_loss=0.03594, over 1443288.29 frames. ], batch size: 59, lr: 7.79e-03, grad_scale: 8.0 +2023-03-21 02:01:30,769 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=59037.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:01:41,643 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6445, 1.9795, 1.5445, 2.4016, 2.5848, 2.4844, 1.9506, 2.0565], + device='cuda:0'), covar=tensor([0.1830, 0.0922, 0.3253, 0.0662, 0.0177, 0.0167, 0.0233, 0.0293], + device='cuda:0'), in_proj_covar=tensor([0.0266, 0.0232, 0.0260, 0.0258, 0.0162, 0.0161, 0.0190, 0.0205], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:01:47,095 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.2205, 4.7562, 4.6032, 5.2432, 5.0965, 5.2189, 4.7461, 4.7332], + device='cuda:0'), covar=tensor([0.0800, 0.2385, 0.2291, 0.0984, 0.0733, 0.1007, 0.0711, 0.1147], + device='cuda:0'), in_proj_covar=tensor([0.0223, 0.0337, 0.0268, 0.0274, 0.0200, 0.0331, 0.0194, 0.0239], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:01:53,461 INFO [train.py:901] (0/2) Epoch 21, batch 2600, loss[loss=0.1503, simple_loss=0.2268, pruned_loss=0.0369, over 7233.00 frames. ], tot_loss[loss=0.1497, simple_loss=0.2275, pruned_loss=0.03593, over 1442929.26 frames. ], batch size: 45, lr: 7.78e-03, grad_scale: 8.0 +2023-03-21 02:02:15,426 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.271e+02 1.926e+02 2.254e+02 2.670e+02 6.116e+02, threshold=4.508e+02, percent-clipped=2.0 +2023-03-21 02:02:18,493 INFO [train.py:901] (0/2) Epoch 21, batch 2650, loss[loss=0.175, simple_loss=0.2505, pruned_loss=0.04969, over 7349.00 frames. ], tot_loss[loss=0.1492, simple_loss=0.227, pruned_loss=0.03573, over 1443043.69 frames. ], batch size: 73, lr: 7.78e-03, grad_scale: 8.0 +2023-03-21 02:02:18,549 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=59131.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:02:34,679 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=59164.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:02:42,791 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.62 vs. limit=2.0 +2023-03-21 02:02:43,361 INFO [train.py:901] (0/2) Epoch 21, batch 2700, loss[loss=0.1621, simple_loss=0.2424, pruned_loss=0.04092, over 7309.00 frames. ], tot_loss[loss=0.149, simple_loss=0.2268, pruned_loss=0.03563, over 1440973.82 frames. ], batch size: 59, lr: 7.78e-03, grad_scale: 8.0 +2023-03-21 02:02:57,581 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0902, 3.2303, 3.4528, 3.2358, 3.6516, 3.1777, 2.9682, 3.4534], + device='cuda:0'), covar=tensor([0.2192, 0.0669, 0.1516, 0.1561, 0.0907, 0.1183, 0.2352, 0.1704], + device='cuda:0'), in_proj_covar=tensor([0.0046, 0.0056, 0.0044, 0.0042, 0.0042, 0.0040, 0.0058, 0.0046], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:03:01,000 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3370, 1.0940, 1.8681, 1.8986, 1.7681, 1.9427, 1.4752, 1.7494], + device='cuda:0'), covar=tensor([0.2238, 0.3041, 0.0910, 0.1320, 0.1083, 0.1602, 0.1730, 0.1824], + device='cuda:0'), in_proj_covar=tensor([0.0058, 0.0063, 0.0049, 0.0045, 0.0049, 0.0051, 0.0072, 0.0050], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:03:05,339 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.436e+02 1.893e+02 2.333e+02 2.936e+02 8.243e+02, threshold=4.666e+02, percent-clipped=3.0 +2023-03-21 02:03:05,490 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=59225.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:03:07,433 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0469, 3.8218, 3.3021, 3.4938, 3.4400, 2.2501, 1.6083, 4.1298], + device='cuda:0'), covar=tensor([0.0041, 0.0075, 0.0103, 0.0072, 0.0087, 0.0426, 0.0599, 0.0043], + device='cuda:0'), in_proj_covar=tensor([0.0090, 0.0076, 0.0094, 0.0081, 0.0107, 0.0118, 0.0118, 0.0087], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 02:03:08,327 INFO [train.py:901] (0/2) Epoch 21, batch 2750, loss[loss=0.1503, simple_loss=0.2261, pruned_loss=0.03723, over 7355.00 frames. ], tot_loss[loss=0.1498, simple_loss=0.2277, pruned_loss=0.03596, over 1441706.88 frames. ], batch size: 51, lr: 7.77e-03, grad_scale: 8.0 +2023-03-21 02:03:23,566 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=59262.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:03:33,225 INFO [train.py:901] (0/2) Epoch 21, batch 2800, loss[loss=0.1735, simple_loss=0.2468, pruned_loss=0.05015, over 7239.00 frames. ], tot_loss[loss=0.1494, simple_loss=0.2273, pruned_loss=0.03579, over 1442452.78 frames. ], batch size: 55, lr: 7.77e-03, grad_scale: 8.0 +2023-03-21 02:03:39,085 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3409, 1.1722, 1.9812, 1.9575, 1.8177, 2.0610, 1.5658, 1.8442], + device='cuda:0'), covar=tensor([0.2848, 0.4562, 0.1028, 0.1541, 0.1108, 0.1666, 0.1815, 0.1709], + device='cuda:0'), in_proj_covar=tensor([0.0057, 0.0062, 0.0049, 0.0046, 0.0048, 0.0050, 0.0071, 0.0050], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:03:39,381 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-03-21 02:03:46,176 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-21.pt +2023-03-21 02:04:06,328 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 02:04:06,383 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 02:04:06,409 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 02:04:06,426 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 02:04:06,611 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 02:04:06,629 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 02:04:06,641 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 02:04:06,748 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 02:04:06,825 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 02:04:06,841 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 02:04:06,857 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 02:04:07,402 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 02:04:07,408 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 02:04:07,416 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 02:04:07,561 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 02:04:07,563 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 02:04:07,625 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 02:04:10,093 INFO [train.py:901] (0/2) Epoch 22, batch 0, loss[loss=0.1482, simple_loss=0.23, pruned_loss=0.03316, over 7317.00 frames. ], tot_loss[loss=0.1482, simple_loss=0.23, pruned_loss=0.03316, over 7317.00 frames. ], batch size: 83, lr: 7.60e-03, grad_scale: 8.0 +2023-03-21 02:04:10,094 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 02:04:35,484 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5149, 4.5291, 4.2742, 4.4426, 4.3922, 4.2837, 4.7969, 4.7726], + device='cuda:0'), covar=tensor([0.0142, 0.0129, 0.0153, 0.0153, 0.0203, 0.0219, 0.0151, 0.0165], + device='cuda:0'), in_proj_covar=tensor([0.0116, 0.0115, 0.0107, 0.0112, 0.0104, 0.0095, 0.0090, 0.0091], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:04:36,421 INFO [train.py:935] (0/2) Epoch 22, validation: loss=0.1648, simple_loss=0.253, pruned_loss=0.03832, over 1622729.00 frames. +2023-03-21 02:04:36,421 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 02:04:43,513 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 02:04:46,413 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.332e+02 1.893e+02 2.208e+02 2.728e+02 4.951e+02, threshold=4.416e+02, percent-clipped=1.0 +2023-03-21 02:04:52,417 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=59337.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:04:53,320 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 02:04:53,578 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.49 vs. limit=5.0 +2023-03-21 02:05:00,407 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 02:05:01,396 INFO [train.py:901] (0/2) Epoch 22, batch 50, loss[loss=0.1664, simple_loss=0.2433, pruned_loss=0.04473, over 7354.00 frames. ], tot_loss[loss=0.1515, simple_loss=0.2307, pruned_loss=0.0362, over 327274.35 frames. ], batch size: 63, lr: 7.59e-03, grad_scale: 8.0 +2023-03-21 02:05:02,416 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 02:05:05,498 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 02:05:16,752 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=59385.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:05:20,202 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1259, 1.3584, 1.2172, 1.2226, 1.3189, 1.1581, 0.9828, 0.8434], + device='cuda:0'), covar=tensor([0.0181, 0.0102, 0.0143, 0.0141, 0.0085, 0.0107, 0.0236, 0.0135], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0026, 0.0025, 0.0026, 0.0023, 0.0025, 0.0027, 0.0033], + device='cuda:0'), out_proj_covar=tensor([3.1472e-05, 2.9114e-05, 2.8524e-05, 2.8943e-05, 2.7033e-05, 2.7606e-05, + 3.0539e-05, 3.7915e-05], device='cuda:0') +2023-03-21 02:05:24,183 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 02:05:24,667 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 02:05:28,579 INFO [train.py:901] (0/2) Epoch 22, batch 100, loss[loss=0.1442, simple_loss=0.2246, pruned_loss=0.03184, over 7298.00 frames. ], tot_loss[loss=0.1495, simple_loss=0.2277, pruned_loss=0.03562, over 573721.22 frames. ], batch size: 86, lr: 7.59e-03, grad_scale: 8.0 +2023-03-21 02:05:38,493 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.392e+02 1.976e+02 2.370e+02 2.738e+02 5.901e+02, threshold=4.739e+02, percent-clipped=2.0 +2023-03-21 02:05:39,690 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1148, 0.9625, 1.7304, 1.7557, 1.6436, 1.8292, 1.2130, 1.7007], + device='cuda:0'), covar=tensor([0.1390, 0.2371, 0.0978, 0.1576, 0.1063, 0.1294, 0.1135, 0.1236], + device='cuda:0'), in_proj_covar=tensor([0.0057, 0.0062, 0.0048, 0.0045, 0.0048, 0.0050, 0.0070, 0.0048], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:05:41,742 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=59431.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:05:53,650 INFO [train.py:901] (0/2) Epoch 22, batch 150, loss[loss=0.1635, simple_loss=0.2488, pruned_loss=0.03906, over 6849.00 frames. ], tot_loss[loss=0.1493, simple_loss=0.2277, pruned_loss=0.03547, over 766350.11 frames. ], batch size: 107, lr: 7.59e-03, grad_scale: 8.0 +2023-03-21 02:06:06,372 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=59479.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:06:17,961 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=59502.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:06:19,298 INFO [train.py:901] (0/2) Epoch 22, batch 200, loss[loss=0.1513, simple_loss=0.2268, pruned_loss=0.03793, over 7314.00 frames. ], tot_loss[loss=0.1501, simple_loss=0.2285, pruned_loss=0.0359, over 918444.82 frames. ], batch size: 59, lr: 7.58e-03, grad_scale: 8.0 +2023-03-21 02:06:24,840 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 02:06:26,923 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=59520.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:06:29,271 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.401e+02 1.757e+02 2.128e+02 2.718e+02 7.175e+02, threshold=4.257e+02, percent-clipped=2.0 +2023-03-21 02:06:29,320 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 02:06:30,174 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.89 vs. limit=2.0 +2023-03-21 02:06:35,287 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 02:06:44,166 INFO [train.py:901] (0/2) Epoch 22, batch 250, loss[loss=0.1641, simple_loss=0.241, pruned_loss=0.0436, over 7360.00 frames. ], tot_loss[loss=0.1501, simple_loss=0.2284, pruned_loss=0.03589, over 1036538.53 frames. ], batch size: 65, lr: 7.58e-03, grad_scale: 8.0 +2023-03-21 02:06:45,296 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7763, 3.7137, 2.8393, 3.2314, 2.4999, 2.0309, 1.4862, 3.7655], + device='cuda:0'), covar=tensor([0.0032, 0.0042, 0.0116, 0.0066, 0.0164, 0.0477, 0.0553, 0.0045], + device='cuda:0'), in_proj_covar=tensor([0.0092, 0.0078, 0.0097, 0.0083, 0.0110, 0.0122, 0.0120, 0.0089], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 02:06:47,683 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 02:06:47,802 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=59562.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:06:49,520 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=59563.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:06:49,539 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0314, 2.4418, 1.8006, 2.8998, 2.9559, 3.0135, 2.3770, 2.3949], + device='cuda:0'), covar=tensor([0.1783, 0.0836, 0.3268, 0.0588, 0.0140, 0.0151, 0.0268, 0.0245], + device='cuda:0'), in_proj_covar=tensor([0.0267, 0.0232, 0.0263, 0.0258, 0.0162, 0.0162, 0.0192, 0.0205], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:06:51,081 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7568, 2.7153, 2.7076, 2.7083, 2.3511, 2.4853, 2.8443, 2.1350], + device='cuda:0'), covar=tensor([0.0471, 0.0553, 0.0460, 0.0591, 0.0499, 0.0687, 0.0507, 0.1263], + device='cuda:0'), in_proj_covar=tensor([0.0326, 0.0332, 0.0266, 0.0359, 0.0305, 0.0299, 0.0339, 0.0288], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:07:00,994 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0967, 3.3848, 2.1746, 3.7703, 2.5291, 3.4886, 1.7889, 2.0592], + device='cuda:0'), covar=tensor([0.0356, 0.0808, 0.2813, 0.0436, 0.0429, 0.0622, 0.3386, 0.2148], + device='cuda:0'), in_proj_covar=tensor([0.0271, 0.0253, 0.0301, 0.0260, 0.0271, 0.0261, 0.0258, 0.0281], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 02:07:08,833 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1553, 3.0747, 3.5699, 3.0786, 3.4020, 2.9848, 2.9517, 3.3505], + device='cuda:0'), covar=tensor([0.1487, 0.0646, 0.0966, 0.1653, 0.0763, 0.1099, 0.1455, 0.1183], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0055, 0.0042, 0.0042, 0.0041, 0.0039, 0.0056, 0.0045], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:07:09,184 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 02:07:09,298 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3590, 4.2041, 3.7668, 3.7275, 3.5853, 2.5995, 1.9309, 4.3946], + device='cuda:0'), covar=tensor([0.0032, 0.0036, 0.0056, 0.0063, 0.0083, 0.0342, 0.0489, 0.0039], + device='cuda:0'), in_proj_covar=tensor([0.0091, 0.0077, 0.0096, 0.0082, 0.0109, 0.0121, 0.0118, 0.0088], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 02:07:10,671 INFO [train.py:901] (0/2) Epoch 22, batch 300, loss[loss=0.1422, simple_loss=0.2211, pruned_loss=0.03166, over 7316.00 frames. ], tot_loss[loss=0.1484, simple_loss=0.227, pruned_loss=0.03491, over 1125293.74 frames. ], batch size: 83, lr: 7.58e-03, grad_scale: 8.0 +2023-03-21 02:07:13,236 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=59610.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:07:17,248 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 02:07:20,809 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.341e+02 1.899e+02 2.338e+02 2.742e+02 5.758e+02, threshold=4.677e+02, percent-clipped=3.0 +2023-03-21 02:07:23,204 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.00 vs. limit=2.0 +2023-03-21 02:07:37,250 INFO [train.py:901] (0/2) Epoch 22, batch 350, loss[loss=0.1301, simple_loss=0.2092, pruned_loss=0.02546, over 7283.00 frames. ], tot_loss[loss=0.1482, simple_loss=0.2268, pruned_loss=0.03484, over 1196599.19 frames. ], batch size: 47, lr: 7.57e-03, grad_scale: 8.0 +2023-03-21 02:07:53,499 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 02:08:02,417 INFO [train.py:901] (0/2) Epoch 22, batch 400, loss[loss=0.1475, simple_loss=0.2238, pruned_loss=0.03555, over 7356.00 frames. ], tot_loss[loss=0.1476, simple_loss=0.2257, pruned_loss=0.03477, over 1248071.28 frames. ], batch size: 63, lr: 7.57e-03, grad_scale: 8.0 +2023-03-21 02:08:12,561 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.110e+02 1.969e+02 2.337e+02 2.871e+02 4.733e+02, threshold=4.673e+02, percent-clipped=1.0 +2023-03-21 02:08:28,720 INFO [train.py:901] (0/2) Epoch 22, batch 450, loss[loss=0.1506, simple_loss=0.2268, pruned_loss=0.03715, over 7279.00 frames. ], tot_loss[loss=0.148, simple_loss=0.2258, pruned_loss=0.03513, over 1291339.98 frames. ], batch size: 77, lr: 7.57e-03, grad_scale: 8.0 +2023-03-21 02:08:35,248 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 02:08:35,760 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 02:08:54,222 INFO [train.py:901] (0/2) Epoch 22, batch 500, loss[loss=0.1466, simple_loss=0.2247, pruned_loss=0.03422, over 7320.00 frames. ], tot_loss[loss=0.1472, simple_loss=0.2249, pruned_loss=0.03469, over 1324796.60 frames. ], batch size: 59, lr: 7.56e-03, grad_scale: 8.0 +2023-03-21 02:09:01,860 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=59820.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:09:04,740 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.315e+02 1.908e+02 2.279e+02 2.698e+02 3.738e+02, threshold=4.558e+02, percent-clipped=0.0 +2023-03-21 02:09:08,362 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 02:09:09,900 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 02:09:10,416 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 02:09:12,500 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 02:09:17,236 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=59848.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:09:17,818 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0973, 2.9117, 3.0637, 3.1195, 2.6829, 2.6869, 3.1420, 2.2857], + device='cuda:0'), covar=tensor([0.0336, 0.0492, 0.0440, 0.0480, 0.0502, 0.0667, 0.0638, 0.1427], + device='cuda:0'), in_proj_covar=tensor([0.0325, 0.0333, 0.0267, 0.0356, 0.0305, 0.0299, 0.0339, 0.0289], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:09:18,141 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 02:09:20,650 INFO [train.py:901] (0/2) Epoch 22, batch 550, loss[loss=0.1399, simple_loss=0.2218, pruned_loss=0.02894, over 7358.00 frames. ], tot_loss[loss=0.1465, simple_loss=0.2245, pruned_loss=0.03428, over 1349274.11 frames. ], batch size: 51, lr: 7.56e-03, grad_scale: 8.0 +2023-03-21 02:09:22,254 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=59858.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:09:27,294 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=59868.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:09:29,167 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.81 vs. limit=2.0 +2023-03-21 02:09:29,802 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 02:09:37,287 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 02:09:37,897 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=59889.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:09:40,671 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 02:09:45,701 INFO [train.py:901] (0/2) Epoch 22, batch 600, loss[loss=0.1515, simple_loss=0.2326, pruned_loss=0.0352, over 7323.00 frames. ], tot_loss[loss=0.1462, simple_loss=0.2241, pruned_loss=0.03414, over 1369427.90 frames. ], batch size: 75, lr: 7.56e-03, grad_scale: 8.0 +2023-03-21 02:09:48,347 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 02:09:48,499 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=59909.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:09:56,960 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.308e+02 1.951e+02 2.184e+02 2.714e+02 3.801e+02, threshold=4.369e+02, percent-clipped=0.0 +2023-03-21 02:10:04,558 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 02:10:09,771 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=59950.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:10:12,113 INFO [train.py:901] (0/2) Epoch 22, batch 650, loss[loss=0.1444, simple_loss=0.2312, pruned_loss=0.02879, over 7314.00 frames. ], tot_loss[loss=0.1463, simple_loss=0.2243, pruned_loss=0.03414, over 1386690.55 frames. ], batch size: 59, lr: 7.56e-03, grad_scale: 8.0 +2023-03-21 02:10:13,155 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 02:10:16,385 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-03-21 02:10:30,192 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 02:10:36,294 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-60000.pt +2023-03-21 02:10:42,487 INFO [train.py:901] (0/2) Epoch 22, batch 700, loss[loss=0.1448, simple_loss=0.2246, pruned_loss=0.03247, over 7293.00 frames. ], tot_loss[loss=0.1469, simple_loss=0.225, pruned_loss=0.03443, over 1401766.14 frames. ], batch size: 66, lr: 7.55e-03, grad_scale: 8.0 +2023-03-21 02:10:44,111 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0931, 2.7246, 2.0262, 2.9074, 3.0822, 2.6459, 2.4413, 2.1863], + device='cuda:0'), covar=tensor([0.1831, 0.0764, 0.2905, 0.0521, 0.0148, 0.0089, 0.0301, 0.0412], + device='cuda:0'), in_proj_covar=tensor([0.0270, 0.0233, 0.0265, 0.0262, 0.0165, 0.0164, 0.0195, 0.0209], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:10:44,471 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 02:10:52,397 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.325e+02 1.930e+02 2.266e+02 2.658e+02 4.058e+02, threshold=4.532e+02, percent-clipped=0.0 +2023-03-21 02:11:00,108 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=60040.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:11:06,180 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=60052.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:11:07,477 INFO [train.py:901] (0/2) Epoch 22, batch 750, loss[loss=0.1665, simple_loss=0.2421, pruned_loss=0.04551, over 7317.00 frames. ], tot_loss[loss=0.1468, simple_loss=0.2248, pruned_loss=0.0344, over 1411886.22 frames. ], batch size: 83, lr: 7.55e-03, grad_scale: 8.0 +2023-03-21 02:11:07,998 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 02:11:08,474 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 02:11:12,667 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=60065.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:11:17,624 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7490, 2.9598, 2.6395, 2.8974, 2.7912, 2.4224, 2.9041, 2.8007], + device='cuda:0'), covar=tensor([0.0892, 0.0666, 0.0994, 0.1443, 0.0897, 0.0846, 0.0672, 0.1347], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0049, 0.0056, 0.0049, 0.0048, 0.0050, 0.0050, 0.0046], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:11:19,058 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=60078.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:11:23,578 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 02:11:29,265 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 02:11:31,914 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=60101.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 02:11:33,736 INFO [train.py:901] (0/2) Epoch 22, batch 800, loss[loss=0.1546, simple_loss=0.2278, pruned_loss=0.04074, over 7323.00 frames. ], tot_loss[loss=0.147, simple_loss=0.2249, pruned_loss=0.03462, over 1420807.38 frames. ], batch size: 59, lr: 7.55e-03, grad_scale: 8.0 +2023-03-21 02:11:35,314 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 02:11:35,997 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7440, 2.7089, 2.8085, 2.9147, 2.4737, 2.4087, 3.0128, 2.1717], + device='cuda:0'), covar=tensor([0.0451, 0.0442, 0.0498, 0.0595, 0.0555, 0.0780, 0.0738, 0.1530], + device='cuda:0'), in_proj_covar=tensor([0.0331, 0.0338, 0.0270, 0.0360, 0.0310, 0.0305, 0.0344, 0.0291], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:11:36,344 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 02:11:37,910 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=60113.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 02:11:40,873 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6575, 2.8062, 2.5478, 2.7459, 2.6325, 2.2902, 2.7131, 2.5983], + device='cuda:0'), covar=tensor([0.0610, 0.0657, 0.0825, 0.1102, 0.1029, 0.0719, 0.1417, 0.1377], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0049, 0.0056, 0.0050, 0.0048, 0.0050, 0.0050, 0.0046], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:11:43,675 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.329e+02 1.981e+02 2.306e+02 2.691e+02 7.119e+02, threshold=4.612e+02, percent-clipped=1.0 +2023-03-21 02:11:44,365 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=60126.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:11:46,792 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 02:11:50,941 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=60139.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:11:58,619 INFO [train.py:901] (0/2) Epoch 22, batch 850, loss[loss=0.169, simple_loss=0.2388, pruned_loss=0.04963, over 7340.00 frames. ], tot_loss[loss=0.1475, simple_loss=0.2255, pruned_loss=0.0348, over 1426457.30 frames. ], batch size: 54, lr: 7.54e-03, grad_scale: 8.0 +2023-03-21 02:12:00,251 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=60158.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:12:05,787 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 02:12:05,797 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 02:12:11,867 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 02:12:13,729 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.89 vs. limit=2.0 +2023-03-21 02:12:14,876 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 02:12:21,381 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=60198.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:12:24,629 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=60204.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:12:25,056 INFO [train.py:901] (0/2) Epoch 22, batch 900, loss[loss=0.1432, simple_loss=0.2292, pruned_loss=0.02857, over 7314.00 frames. ], tot_loss[loss=0.1472, simple_loss=0.2251, pruned_loss=0.03469, over 1430751.13 frames. ], batch size: 80, lr: 7.54e-03, grad_scale: 8.0 +2023-03-21 02:12:25,605 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=60206.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:12:33,484 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.70 vs. limit=5.0 +2023-03-21 02:12:35,133 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.240e+02 1.857e+02 2.134e+02 2.609e+02 4.521e+02, threshold=4.269e+02, percent-clipped=0.0 +2023-03-21 02:12:45,212 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=60245.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:12:45,803 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5407, 1.0558, 1.9089, 2.0817, 1.6406, 2.1055, 1.7265, 1.8767], + device='cuda:0'), covar=tensor([0.2172, 0.2357, 0.0811, 0.1199, 0.2244, 0.1303, 0.1779, 0.2595], + device='cuda:0'), in_proj_covar=tensor([0.0057, 0.0061, 0.0048, 0.0043, 0.0046, 0.0048, 0.0071, 0.0048], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 02:12:50,761 INFO [train.py:901] (0/2) Epoch 22, batch 950, loss[loss=0.1513, simple_loss=0.2355, pruned_loss=0.03357, over 7331.00 frames. ], tot_loss[loss=0.1476, simple_loss=0.2258, pruned_loss=0.03466, over 1434686.18 frames. ], batch size: 59, lr: 7.54e-03, grad_scale: 8.0 +2023-03-21 02:12:52,697 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 02:12:53,458 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=60259.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:13:16,268 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 02:13:16,719 INFO [train.py:901] (0/2) Epoch 22, batch 1000, loss[loss=0.1381, simple_loss=0.2198, pruned_loss=0.02825, over 7322.00 frames. ], tot_loss[loss=0.1473, simple_loss=0.2256, pruned_loss=0.03453, over 1436066.72 frames. ], batch size: 75, lr: 7.53e-03, grad_scale: 8.0 +2023-03-21 02:13:26,709 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.173e+02 1.855e+02 2.168e+02 2.543e+02 4.857e+02, threshold=4.336e+02, percent-clipped=1.0 +2023-03-21 02:13:36,609 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 02:13:42,925 INFO [train.py:901] (0/2) Epoch 22, batch 1050, loss[loss=0.1564, simple_loss=0.2418, pruned_loss=0.03545, over 7294.00 frames. ], tot_loss[loss=0.148, simple_loss=0.2263, pruned_loss=0.03484, over 1437193.80 frames. ], batch size: 83, lr: 7.53e-03, grad_scale: 8.0 +2023-03-21 02:13:49,020 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8947, 4.0418, 3.8893, 4.0199, 3.8382, 4.0445, 4.2573, 4.2746], + device='cuda:0'), covar=tensor([0.0193, 0.0146, 0.0184, 0.0153, 0.0276, 0.0358, 0.0233, 0.0184], + device='cuda:0'), in_proj_covar=tensor([0.0114, 0.0114, 0.0106, 0.0112, 0.0105, 0.0095, 0.0091, 0.0091], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:13:53,794 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.75 vs. limit=5.0 +2023-03-21 02:13:56,630 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9554, 3.1920, 2.7793, 3.1972, 2.9807, 2.4238, 3.0912, 2.9292], + device='cuda:0'), covar=tensor([0.0605, 0.0727, 0.0857, 0.0734, 0.0646, 0.0958, 0.0796, 0.1254], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0048, 0.0056, 0.0048, 0.0047, 0.0049, 0.0050, 0.0046], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:13:58,051 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 02:14:01,982 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 02:14:03,578 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=60396.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 02:14:08,313 INFO [train.py:901] (0/2) Epoch 22, batch 1100, loss[loss=0.1597, simple_loss=0.231, pruned_loss=0.04423, over 7287.00 frames. ], tot_loss[loss=0.1474, simple_loss=0.2256, pruned_loss=0.03458, over 1437486.20 frames. ], batch size: 77, lr: 7.53e-03, grad_scale: 8.0 +2023-03-21 02:14:09,888 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=60408.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 02:14:17,041 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=60421.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:14:18,715 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9398, 2.6610, 2.7060, 2.8945, 2.4044, 2.3854, 2.9130, 2.2752], + device='cuda:0'), covar=tensor([0.0495, 0.0502, 0.0468, 0.0516, 0.0545, 0.0753, 0.0526, 0.1404], + device='cuda:0'), in_proj_covar=tensor([0.0325, 0.0328, 0.0266, 0.0356, 0.0303, 0.0298, 0.0337, 0.0286], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:14:19,038 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.201e+02 1.927e+02 2.269e+02 2.813e+02 4.355e+02, threshold=4.537e+02, percent-clipped=1.0 +2023-03-21 02:14:22,880 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3138, 1.6316, 1.3823, 1.4546, 1.6608, 1.4017, 1.4226, 0.9738], + device='cuda:0'), covar=tensor([0.0117, 0.0129, 0.0277, 0.0123, 0.0060, 0.0098, 0.0373, 0.0208], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0026, 0.0026, 0.0026, 0.0024, 0.0025, 0.0027, 0.0034], + device='cuda:0'), out_proj_covar=tensor([3.1091e-05, 2.9517e-05, 3.0283e-05, 2.9420e-05, 2.7380e-05, 2.7472e-05, + 3.0985e-05, 3.9153e-05], device='cuda:0') +2023-03-21 02:14:24,264 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=60434.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:14:31,373 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 02:14:31,864 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 02:14:34,876 INFO [train.py:901] (0/2) Epoch 22, batch 1150, loss[loss=0.1892, simple_loss=0.2563, pruned_loss=0.06104, over 6686.00 frames. ], tot_loss[loss=0.1475, simple_loss=0.226, pruned_loss=0.03451, over 1438303.61 frames. ], batch size: 106, lr: 7.52e-03, grad_scale: 8.0 +2023-03-21 02:14:43,375 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 02:14:44,364 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 02:14:49,949 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2924, 3.8362, 4.0422, 4.0164, 3.9040, 3.8826, 4.1902, 3.7356], + device='cuda:0'), covar=tensor([0.0127, 0.0169, 0.0117, 0.0135, 0.0399, 0.0107, 0.0128, 0.0166], + device='cuda:0'), in_proj_covar=tensor([0.0079, 0.0086, 0.0084, 0.0077, 0.0153, 0.0095, 0.0093, 0.0093], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:14:51,688 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.09 vs. limit=5.0 +2023-03-21 02:14:59,449 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=60504.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:14:59,863 INFO [train.py:901] (0/2) Epoch 22, batch 1200, loss[loss=0.1373, simple_loss=0.2172, pruned_loss=0.02864, over 7254.00 frames. ], tot_loss[loss=0.1476, simple_loss=0.2263, pruned_loss=0.03447, over 1440491.29 frames. ], batch size: 89, lr: 7.52e-03, grad_scale: 8.0 +2023-03-21 02:15:07,656 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.76 vs. limit=5.0 +2023-03-21 02:15:11,231 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.263e+02 2.000e+02 2.319e+02 2.999e+02 4.801e+02, threshold=4.638e+02, percent-clipped=2.0 +2023-03-21 02:15:16,601 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-21 02:15:17,714 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 02:15:21,365 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=60545.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:15:24,833 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=60552.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:15:25,888 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=60554.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:15:26,306 INFO [train.py:901] (0/2) Epoch 22, batch 1250, loss[loss=0.1455, simple_loss=0.2273, pruned_loss=0.03185, over 7294.00 frames. ], tot_loss[loss=0.1473, simple_loss=0.2254, pruned_loss=0.03457, over 1438790.86 frames. ], batch size: 68, lr: 7.52e-03, grad_scale: 8.0 +2023-03-21 02:15:36,004 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2220, 2.8210, 1.9361, 2.9834, 3.0644, 3.1696, 2.8893, 2.7035], + device='cuda:0'), covar=tensor([0.1903, 0.0780, 0.3428, 0.0616, 0.0154, 0.0114, 0.0260, 0.0286], + device='cuda:0'), in_proj_covar=tensor([0.0272, 0.0237, 0.0267, 0.0265, 0.0167, 0.0167, 0.0197, 0.0210], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:15:40,285 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 02:15:43,816 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 02:15:45,422 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 02:15:45,948 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=60593.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:15:52,799 INFO [train.py:901] (0/2) Epoch 22, batch 1300, loss[loss=0.1508, simple_loss=0.2257, pruned_loss=0.03799, over 7279.00 frames. ], tot_loss[loss=0.147, simple_loss=0.2253, pruned_loss=0.03433, over 1439284.49 frames. ], batch size: 57, lr: 7.51e-03, grad_scale: 8.0 +2023-03-21 02:16:02,812 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.367e+02 2.027e+02 2.390e+02 2.761e+02 3.911e+02, threshold=4.781e+02, percent-clipped=0.0 +2023-03-21 02:16:09,453 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 02:16:11,497 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 02:16:15,022 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 02:16:18,028 INFO [train.py:901] (0/2) Epoch 22, batch 1350, loss[loss=0.1382, simple_loss=0.2256, pruned_loss=0.02541, over 7272.00 frames. ], tot_loss[loss=0.1466, simple_loss=0.2252, pruned_loss=0.03399, over 1441599.43 frames. ], batch size: 64, lr: 7.51e-03, grad_scale: 8.0 +2023-03-21 02:16:25,968 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 02:16:40,008 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=60696.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:16:44,542 INFO [train.py:901] (0/2) Epoch 22, batch 1400, loss[loss=0.1172, simple_loss=0.1925, pruned_loss=0.02098, over 7196.00 frames. ], tot_loss[loss=0.147, simple_loss=0.2255, pruned_loss=0.03428, over 1443603.46 frames. ], batch size: 39, lr: 7.51e-03, grad_scale: 8.0 +2023-03-21 02:16:46,177 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=60708.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:16:51,219 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=60718.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:16:52,612 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=60721.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:16:53,637 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3109, 1.5486, 1.1976, 1.3573, 1.4999, 1.3744, 1.3161, 0.8097], + device='cuda:0'), covar=tensor([0.0125, 0.0140, 0.0320, 0.0165, 0.0087, 0.0118, 0.0189, 0.0168], + device='cuda:0'), in_proj_covar=tensor([0.0027, 0.0026, 0.0026, 0.0026, 0.0024, 0.0025, 0.0027, 0.0034], + device='cuda:0'), out_proj_covar=tensor([3.1300e-05, 2.9188e-05, 3.0062e-05, 2.9593e-05, 2.7548e-05, 2.7797e-05, + 3.1142e-05, 3.8787e-05], device='cuda:0') +2023-03-21 02:16:54,278 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-03-21 02:16:54,498 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.546e+02 1.876e+02 2.210e+02 2.686e+02 7.585e+02, threshold=4.419e+02, percent-clipped=2.0 +2023-03-21 02:16:59,040 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=60734.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:16:59,439 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 02:17:04,057 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=60744.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:17:09,487 INFO [train.py:901] (0/2) Epoch 22, batch 1450, loss[loss=0.119, simple_loss=0.1941, pruned_loss=0.02195, over 7181.00 frames. ], tot_loss[loss=0.148, simple_loss=0.2264, pruned_loss=0.03476, over 1443856.20 frames. ], batch size: 39, lr: 7.51e-03, grad_scale: 8.0 +2023-03-21 02:17:10,051 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=60756.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:17:17,101 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=60769.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:17:22,222 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=60779.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:17:24,141 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 02:17:24,184 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=60782.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:17:36,006 INFO [train.py:901] (0/2) Epoch 22, batch 1500, loss[loss=0.1478, simple_loss=0.2204, pruned_loss=0.03763, over 7254.00 frames. ], tot_loss[loss=0.1482, simple_loss=0.2262, pruned_loss=0.03504, over 1443958.04 frames. ], batch size: 47, lr: 7.50e-03, grad_scale: 8.0 +2023-03-21 02:17:39,952 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 02:17:41,634 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6307, 2.7926, 2.3680, 2.6979, 2.6798, 2.4454, 2.6734, 2.5595], + device='cuda:0'), covar=tensor([0.0902, 0.0641, 0.1287, 0.0876, 0.0760, 0.0513, 0.0753, 0.1181], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0048, 0.0056, 0.0048, 0.0046, 0.0049, 0.0049, 0.0045], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:17:45,914 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.359e+02 2.001e+02 2.411e+02 2.741e+02 4.472e+02, threshold=4.822e+02, percent-clipped=1.0 +2023-03-21 02:17:47,278 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.92 vs. limit=2.0 +2023-03-21 02:18:00,613 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=60854.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:18:01,626 INFO [train.py:901] (0/2) Epoch 22, batch 1550, loss[loss=0.1411, simple_loss=0.2264, pruned_loss=0.02794, over 7284.00 frames. ], tot_loss[loss=0.1485, simple_loss=0.2265, pruned_loss=0.03521, over 1443779.41 frames. ], batch size: 57, lr: 7.50e-03, grad_scale: 8.0 +2023-03-21 02:18:04,640 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 02:18:11,390 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6813, 2.8906, 3.5452, 3.6199, 3.5874, 3.6704, 3.5807, 3.3575], + device='cuda:0'), covar=tensor([0.0040, 0.0161, 0.0051, 0.0057, 0.0052, 0.0049, 0.0065, 0.0077], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0058, 0.0051, 0.0048, 0.0048, 0.0052, 0.0046, 0.0063], + device='cuda:0'), out_proj_covar=tensor([7.9428e-05, 1.3615e-04, 1.0993e-04, 9.7482e-05, 9.5172e-05, 1.0508e-04, + 1.0187e-04, 1.3122e-04], device='cuda:0') +2023-03-21 02:18:25,794 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=60902.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:18:27,247 INFO [train.py:901] (0/2) Epoch 22, batch 1600, loss[loss=0.1634, simple_loss=0.231, pruned_loss=0.04795, over 7279.00 frames. ], tot_loss[loss=0.1482, simple_loss=0.226, pruned_loss=0.03521, over 1442965.05 frames. ], batch size: 57, lr: 7.50e-03, grad_scale: 8.0 +2023-03-21 02:18:34,804 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 02:18:35,773 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 02:18:37,318 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.373e+02 1.951e+02 2.358e+02 2.775e+02 4.018e+02, threshold=4.716e+02, percent-clipped=0.0 +2023-03-21 02:18:38,367 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 02:18:49,094 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 02:18:53,672 INFO [train.py:901] (0/2) Epoch 22, batch 1650, loss[loss=0.1595, simple_loss=0.2353, pruned_loss=0.04185, over 7283.00 frames. ], tot_loss[loss=0.1485, simple_loss=0.2263, pruned_loss=0.03534, over 1442696.42 frames. ], batch size: 57, lr: 7.49e-03, grad_scale: 8.0 +2023-03-21 02:18:53,695 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 02:19:02,336 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 02:19:03,381 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5033, 5.0575, 5.1706, 5.0977, 4.8606, 4.5130, 5.1805, 4.9206], + device='cuda:0'), covar=tensor([0.0405, 0.0380, 0.0332, 0.0377, 0.0365, 0.0332, 0.0272, 0.0457], + device='cuda:0'), in_proj_covar=tensor([0.0193, 0.0226, 0.0171, 0.0170, 0.0136, 0.0203, 0.0180, 0.0135], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 02:19:04,006 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2143, 2.9109, 3.1098, 3.1358, 2.8293, 2.7426, 3.2359, 2.5648], + device='cuda:0'), covar=tensor([0.0356, 0.0361, 0.0420, 0.0483, 0.0437, 0.0592, 0.0557, 0.1317], + device='cuda:0'), in_proj_covar=tensor([0.0327, 0.0330, 0.0265, 0.0352, 0.0303, 0.0299, 0.0339, 0.0287], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:19:19,353 INFO [train.py:901] (0/2) Epoch 22, batch 1700, loss[loss=0.1243, simple_loss=0.1932, pruned_loss=0.02768, over 6991.00 frames. ], tot_loss[loss=0.1484, simple_loss=0.2261, pruned_loss=0.03535, over 1441270.23 frames. ], batch size: 35, lr: 7.49e-03, grad_scale: 16.0 +2023-03-21 02:19:19,381 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 02:19:23,996 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 02:19:30,088 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.186e+02 1.914e+02 2.304e+02 2.748e+02 4.764e+02, threshold=4.609e+02, percent-clipped=1.0 +2023-03-21 02:19:35,132 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 02:19:38,961 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7497, 2.4130, 2.8897, 2.7896, 2.8816, 2.5962, 2.0712, 2.7516], + device='cuda:0'), covar=tensor([0.1591, 0.0859, 0.1103, 0.1124, 0.1091, 0.1143, 0.3397, 0.1563], + device='cuda:0'), in_proj_covar=tensor([0.0045, 0.0056, 0.0042, 0.0041, 0.0042, 0.0039, 0.0057, 0.0045], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:19:45,765 INFO [train.py:901] (0/2) Epoch 22, batch 1750, loss[loss=0.1501, simple_loss=0.2321, pruned_loss=0.03403, over 7304.00 frames. ], tot_loss[loss=0.1481, simple_loss=0.226, pruned_loss=0.03516, over 1441708.83 frames. ], batch size: 59, lr: 7.49e-03, grad_scale: 16.0 +2023-03-21 02:19:48,930 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=61061.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:19:55,441 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=61074.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:20:00,890 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 02:20:01,918 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 02:20:09,573 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8943, 4.1440, 4.6056, 4.4756, 4.0880, 4.3407, 4.6684, 4.3245], + device='cuda:0'), covar=tensor([0.0108, 0.0171, 0.0109, 0.0135, 0.0515, 0.0123, 0.0139, 0.0133], + device='cuda:0'), in_proj_covar=tensor([0.0081, 0.0091, 0.0088, 0.0079, 0.0159, 0.0099, 0.0094, 0.0096], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:20:10,967 INFO [train.py:901] (0/2) Epoch 22, batch 1800, loss[loss=0.1446, simple_loss=0.2231, pruned_loss=0.03305, over 7238.00 frames. ], tot_loss[loss=0.1473, simple_loss=0.2252, pruned_loss=0.03469, over 1439717.06 frames. ], batch size: 93, lr: 7.48e-03, grad_scale: 16.0 +2023-03-21 02:20:20,010 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=61122.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:20:21,382 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.437e+02 1.953e+02 2.262e+02 2.819e+02 5.491e+02, threshold=4.524e+02, percent-clipped=2.0 +2023-03-21 02:20:22,948 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 02:20:30,871 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.80 vs. limit=2.0 +2023-03-21 02:20:37,055 INFO [train.py:901] (0/2) Epoch 22, batch 1850, loss[loss=0.1359, simple_loss=0.2195, pruned_loss=0.02619, over 7265.00 frames. ], tot_loss[loss=0.1476, simple_loss=0.2256, pruned_loss=0.03484, over 1440753.94 frames. ], batch size: 89, lr: 7.48e-03, grad_scale: 8.0 +2023-03-21 02:20:37,062 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 02:20:47,567 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 02:20:50,769 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-03-21 02:20:56,709 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9268, 3.1904, 2.7260, 3.0602, 2.9774, 2.6318, 2.9701, 2.9567], + device='cuda:0'), covar=tensor([0.0725, 0.0573, 0.0698, 0.1196, 0.0762, 0.0726, 0.1144, 0.0765], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0049, 0.0056, 0.0050, 0.0047, 0.0050, 0.0050, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:21:03,039 INFO [train.py:901] (0/2) Epoch 22, batch 1900, loss[loss=0.1458, simple_loss=0.2285, pruned_loss=0.03153, over 7248.00 frames. ], tot_loss[loss=0.1474, simple_loss=0.2256, pruned_loss=0.03461, over 1439825.65 frames. ], batch size: 55, lr: 7.48e-03, grad_scale: 8.0 +2023-03-21 02:21:04,548 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 02:21:13,002 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.85 vs. limit=5.0 +2023-03-21 02:21:14,170 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.375e+02 1.892e+02 2.344e+02 2.726e+02 6.407e+02, threshold=4.688e+02, percent-clipped=1.0 +2023-03-21 02:21:28,657 INFO [train.py:901] (0/2) Epoch 22, batch 1950, loss[loss=0.1486, simple_loss=0.2268, pruned_loss=0.03515, over 7280.00 frames. ], tot_loss[loss=0.1476, simple_loss=0.2261, pruned_loss=0.0346, over 1441081.52 frames. ], batch size: 57, lr: 7.47e-03, grad_scale: 8.0 +2023-03-21 02:21:29,203 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 02:21:40,214 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 02:21:45,332 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 02:21:45,843 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 02:21:54,364 INFO [train.py:901] (0/2) Epoch 22, batch 2000, loss[loss=0.1448, simple_loss=0.2265, pruned_loss=0.03155, over 7210.00 frames. ], tot_loss[loss=0.1478, simple_loss=0.2261, pruned_loss=0.03475, over 1441524.35 frames. ], batch size: 50, lr: 7.47e-03, grad_scale: 8.0 +2023-03-21 02:22:02,591 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 02:22:05,610 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.543e+02 1.924e+02 2.207e+02 2.701e+02 5.428e+02, threshold=4.414e+02, percent-clipped=1.0 +2023-03-21 02:22:12,590 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 02:22:20,032 INFO [train.py:901] (0/2) Epoch 22, batch 2050, loss[loss=0.1289, simple_loss=0.2109, pruned_loss=0.02343, over 7335.00 frames. ], tot_loss[loss=0.1472, simple_loss=0.2252, pruned_loss=0.03461, over 1439831.60 frames. ], batch size: 44, lr: 7.47e-03, grad_scale: 8.0 +2023-03-21 02:22:21,068 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 02:22:30,214 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=61374.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:22:46,756 INFO [train.py:901] (0/2) Epoch 22, batch 2100, loss[loss=0.1373, simple_loss=0.2205, pruned_loss=0.02702, over 7322.00 frames. ], tot_loss[loss=0.1462, simple_loss=0.2245, pruned_loss=0.03396, over 1442264.77 frames. ], batch size: 75, lr: 7.47e-03, grad_scale: 8.0 +2023-03-21 02:22:52,694 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=61417.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:22:54,165 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 02:22:55,230 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=61422.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:22:57,129 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.451e+02 2.046e+02 2.349e+02 2.806e+02 4.447e+02, threshold=4.697e+02, percent-clipped=1.0 +2023-03-21 02:22:57,664 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 02:23:11,615 INFO [train.py:901] (0/2) Epoch 22, batch 2150, loss[loss=0.1439, simple_loss=0.2267, pruned_loss=0.03056, over 7344.00 frames. ], tot_loss[loss=0.1467, simple_loss=0.2253, pruned_loss=0.03411, over 1443599.92 frames. ], batch size: 73, lr: 7.46e-03, grad_scale: 8.0 +2023-03-21 02:23:34,054 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4269, 3.9710, 3.9625, 4.4232, 4.3210, 4.4522, 3.8648, 4.0590], + device='cuda:0'), covar=tensor([0.0896, 0.2538, 0.2213, 0.1148, 0.0924, 0.1309, 0.0781, 0.1109], + device='cuda:0'), in_proj_covar=tensor([0.0221, 0.0340, 0.0267, 0.0270, 0.0202, 0.0334, 0.0194, 0.0239], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:23:38,227 INFO [train.py:901] (0/2) Epoch 22, batch 2200, loss[loss=0.1546, simple_loss=0.2318, pruned_loss=0.03867, over 7249.00 frames. ], tot_loss[loss=0.1475, simple_loss=0.2261, pruned_loss=0.03445, over 1444315.57 frames. ], batch size: 55, lr: 7.46e-03, grad_scale: 8.0 +2023-03-21 02:23:42,247 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 02:23:48,794 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.260e+02 1.877e+02 2.439e+02 2.820e+02 5.772e+02, threshold=4.878e+02, percent-clipped=2.0 +2023-03-21 02:24:01,089 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1594, 3.8404, 3.8551, 3.9627, 3.7854, 3.7515, 4.1742, 3.7145], + device='cuda:0'), covar=tensor([0.0162, 0.0155, 0.0141, 0.0142, 0.0506, 0.0142, 0.0131, 0.0172], + device='cuda:0'), in_proj_covar=tensor([0.0080, 0.0089, 0.0087, 0.0077, 0.0158, 0.0098, 0.0093, 0.0095], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:24:03,966 INFO [train.py:901] (0/2) Epoch 22, batch 2250, loss[loss=0.134, simple_loss=0.2099, pruned_loss=0.02903, over 7229.00 frames. ], tot_loss[loss=0.1477, simple_loss=0.2264, pruned_loss=0.03443, over 1442653.22 frames. ], batch size: 45, lr: 7.46e-03, grad_scale: 8.0 +2023-03-21 02:24:17,277 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 02:24:17,289 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 02:24:21,558 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.33 vs. limit=2.0 +2023-03-21 02:24:29,176 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 02:24:30,183 INFO [train.py:901] (0/2) Epoch 22, batch 2300, loss[loss=0.1446, simple_loss=0.223, pruned_loss=0.03303, over 7276.00 frames. ], tot_loss[loss=0.1472, simple_loss=0.2258, pruned_loss=0.03431, over 1442553.78 frames. ], batch size: 70, lr: 7.45e-03, grad_scale: 8.0 +2023-03-21 02:24:40,703 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.277e+02 1.952e+02 2.374e+02 2.759e+02 5.347e+02, threshold=4.748e+02, percent-clipped=1.0 +2023-03-21 02:24:42,909 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3824, 1.1027, 1.6222, 1.8522, 1.6269, 1.6955, 1.2428, 1.7378], + device='cuda:0'), covar=tensor([0.1612, 0.2306, 0.1303, 0.0873, 0.1148, 0.0808, 0.0876, 0.1815], + device='cuda:0'), in_proj_covar=tensor([0.0060, 0.0064, 0.0050, 0.0045, 0.0047, 0.0050, 0.0074, 0.0050], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:24:56,190 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5518, 3.6681, 3.5082, 3.6363, 3.4312, 3.5922, 3.8709, 3.9132], + device='cuda:0'), covar=tensor([0.0228, 0.0183, 0.0233, 0.0180, 0.0288, 0.0313, 0.0283, 0.0197], + device='cuda:0'), in_proj_covar=tensor([0.0111, 0.0112, 0.0104, 0.0109, 0.0102, 0.0093, 0.0092, 0.0089], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:24:56,610 INFO [train.py:901] (0/2) Epoch 22, batch 2350, loss[loss=0.1366, simple_loss=0.2178, pruned_loss=0.02774, over 7264.00 frames. ], tot_loss[loss=0.1469, simple_loss=0.2253, pruned_loss=0.03427, over 1440775.16 frames. ], batch size: 89, lr: 7.45e-03, grad_scale: 8.0 +2023-03-21 02:25:13,895 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.59 vs. limit=2.0 +2023-03-21 02:25:16,086 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 02:25:21,582 INFO [train.py:901] (0/2) Epoch 22, batch 2400, loss[loss=0.1342, simple_loss=0.2157, pruned_loss=0.02634, over 7254.00 frames. ], tot_loss[loss=0.1475, simple_loss=0.2263, pruned_loss=0.0344, over 1443751.22 frames. ], batch size: 89, lr: 7.45e-03, grad_scale: 8.0 +2023-03-21 02:25:21,621 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 02:25:28,435 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=61717.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:25:32,927 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.427e+02 1.971e+02 2.295e+02 2.794e+02 4.281e+02, threshold=4.590e+02, percent-clipped=0.0 +2023-03-21 02:25:33,465 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 02:25:34,608 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1182, 4.0983, 3.3670, 3.5410, 3.1273, 2.3974, 1.7208, 4.1175], + device='cuda:0'), covar=tensor([0.0039, 0.0041, 0.0090, 0.0068, 0.0112, 0.0381, 0.0537, 0.0048], + device='cuda:0'), in_proj_covar=tensor([0.0096, 0.0076, 0.0100, 0.0086, 0.0111, 0.0122, 0.0122, 0.0091], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 02:25:35,992 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 02:25:36,121 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=61732.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:25:42,253 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=61743.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:25:48,404 INFO [train.py:901] (0/2) Epoch 22, batch 2450, loss[loss=0.1565, simple_loss=0.2379, pruned_loss=0.03759, over 7271.00 frames. ], tot_loss[loss=0.1473, simple_loss=0.2263, pruned_loss=0.03414, over 1445508.33 frames. ], batch size: 52, lr: 7.44e-03, grad_scale: 8.0 +2023-03-21 02:25:53,488 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=61765.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:26:03,067 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 02:26:07,700 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=61793.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:26:08,149 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0826, 2.2821, 2.3661, 2.4107, 2.3316, 2.1168, 2.0395, 1.8232], + device='cuda:0'), covar=tensor([0.0469, 0.0372, 0.0274, 0.0127, 0.0726, 0.0380, 0.0214, 0.0314], + device='cuda:0'), in_proj_covar=tensor([0.0029, 0.0027, 0.0027, 0.0026, 0.0028, 0.0026, 0.0030, 0.0030], + device='cuda:0'), out_proj_covar=tensor([7.3012e-05, 7.2144e-05, 6.7978e-05, 6.7596e-05, 7.1592e-05, 6.7639e-05, + 7.3398e-05, 7.6351e-05], device='cuda:0') +2023-03-21 02:26:14,141 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=61804.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:26:14,520 INFO [train.py:901] (0/2) Epoch 22, batch 2500, loss[loss=0.143, simple_loss=0.2264, pruned_loss=0.02977, over 7339.00 frames. ], tot_loss[loss=0.1469, simple_loss=0.2257, pruned_loss=0.03408, over 1446114.66 frames. ], batch size: 59, lr: 7.44e-03, grad_scale: 8.0 +2023-03-21 02:26:25,714 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.431e+02 1.926e+02 2.287e+02 2.628e+02 4.851e+02, threshold=4.574e+02, percent-clipped=1.0 +2023-03-21 02:26:29,174 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 02:26:40,367 INFO [train.py:901] (0/2) Epoch 22, batch 2550, loss[loss=0.1543, simple_loss=0.2359, pruned_loss=0.03637, over 7243.00 frames. ], tot_loss[loss=0.1468, simple_loss=0.2254, pruned_loss=0.0341, over 1443745.70 frames. ], batch size: 55, lr: 7.44e-03, grad_scale: 8.0 +2023-03-21 02:26:51,907 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.4739, 4.8529, 4.8358, 5.4318, 5.2358, 5.3932, 4.9790, 4.8745], + device='cuda:0'), covar=tensor([0.0676, 0.2663, 0.2172, 0.0982, 0.0881, 0.1132, 0.0634, 0.1100], + device='cuda:0'), in_proj_covar=tensor([0.0218, 0.0340, 0.0264, 0.0265, 0.0200, 0.0327, 0.0194, 0.0238], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:26:52,088 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.57 vs. limit=5.0 +2023-03-21 02:27:05,956 INFO [train.py:901] (0/2) Epoch 22, batch 2600, loss[loss=0.141, simple_loss=0.2261, pruned_loss=0.02795, over 7293.00 frames. ], tot_loss[loss=0.1469, simple_loss=0.2252, pruned_loss=0.03434, over 1444244.42 frames. ], batch size: 68, lr: 7.44e-03, grad_scale: 8.0 +2023-03-21 02:27:10,508 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=61914.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:27:16,187 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.332e+02 1.945e+02 2.250e+02 2.728e+02 4.156e+02, threshold=4.499e+02, percent-clipped=0.0 +2023-03-21 02:27:20,664 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.89 vs. limit=2.0 +2023-03-21 02:27:26,329 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8764, 3.3226, 3.7797, 3.7686, 3.8752, 3.9334, 3.9373, 3.7022], + device='cuda:0'), covar=tensor([0.0029, 0.0099, 0.0030, 0.0037, 0.0030, 0.0027, 0.0035, 0.0049], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0059, 0.0051, 0.0049, 0.0048, 0.0053, 0.0046, 0.0063], + device='cuda:0'), out_proj_covar=tensor([8.0337e-05, 1.3785e-04, 1.1066e-04, 9.7118e-05, 9.6592e-05, 1.0586e-04, + 1.0138e-04, 1.3102e-04], device='cuda:0') +2023-03-21 02:27:31,132 INFO [train.py:901] (0/2) Epoch 22, batch 2650, loss[loss=0.1302, simple_loss=0.2058, pruned_loss=0.02731, over 7161.00 frames. ], tot_loss[loss=0.1472, simple_loss=0.2254, pruned_loss=0.03456, over 1443771.27 frames. ], batch size: 41, lr: 7.43e-03, grad_scale: 8.0 +2023-03-21 02:27:41,141 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=61975.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:27:54,827 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8155, 2.3936, 1.5730, 2.4486, 2.5967, 2.5519, 2.0303, 2.5332], + device='cuda:0'), covar=tensor([0.1793, 0.0834, 0.3494, 0.0674, 0.0169, 0.0209, 0.0223, 0.0254], + device='cuda:0'), in_proj_covar=tensor([0.0263, 0.0230, 0.0263, 0.0264, 0.0167, 0.0170, 0.0197, 0.0209], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:27:56,093 INFO [train.py:901] (0/2) Epoch 22, batch 2700, loss[loss=0.1352, simple_loss=0.2082, pruned_loss=0.03105, over 7299.00 frames. ], tot_loss[loss=0.1468, simple_loss=0.225, pruned_loss=0.03427, over 1443259.64 frames. ], batch size: 42, lr: 7.43e-03, grad_scale: 8.0 +2023-03-21 02:28:06,225 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.177e+02 2.000e+02 2.344e+02 2.679e+02 5.407e+02, threshold=4.688e+02, percent-clipped=1.0 +2023-03-21 02:28:12,451 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-21 02:28:12,754 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7123, 3.1834, 3.5744, 3.7598, 3.8553, 3.8295, 3.7997, 3.6385], + device='cuda:0'), covar=tensor([0.0030, 0.0106, 0.0035, 0.0038, 0.0031, 0.0028, 0.0035, 0.0050], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0059, 0.0051, 0.0048, 0.0048, 0.0052, 0.0045, 0.0063], + device='cuda:0'), out_proj_covar=tensor([7.9939e-05, 1.3752e-04, 1.0946e-04, 9.6006e-05, 9.5808e-05, 1.0538e-04, + 9.9571e-05, 1.2903e-04], device='cuda:0') +2023-03-21 02:28:19,833 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0573, 2.6654, 1.9113, 2.8508, 3.0828, 2.8596, 2.4521, 2.8004], + device='cuda:0'), covar=tensor([0.1827, 0.0810, 0.3235, 0.0513, 0.0166, 0.0156, 0.0275, 0.0305], + device='cuda:0'), in_proj_covar=tensor([0.0267, 0.0233, 0.0265, 0.0267, 0.0169, 0.0171, 0.0200, 0.0211], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:28:20,673 INFO [train.py:901] (0/2) Epoch 22, batch 2750, loss[loss=0.1572, simple_loss=0.2269, pruned_loss=0.04376, over 7262.00 frames. ], tot_loss[loss=0.1474, simple_loss=0.226, pruned_loss=0.03446, over 1444331.34 frames. ], batch size: 47, lr: 7.43e-03, grad_scale: 8.0 +2023-03-21 02:28:36,985 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=62088.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:28:42,356 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=62099.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:28:44,420 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9261, 2.6545, 2.3096, 3.8583, 1.8436, 3.6466, 1.5013, 2.7159], + device='cuda:0'), covar=tensor([0.0120, 0.0945, 0.1765, 0.0126, 0.3773, 0.0156, 0.1205, 0.0309], + device='cuda:0'), in_proj_covar=tensor([0.0193, 0.0262, 0.0278, 0.0185, 0.0266, 0.0194, 0.0252, 0.0227], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:28:45,255 INFO [train.py:901] (0/2) Epoch 22, batch 2800, loss[loss=0.158, simple_loss=0.2303, pruned_loss=0.04282, over 7238.00 frames. ], tot_loss[loss=0.1485, simple_loss=0.2271, pruned_loss=0.03489, over 1444222.39 frames. ], batch size: 55, lr: 7.42e-03, grad_scale: 8.0 +2023-03-21 02:28:53,562 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=62122.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:28:55,319 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.488e+02 1.965e+02 2.306e+02 2.892e+02 4.105e+02, threshold=4.613e+02, percent-clipped=0.0 +2023-03-21 02:28:57,662 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-22.pt +2023-03-21 02:29:16,047 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 02:29:16,123 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 02:29:16,156 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 02:29:16,180 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 02:29:16,805 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 02:29:16,828 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 02:29:16,843 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 02:29:16,991 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 02:29:17,098 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 02:29:17,119 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 02:29:17,144 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 02:29:17,383 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 02:29:17,391 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 02:29:17,402 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 02:29:17,606 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 02:29:17,610 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 02:29:17,697 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 02:29:20,240 INFO [train.py:901] (0/2) Epoch 23, batch 0, loss[loss=0.104, simple_loss=0.1676, pruned_loss=0.02019, over 6376.00 frames. ], tot_loss[loss=0.104, simple_loss=0.1676, pruned_loss=0.02019, over 6376.00 frames. ], batch size: 27, lr: 7.26e-03, grad_scale: 8.0 +2023-03-21 02:29:20,242 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 02:29:36,392 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5835, 2.7690, 2.6724, 2.9111, 2.7113, 2.5831, 2.9594, 2.6761], + device='cuda:0'), covar=tensor([0.0790, 0.0712, 0.0655, 0.0647, 0.0718, 0.0494, 0.0632, 0.1085], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0050, 0.0057, 0.0051, 0.0049, 0.0051, 0.0051, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:29:45,949 INFO [train.py:935] (0/2) Epoch 23, validation: loss=0.1647, simple_loss=0.2529, pruned_loss=0.03826, over 1622729.00 frames. +2023-03-21 02:29:45,950 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 02:29:53,088 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 02:29:58,627 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-03-21 02:29:59,015 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.86 vs. limit=5.0 +2023-03-21 02:30:02,346 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8653, 3.8853, 3.2207, 3.4498, 2.8361, 2.3484, 1.8413, 3.9124], + device='cuda:0'), covar=tensor([0.0050, 0.0050, 0.0088, 0.0067, 0.0150, 0.0469, 0.0574, 0.0051], + device='cuda:0'), in_proj_covar=tensor([0.0097, 0.0077, 0.0100, 0.0085, 0.0112, 0.0123, 0.0123, 0.0090], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 02:30:04,928 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 02:30:11,892 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 02:30:12,222 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.80 vs. limit=2.0 +2023-03-21 02:30:12,387 INFO [train.py:901] (0/2) Epoch 23, batch 50, loss[loss=0.1201, simple_loss=0.1899, pruned_loss=0.02515, over 6973.00 frames. ], tot_loss[loss=0.1439, simple_loss=0.2223, pruned_loss=0.03274, over 324111.61 frames. ], batch size: 35, lr: 7.26e-03, grad_scale: 8.0 +2023-03-21 02:30:13,884 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 02:30:14,506 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=62183.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:30:16,300 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 02:30:32,101 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 02:30:33,099 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 02:30:34,683 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4263, 3.7695, 3.9732, 4.0852, 4.0055, 3.8766, 4.3632, 3.7363], + device='cuda:0'), covar=tensor([0.0111, 0.0187, 0.0126, 0.0120, 0.0402, 0.0135, 0.0111, 0.0187], + device='cuda:0'), in_proj_covar=tensor([0.0080, 0.0088, 0.0085, 0.0075, 0.0153, 0.0096, 0.0091, 0.0094], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:30:35,163 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2016, 3.6689, 3.7856, 3.8737, 3.7797, 3.6895, 4.1210, 3.6005], + device='cuda:0'), covar=tensor([0.0117, 0.0173, 0.0132, 0.0132, 0.0393, 0.0132, 0.0109, 0.0175], + device='cuda:0'), in_proj_covar=tensor([0.0080, 0.0088, 0.0085, 0.0075, 0.0153, 0.0096, 0.0091, 0.0094], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:30:36,059 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.420e+02 2.038e+02 2.389e+02 2.944e+02 6.721e+02, threshold=4.778e+02, percent-clipped=2.0 +2023-03-21 02:30:37,554 INFO [train.py:901] (0/2) Epoch 23, batch 100, loss[loss=0.1562, simple_loss=0.2275, pruned_loss=0.04248, over 7310.00 frames. ], tot_loss[loss=0.1471, simple_loss=0.2261, pruned_loss=0.03406, over 574133.51 frames. ], batch size: 49, lr: 7.26e-03, grad_scale: 8.0 +2023-03-21 02:30:59,314 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=62270.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:31:00,859 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=62273.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:31:03,783 INFO [train.py:901] (0/2) Epoch 23, batch 150, loss[loss=0.136, simple_loss=0.2128, pruned_loss=0.02966, over 7279.00 frames. ], tot_loss[loss=0.147, simple_loss=0.2257, pruned_loss=0.03417, over 766858.79 frames. ], batch size: 70, lr: 7.26e-03, grad_scale: 8.0 +2023-03-21 02:31:28,323 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.231e+02 1.982e+02 2.401e+02 2.805e+02 5.339e+02, threshold=4.801e+02, percent-clipped=1.0 +2023-03-21 02:31:29,902 INFO [train.py:901] (0/2) Epoch 23, batch 200, loss[loss=0.157, simple_loss=0.2399, pruned_loss=0.03701, over 7306.00 frames. ], tot_loss[loss=0.1475, simple_loss=0.2258, pruned_loss=0.03462, over 915756.20 frames. ], batch size: 86, lr: 7.25e-03, grad_scale: 8.0 +2023-03-21 02:31:33,206 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=62334.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:31:33,553 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 02:31:38,069 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 02:31:44,038 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 02:31:44,659 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3628, 2.2862, 2.1649, 3.3150, 1.5527, 3.3945, 1.3261, 2.7985], + device='cuda:0'), covar=tensor([0.0108, 0.0988, 0.1520, 0.0118, 0.3689, 0.0151, 0.1068, 0.0328], + device='cuda:0'), in_proj_covar=tensor([0.0192, 0.0261, 0.0276, 0.0185, 0.0266, 0.0194, 0.0252, 0.0228], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:31:55,512 INFO [train.py:901] (0/2) Epoch 23, batch 250, loss[loss=0.135, simple_loss=0.2141, pruned_loss=0.02796, over 7352.00 frames. ], tot_loss[loss=0.1462, simple_loss=0.2244, pruned_loss=0.03404, over 1033199.14 frames. ], batch size: 44, lr: 7.25e-03, grad_scale: 8.0 +2023-03-21 02:31:57,036 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 02:32:00,101 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=62388.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:32:01,643 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=62391.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:32:05,659 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=62399.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:32:18,548 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 02:32:20,562 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.342e+02 1.900e+02 2.217e+02 2.652e+02 4.046e+02, threshold=4.434e+02, percent-clipped=0.0 +2023-03-21 02:32:21,208 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5279, 2.8496, 2.3949, 2.8166, 2.7956, 2.2945, 2.8794, 2.5759], + device='cuda:0'), covar=tensor([0.1148, 0.1107, 0.1109, 0.0938, 0.1040, 0.1122, 0.0859, 0.0987], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0049, 0.0057, 0.0050, 0.0049, 0.0050, 0.0050, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:32:22,052 INFO [train.py:901] (0/2) Epoch 23, batch 300, loss[loss=0.1451, simple_loss=0.2263, pruned_loss=0.03199, over 7305.00 frames. ], tot_loss[loss=0.1464, simple_loss=0.2251, pruned_loss=0.03389, over 1127021.68 frames. ], batch size: 80, lr: 7.25e-03, grad_scale: 8.0 +2023-03-21 02:32:25,528 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=62436.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:32:27,068 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 02:32:30,667 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2404, 3.0424, 2.4722, 3.7265, 2.8075, 3.1455, 1.6638, 2.2397], + device='cuda:0'), covar=tensor([0.0364, 0.0782, 0.1988, 0.0435, 0.0379, 0.0669, 0.2955, 0.1873], + device='cuda:0'), in_proj_covar=tensor([0.0269, 0.0252, 0.0289, 0.0264, 0.0269, 0.0264, 0.0253, 0.0277], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 02:32:31,075 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=62447.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:32:32,177 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7299, 3.6874, 2.8721, 3.2388, 2.5970, 2.1681, 1.8651, 3.7931], + device='cuda:0'), covar=tensor([0.0045, 0.0044, 0.0131, 0.0076, 0.0140, 0.0475, 0.0505, 0.0038], + device='cuda:0'), in_proj_covar=tensor([0.0097, 0.0077, 0.0100, 0.0086, 0.0112, 0.0124, 0.0124, 0.0091], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 02:32:33,694 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=62452.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:32:36,534 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.58 vs. limit=2.0 +2023-03-21 02:32:46,609 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=62478.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:32:47,039 INFO [train.py:901] (0/2) Epoch 23, batch 350, loss[loss=0.1528, simple_loss=0.2303, pruned_loss=0.03766, over 7350.00 frames. ], tot_loss[loss=0.1471, simple_loss=0.2262, pruned_loss=0.034, over 1197698.84 frames. ], batch size: 63, lr: 7.24e-03, grad_scale: 8.0 +2023-03-21 02:32:59,304 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6624, 1.9234, 1.9333, 1.9074, 2.1231, 1.8158, 1.7274, 1.4959], + device='cuda:0'), covar=tensor([0.0329, 0.0255, 0.0118, 0.0171, 0.0331, 0.0256, 0.0351, 0.0259], + device='cuda:0'), in_proj_covar=tensor([0.0029, 0.0028, 0.0027, 0.0027, 0.0028, 0.0026, 0.0030, 0.0030], + device='cuda:0'), out_proj_covar=tensor([7.3878e-05, 7.3516e-05, 6.9459e-05, 6.8283e-05, 7.2103e-05, 6.8717e-05, + 7.5191e-05, 7.6978e-05], device='cuda:0') +2023-03-21 02:33:00,204 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 02:33:06,478 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-03-21 02:33:08,714 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7972, 3.9010, 3.7876, 3.9151, 3.6182, 3.9859, 4.2591, 4.3258], + device='cuda:0'), covar=tensor([0.0228, 0.0177, 0.0235, 0.0186, 0.0340, 0.0245, 0.0196, 0.0133], + device='cuda:0'), in_proj_covar=tensor([0.0112, 0.0114, 0.0104, 0.0109, 0.0104, 0.0094, 0.0092, 0.0087], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:33:11,593 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.318e+02 1.830e+02 2.167e+02 2.718e+02 5.157e+02, threshold=4.334e+02, percent-clipped=1.0 +2023-03-21 02:33:13,134 INFO [train.py:901] (0/2) Epoch 23, batch 400, loss[loss=0.1477, simple_loss=0.2294, pruned_loss=0.03301, over 7303.00 frames. ], tot_loss[loss=0.1478, simple_loss=0.2267, pruned_loss=0.03445, over 1251585.81 frames. ], batch size: 68, lr: 7.24e-03, grad_scale: 8.0 +2023-03-21 02:33:21,335 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=62545.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:33:33,726 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=62570.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:33:38,740 INFO [train.py:901] (0/2) Epoch 23, batch 450, loss[loss=0.1183, simple_loss=0.1901, pruned_loss=0.02329, over 7039.00 frames. ], tot_loss[loss=0.1473, simple_loss=0.2261, pruned_loss=0.0342, over 1294044.42 frames. ], batch size: 35, lr: 7.24e-03, grad_scale: 8.0 +2023-03-21 02:33:41,218 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 02:33:41,718 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 02:33:41,808 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=62585.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:33:53,388 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=62606.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:33:59,502 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=62618.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:34:03,460 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.382e+02 1.923e+02 2.343e+02 2.764e+02 5.290e+02, threshold=4.685e+02, percent-clipped=5.0 +2023-03-21 02:34:04,973 INFO [train.py:901] (0/2) Epoch 23, batch 500, loss[loss=0.1517, simple_loss=0.2282, pruned_loss=0.03755, over 7261.00 frames. ], tot_loss[loss=0.1462, simple_loss=0.2247, pruned_loss=0.03381, over 1326598.31 frames. ], batch size: 64, lr: 7.24e-03, grad_scale: 8.0 +2023-03-21 02:34:05,039 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=62629.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:34:13,884 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=62646.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:34:16,276 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 02:34:16,384 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6379, 2.2810, 2.7933, 2.7654, 2.7630, 2.3308, 2.0580, 2.5970], + device='cuda:0'), covar=tensor([0.1747, 0.1183, 0.1215, 0.1284, 0.0950, 0.1377, 0.2900, 0.1770], + device='cuda:0'), in_proj_covar=tensor([0.0046, 0.0056, 0.0043, 0.0043, 0.0043, 0.0040, 0.0058, 0.0046], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:34:17,800 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 02:34:18,307 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 02:34:20,309 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 02:34:24,531 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=62666.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:34:24,863 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 02:34:31,467 INFO [train.py:901] (0/2) Epoch 23, batch 550, loss[loss=0.1556, simple_loss=0.2291, pruned_loss=0.04107, over 7213.00 frames. ], tot_loss[loss=0.1461, simple_loss=0.2248, pruned_loss=0.03375, over 1352685.59 frames. ], batch size: 50, lr: 7.23e-03, grad_scale: 8.0 +2023-03-21 02:34:32,738 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.32 vs. limit=2.0 +2023-03-21 02:34:37,037 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 02:34:45,196 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 02:34:47,356 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3916, 3.4265, 2.5702, 4.0274, 3.1587, 3.3665, 1.8556, 2.4227], + device='cuda:0'), covar=tensor([0.0376, 0.0745, 0.2140, 0.0377, 0.0422, 0.0558, 0.2840, 0.1878], + device='cuda:0'), in_proj_covar=tensor([0.0268, 0.0253, 0.0291, 0.0264, 0.0268, 0.0264, 0.0253, 0.0275], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0004], + device='cuda:0') +2023-03-21 02:34:49,162 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 02:34:55,040 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.352e+02 2.048e+02 2.314e+02 2.629e+02 3.954e+02, threshold=4.629e+02, percent-clipped=0.0 +2023-03-21 02:34:55,747 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=62727.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:34:56,116 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 02:34:56,638 INFO [train.py:901] (0/2) Epoch 23, batch 600, loss[loss=0.1165, simple_loss=0.1789, pruned_loss=0.02708, over 6215.00 frames. ], tot_loss[loss=0.1449, simple_loss=0.2231, pruned_loss=0.03339, over 1369464.77 frames. ], batch size: 27, lr: 7.23e-03, grad_scale: 8.0 +2023-03-21 02:35:05,750 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=62747.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:35:13,815 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 02:35:17,470 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2666, 4.2863, 3.6564, 3.7388, 3.6032, 2.6326, 1.9196, 4.2819], + device='cuda:0'), covar=tensor([0.0039, 0.0029, 0.0073, 0.0060, 0.0076, 0.0375, 0.0517, 0.0043], + device='cuda:0'), in_proj_covar=tensor([0.0095, 0.0076, 0.0097, 0.0084, 0.0110, 0.0122, 0.0121, 0.0090], + device='cuda:0'), out_proj_covar=tensor([1.1218e-04, 9.8773e-05, 1.2154e-04, 1.0919e-04, 1.3472e-04, 1.5145e-04, + 1.5292e-04, 1.0831e-04], device='cuda:0') +2023-03-21 02:35:20,005 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=62773.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:35:22,440 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=62778.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:35:22,851 INFO [train.py:901] (0/2) Epoch 23, batch 650, loss[loss=0.1502, simple_loss=0.2361, pruned_loss=0.03217, over 7286.00 frames. ], tot_loss[loss=0.1454, simple_loss=0.2236, pruned_loss=0.03357, over 1386283.62 frames. ], batch size: 77, lr: 7.23e-03, grad_scale: 8.0 +2023-03-21 02:35:23,896 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 02:35:36,814 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1662, 3.6308, 3.7816, 3.8607, 3.7014, 3.6897, 4.0261, 3.6323], + device='cuda:0'), covar=tensor([0.0136, 0.0197, 0.0138, 0.0148, 0.0462, 0.0143, 0.0167, 0.0176], + device='cuda:0'), in_proj_covar=tensor([0.0084, 0.0090, 0.0089, 0.0077, 0.0159, 0.0099, 0.0095, 0.0098], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:35:40,840 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 02:35:46,853 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.383e+02 1.878e+02 2.197e+02 2.481e+02 4.239e+02, threshold=4.395e+02, percent-clipped=0.0 +2023-03-21 02:35:46,941 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=62826.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:35:48,390 INFO [train.py:901] (0/2) Epoch 23, batch 700, loss[loss=0.1484, simple_loss=0.2317, pruned_loss=0.03256, over 7271.00 frames. ], tot_loss[loss=0.1445, simple_loss=0.2226, pruned_loss=0.03318, over 1395476.92 frames. ], batch size: 70, lr: 7.22e-03, grad_scale: 8.0 +2023-03-21 02:35:48,907 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 02:35:51,042 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=62834.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:36:12,752 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 02:36:13,275 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 02:36:14,745 INFO [train.py:901] (0/2) Epoch 23, batch 750, loss[loss=0.1271, simple_loss=0.1999, pruned_loss=0.02713, over 7144.00 frames. ], tot_loss[loss=0.1462, simple_loss=0.2245, pruned_loss=0.03401, over 1406216.10 frames. ], batch size: 41, lr: 7.22e-03, grad_scale: 8.0 +2023-03-21 02:36:25,844 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=62901.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:36:26,744 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 02:36:30,368 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3751, 1.0863, 1.6156, 1.8353, 1.5749, 1.8348, 1.2693, 1.6656], + device='cuda:0'), covar=tensor([0.3435, 0.5534, 0.1021, 0.0892, 0.1485, 0.1564, 0.1701, 0.1999], + device='cuda:0'), in_proj_covar=tensor([0.0059, 0.0064, 0.0050, 0.0045, 0.0048, 0.0049, 0.0073, 0.0051], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:36:31,725 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 02:36:37,187 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 02:36:38,152 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.391e+02 1.862e+02 2.148e+02 2.738e+02 4.952e+02, threshold=4.297e+02, percent-clipped=2.0 +2023-03-21 02:36:38,192 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 02:36:38,816 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9221, 1.9759, 1.9392, 1.9240, 2.1419, 2.0116, 1.5908, 1.6960], + device='cuda:0'), covar=tensor([0.0312, 0.0338, 0.0395, 0.0175, 0.0708, 0.0435, 0.0498, 0.0321], + device='cuda:0'), in_proj_covar=tensor([0.0029, 0.0028, 0.0028, 0.0026, 0.0029, 0.0027, 0.0031, 0.0031], + device='cuda:0'), out_proj_covar=tensor([7.4586e-05, 7.3877e-05, 7.0568e-05, 6.8130e-05, 7.3088e-05, 6.9020e-05, + 7.5563e-05, 7.7998e-05], device='cuda:0') +2023-03-21 02:36:40,262 INFO [train.py:901] (0/2) Epoch 23, batch 800, loss[loss=0.1432, simple_loss=0.2364, pruned_loss=0.02499, over 7139.00 frames. ], tot_loss[loss=0.1458, simple_loss=0.2245, pruned_loss=0.03355, over 1415888.05 frames. ], batch size: 98, lr: 7.22e-03, grad_scale: 8.0 +2023-03-21 02:36:40,352 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=62929.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:36:45,041 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3641, 1.1355, 1.6140, 1.8076, 1.5817, 1.7877, 1.2480, 1.6383], + device='cuda:0'), covar=tensor([0.1941, 0.2860, 0.1078, 0.1045, 0.2618, 0.1579, 0.1228, 0.2267], + device='cuda:0'), in_proj_covar=tensor([0.0059, 0.0063, 0.0050, 0.0045, 0.0047, 0.0048, 0.0072, 0.0050], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:36:47,014 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=62941.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:36:49,425 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 02:37:04,840 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=62977.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:37:05,834 INFO [train.py:901] (0/2) Epoch 23, batch 850, loss[loss=0.1168, simple_loss=0.1973, pruned_loss=0.01818, over 7220.00 frames. ], tot_loss[loss=0.1463, simple_loss=0.225, pruned_loss=0.0338, over 1423854.47 frames. ], batch size: 39, lr: 7.22e-03, grad_scale: 8.0 +2023-03-21 02:37:08,316 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 02:37:08,328 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 02:37:12,759 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.68 vs. limit=2.0 +2023-03-21 02:37:14,429 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 02:37:17,823 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 02:37:29,111 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=63022.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:37:31,019 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.346e+02 2.012e+02 2.274e+02 2.694e+02 4.230e+02, threshold=4.549e+02, percent-clipped=0.0 +2023-03-21 02:37:32,558 INFO [train.py:901] (0/2) Epoch 23, batch 900, loss[loss=0.152, simple_loss=0.2341, pruned_loss=0.03493, over 7283.00 frames. ], tot_loss[loss=0.1473, simple_loss=0.2259, pruned_loss=0.03438, over 1427485.18 frames. ], batch size: 66, lr: 7.21e-03, grad_scale: 8.0 +2023-03-21 02:37:41,700 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=63047.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:37:49,555 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.1375, 4.6984, 4.5289, 5.1204, 5.0122, 5.1324, 4.6537, 4.7138], + device='cuda:0'), covar=tensor([0.0707, 0.2194, 0.2069, 0.1080, 0.0882, 0.1065, 0.0641, 0.0920], + device='cuda:0'), in_proj_covar=tensor([0.0217, 0.0333, 0.0261, 0.0262, 0.0197, 0.0324, 0.0191, 0.0234], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:37:55,507 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 02:37:57,521 INFO [train.py:901] (0/2) Epoch 23, batch 950, loss[loss=0.1485, simple_loss=0.2356, pruned_loss=0.03068, over 7302.00 frames. ], tot_loss[loss=0.1472, simple_loss=0.2258, pruned_loss=0.03429, over 1432293.03 frames. ], batch size: 75, lr: 7.21e-03, grad_scale: 8.0 +2023-03-21 02:37:59,882 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.02 vs. limit=2.0 +2023-03-21 02:38:05,558 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=63095.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:38:06,115 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9475, 4.3842, 4.3902, 4.3358, 4.3502, 3.9934, 4.4392, 4.3202], + device='cuda:0'), covar=tensor([0.0482, 0.0392, 0.0381, 0.0495, 0.0290, 0.0432, 0.0328, 0.0404], + device='cuda:0'), in_proj_covar=tensor([0.0198, 0.0227, 0.0175, 0.0171, 0.0138, 0.0207, 0.0180, 0.0135], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 02:38:19,027 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 02:38:22,018 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.310e+02 2.026e+02 2.334e+02 2.791e+02 5.736e+02, threshold=4.667e+02, percent-clipped=2.0 +2023-03-21 02:38:23,496 INFO [train.py:901] (0/2) Epoch 23, batch 1000, loss[loss=0.1507, simple_loss=0.2337, pruned_loss=0.03383, over 7279.00 frames. ], tot_loss[loss=0.1474, simple_loss=0.2257, pruned_loss=0.03455, over 1432794.71 frames. ], batch size: 57, lr: 7.21e-03, grad_scale: 8.0 +2023-03-21 02:38:23,559 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=63129.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:38:38,039 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 02:38:39,683 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7604, 2.4313, 1.6607, 2.7474, 2.4935, 2.3325, 1.9790, 2.2475], + device='cuda:0'), covar=tensor([0.1932, 0.0718, 0.3124, 0.0610, 0.0168, 0.0170, 0.0267, 0.0236], + device='cuda:0'), in_proj_covar=tensor([0.0266, 0.0233, 0.0264, 0.0261, 0.0170, 0.0170, 0.0200, 0.0212], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:38:41,577 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9515, 2.0280, 1.9174, 1.8324, 2.0519, 1.8721, 1.6941, 1.6379], + device='cuda:0'), covar=tensor([0.0437, 0.0385, 0.0284, 0.0214, 0.0543, 0.0549, 0.0385, 0.0380], + device='cuda:0'), in_proj_covar=tensor([0.0029, 0.0028, 0.0028, 0.0026, 0.0028, 0.0027, 0.0031, 0.0030], + device='cuda:0'), out_proj_covar=tensor([7.4170e-05, 7.3879e-05, 7.0440e-05, 6.7726e-05, 7.2665e-05, 6.9353e-05, + 7.5493e-05, 7.7507e-05], device='cuda:0') +2023-03-21 02:38:45,170 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0856, 2.7989, 2.7869, 2.9431, 2.6015, 2.3993, 3.0796, 2.1398], + device='cuda:0'), covar=tensor([0.0614, 0.0543, 0.0481, 0.0714, 0.0668, 0.0919, 0.0817, 0.1420], + device='cuda:0'), in_proj_covar=tensor([0.0328, 0.0332, 0.0266, 0.0359, 0.0308, 0.0302, 0.0344, 0.0285], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:38:48,455 INFO [train.py:901] (0/2) Epoch 23, batch 1050, loss[loss=0.1695, simple_loss=0.2545, pruned_loss=0.04228, over 7287.00 frames. ], tot_loss[loss=0.1476, simple_loss=0.2261, pruned_loss=0.03454, over 1432928.14 frames. ], batch size: 68, lr: 7.20e-03, grad_scale: 16.0 +2023-03-21 02:39:00,667 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 02:39:01,258 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=63201.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:39:04,614 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 02:39:11,169 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=63221.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:39:13,463 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.351e+02 1.843e+02 2.146e+02 2.547e+02 4.029e+02, threshold=4.291e+02, percent-clipped=0.0 +2023-03-21 02:39:14,992 INFO [train.py:901] (0/2) Epoch 23, batch 1100, loss[loss=0.1561, simple_loss=0.2351, pruned_loss=0.03855, over 7315.00 frames. ], tot_loss[loss=0.1475, simple_loss=0.226, pruned_loss=0.03448, over 1435309.97 frames. ], batch size: 80, lr: 7.20e-03, grad_scale: 16.0 +2023-03-21 02:39:21,046 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=63241.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:39:24,942 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=63249.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:39:33,003 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 02:39:33,486 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 02:39:38,715 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6580, 4.2527, 4.0892, 4.6567, 4.5439, 4.6413, 4.0674, 4.2224], + device='cuda:0'), covar=tensor([0.0972, 0.2420, 0.2184, 0.0969, 0.0895, 0.1177, 0.0874, 0.1018], + device='cuda:0'), in_proj_covar=tensor([0.0220, 0.0340, 0.0263, 0.0263, 0.0199, 0.0328, 0.0196, 0.0237], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:39:40,619 INFO [train.py:901] (0/2) Epoch 23, batch 1150, loss[loss=0.1528, simple_loss=0.2344, pruned_loss=0.0356, over 7349.00 frames. ], tot_loss[loss=0.1466, simple_loss=0.2254, pruned_loss=0.0339, over 1437221.97 frames. ], batch size: 54, lr: 7.20e-03, grad_scale: 8.0 +2023-03-21 02:39:42,882 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=63282.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:39:43,582 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.83 vs. limit=2.0 +2023-03-21 02:39:44,328 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7124, 4.2396, 4.1676, 4.7187, 4.6095, 4.6758, 4.1256, 4.2605], + device='cuda:0'), covar=tensor([0.0914, 0.2351, 0.2151, 0.0970, 0.0866, 0.1184, 0.0848, 0.1103], + device='cuda:0'), in_proj_covar=tensor([0.0220, 0.0340, 0.0263, 0.0263, 0.0199, 0.0328, 0.0196, 0.0237], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:39:46,375 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=63289.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:39:46,853 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 02:39:47,391 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 02:40:03,185 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=63322.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:40:05,566 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.308e+02 1.992e+02 2.320e+02 2.694e+02 4.886e+02, threshold=4.640e+02, percent-clipped=2.0 +2023-03-21 02:40:06,564 INFO [train.py:901] (0/2) Epoch 23, batch 1200, loss[loss=0.1377, simple_loss=0.2137, pruned_loss=0.03089, over 7276.00 frames. ], tot_loss[loss=0.1466, simple_loss=0.2252, pruned_loss=0.03397, over 1439051.33 frames. ], batch size: 57, lr: 7.20e-03, grad_scale: 8.0 +2023-03-21 02:40:16,857 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1088, 2.8494, 2.9860, 2.9070, 2.6868, 2.5728, 3.0682, 2.3915], + device='cuda:0'), covar=tensor([0.0410, 0.0536, 0.0501, 0.0499, 0.0479, 0.0685, 0.0507, 0.1460], + device='cuda:0'), in_proj_covar=tensor([0.0328, 0.0333, 0.0266, 0.0356, 0.0306, 0.0301, 0.0342, 0.0284], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:40:19,697 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 02:40:21,049 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.80 vs. limit=5.0 +2023-03-21 02:40:28,436 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=63370.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:40:32,884 INFO [train.py:901] (0/2) Epoch 23, batch 1250, loss[loss=0.1359, simple_loss=0.2126, pruned_loss=0.02965, over 7268.00 frames. ], tot_loss[loss=0.1465, simple_loss=0.225, pruned_loss=0.03393, over 1439875.69 frames. ], batch size: 52, lr: 7.19e-03, grad_scale: 8.0 +2023-03-21 02:40:44,316 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 02:40:48,306 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 02:40:49,759 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 02:40:53,802 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=63420.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:40:54,808 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=63422.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:40:57,167 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.231e+02 1.941e+02 2.406e+02 2.813e+02 7.014e+02, threshold=4.811e+02, percent-clipped=2.0 +2023-03-21 02:40:58,173 INFO [train.py:901] (0/2) Epoch 23, batch 1300, loss[loss=0.1317, simple_loss=0.1933, pruned_loss=0.03511, over 6085.00 frames. ], tot_loss[loss=0.1461, simple_loss=0.2246, pruned_loss=0.03383, over 1440225.12 frames. ], batch size: 26, lr: 7.19e-03, grad_scale: 8.0 +2023-03-21 02:40:58,270 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=63429.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:41:14,484 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 02:41:16,516 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 02:41:20,024 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 02:41:23,148 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2350, 1.5919, 1.2946, 1.3093, 1.4857, 1.5139, 1.1985, 1.0784], + device='cuda:0'), covar=tensor([0.0148, 0.0100, 0.0208, 0.0114, 0.0067, 0.0085, 0.0139, 0.0126], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0027, 0.0027, 0.0027, 0.0026, 0.0026, 0.0028, 0.0035], + device='cuda:0'), out_proj_covar=tensor([3.2932e-05, 3.0298e-05, 3.0900e-05, 3.0454e-05, 2.9478e-05, 2.9009e-05, + 3.2545e-05, 4.0507e-05], device='cuda:0') +2023-03-21 02:41:23,537 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=63477.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:41:24,135 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4034, 3.8959, 2.3193, 4.0319, 3.1327, 3.5182, 1.9393, 2.0634], + device='cuda:0'), covar=tensor([0.0362, 0.0715, 0.3076, 0.0400, 0.0577, 0.0698, 0.3727, 0.2552], + device='cuda:0'), in_proj_covar=tensor([0.0267, 0.0253, 0.0289, 0.0263, 0.0268, 0.0261, 0.0252, 0.0272], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:41:24,442 INFO [train.py:901] (0/2) Epoch 23, batch 1350, loss[loss=0.1381, simple_loss=0.2176, pruned_loss=0.02931, over 7262.00 frames. ], tot_loss[loss=0.1456, simple_loss=0.224, pruned_loss=0.0336, over 1441094.88 frames. ], batch size: 47, lr: 7.19e-03, grad_scale: 8.0 +2023-03-21 02:41:25,558 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=63481.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:41:26,595 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=63483.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:41:30,042 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 02:41:38,236 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8453, 1.9656, 1.9122, 1.9089, 2.1909, 2.0252, 1.8176, 1.6136], + device='cuda:0'), covar=tensor([0.0448, 0.0344, 0.0270, 0.0132, 0.0381, 0.0275, 0.0259, 0.0243], + device='cuda:0'), in_proj_covar=tensor([0.0029, 0.0028, 0.0028, 0.0026, 0.0028, 0.0027, 0.0031, 0.0031], + device='cuda:0'), out_proj_covar=tensor([7.3793e-05, 7.3596e-05, 7.0823e-05, 6.7384e-05, 7.2624e-05, 6.9372e-05, + 7.5453e-05, 7.8034e-05], device='cuda:0') +2023-03-21 02:41:48,580 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.437e+02 1.965e+02 2.363e+02 2.855e+02 5.705e+02, threshold=4.726e+02, percent-clipped=1.0 +2023-03-21 02:41:49,578 INFO [train.py:901] (0/2) Epoch 23, batch 1400, loss[loss=0.1576, simple_loss=0.2321, pruned_loss=0.04154, over 7327.00 frames. ], tot_loss[loss=0.1455, simple_loss=0.224, pruned_loss=0.03348, over 1439813.61 frames. ], batch size: 59, lr: 7.18e-03, grad_scale: 8.0 +2023-03-21 02:42:03,229 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 02:42:07,382 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3064, 3.6767, 4.2302, 4.4395, 4.2812, 4.3369, 4.2851, 4.2099], + device='cuda:0'), covar=tensor([0.0026, 0.0094, 0.0034, 0.0027, 0.0029, 0.0027, 0.0026, 0.0046], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0060, 0.0052, 0.0050, 0.0048, 0.0053, 0.0046, 0.0064], + device='cuda:0'), out_proj_covar=tensor([8.0997e-05, 1.3791e-04, 1.1145e-04, 9.8859e-05, 9.5617e-05, 1.0576e-04, + 1.0013e-04, 1.3035e-04], device='cuda:0') +2023-03-21 02:42:10,889 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=63569.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:42:14,884 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=63577.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:42:15,833 INFO [train.py:901] (0/2) Epoch 23, batch 1450, loss[loss=0.1474, simple_loss=0.2289, pruned_loss=0.03298, over 7342.00 frames. ], tot_loss[loss=0.1461, simple_loss=0.2247, pruned_loss=0.03375, over 1440213.00 frames. ], batch size: 63, lr: 7.18e-03, grad_scale: 8.0 +2023-03-21 02:42:19,373 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=63586.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:42:25,806 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 02:42:27,958 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.89 vs. limit=2.0 +2023-03-21 02:42:39,411 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0733, 2.7518, 1.8477, 3.1931, 2.9154, 3.1527, 2.4954, 2.6358], + device='cuda:0'), covar=tensor([0.2050, 0.0837, 0.3752, 0.0531, 0.0147, 0.0168, 0.0306, 0.0300], + device='cuda:0'), in_proj_covar=tensor([0.0261, 0.0232, 0.0257, 0.0260, 0.0169, 0.0169, 0.0198, 0.0209], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:42:40,429 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3085, 2.9112, 3.1869, 3.0703, 2.7774, 2.4514, 3.3289, 2.4830], + device='cuda:0'), covar=tensor([0.0411, 0.0372, 0.0419, 0.0468, 0.0529, 0.0741, 0.0661, 0.1397], + device='cuda:0'), in_proj_covar=tensor([0.0330, 0.0334, 0.0267, 0.0359, 0.0308, 0.0302, 0.0344, 0.0286], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:42:40,734 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.192e+02 1.916e+02 2.282e+02 2.775e+02 5.456e+02, threshold=4.565e+02, percent-clipped=1.0 +2023-03-21 02:42:42,330 INFO [train.py:901] (0/2) Epoch 23, batch 1500, loss[loss=0.1458, simple_loss=0.2233, pruned_loss=0.03411, over 7360.00 frames. ], tot_loss[loss=0.1467, simple_loss=0.2255, pruned_loss=0.03391, over 1441906.95 frames. ], batch size: 73, lr: 7.18e-03, grad_scale: 8.0 +2023-03-21 02:42:42,852 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 02:42:42,954 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=63630.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:42:51,555 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=63647.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:43:07,531 INFO [train.py:901] (0/2) Epoch 23, batch 1550, loss[loss=0.1468, simple_loss=0.2219, pruned_loss=0.03584, over 7278.00 frames. ], tot_loss[loss=0.1461, simple_loss=0.2254, pruned_loss=0.03341, over 1443881.50 frames. ], batch size: 57, lr: 7.18e-03, grad_scale: 8.0 +2023-03-21 02:43:07,564 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 02:43:32,997 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.430e+02 1.854e+02 2.117e+02 2.642e+02 5.198e+02, threshold=4.235e+02, percent-clipped=1.0 +2023-03-21 02:43:33,930 INFO [train.py:901] (0/2) Epoch 23, batch 1600, loss[loss=0.1664, simple_loss=0.2326, pruned_loss=0.05007, over 7268.00 frames. ], tot_loss[loss=0.1459, simple_loss=0.2251, pruned_loss=0.03335, over 1444683.31 frames. ], batch size: 52, lr: 7.17e-03, grad_scale: 8.0 +2023-03-21 02:43:40,472 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 02:43:40,993 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 02:43:44,007 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 02:43:49,187 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8382, 2.0781, 2.0744, 2.0414, 2.3161, 2.0195, 1.7236, 1.7298], + device='cuda:0'), covar=tensor([0.0468, 0.0397, 0.0215, 0.0242, 0.0454, 0.0396, 0.0372, 0.0305], + device='cuda:0'), in_proj_covar=tensor([0.0029, 0.0029, 0.0028, 0.0026, 0.0029, 0.0027, 0.0031, 0.0031], + device='cuda:0'), out_proj_covar=tensor([7.4602e-05, 7.4923e-05, 7.1549e-05, 6.8174e-05, 7.4027e-05, 7.0385e-05, + 7.6514e-05, 7.8303e-05], device='cuda:0') +2023-03-21 02:43:54,121 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 02:43:57,675 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 02:43:57,725 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=63776.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:43:58,737 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=63778.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:43:59,174 INFO [train.py:901] (0/2) Epoch 23, batch 1650, loss[loss=0.1365, simple_loss=0.2169, pruned_loss=0.02805, over 7354.00 frames. ], tot_loss[loss=0.1458, simple_loss=0.2249, pruned_loss=0.03333, over 1444045.60 frames. ], batch size: 51, lr: 7.17e-03, grad_scale: 8.0 +2023-03-21 02:44:05,670 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 02:44:16,379 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=63810.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:44:24,304 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 02:44:24,706 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.358e+02 1.919e+02 2.326e+02 2.793e+02 7.280e+02, threshold=4.653e+02, percent-clipped=3.0 +2023-03-21 02:44:25,741 INFO [train.py:901] (0/2) Epoch 23, batch 1700, loss[loss=0.149, simple_loss=0.2311, pruned_loss=0.03341, over 7343.00 frames. ], tot_loss[loss=0.1467, simple_loss=0.2257, pruned_loss=0.03382, over 1444464.46 frames. ], batch size: 54, lr: 7.17e-03, grad_scale: 8.0 +2023-03-21 02:44:28,345 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 02:44:33,946 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=63845.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:44:39,465 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 02:44:47,238 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=63871.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:44:50,237 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=63877.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:44:51,127 INFO [train.py:901] (0/2) Epoch 23, batch 1750, loss[loss=0.1371, simple_loss=0.2278, pruned_loss=0.02319, over 7342.00 frames. ], tot_loss[loss=0.1468, simple_loss=0.226, pruned_loss=0.03377, over 1446031.14 frames. ], batch size: 73, lr: 7.16e-03, grad_scale: 8.0 +2023-03-21 02:45:03,778 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 02:45:04,760 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 02:45:05,894 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=63906.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:45:08,144 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.01 vs. limit=2.0 +2023-03-21 02:45:11,066 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=63916.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:45:14,772 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.12 vs. limit=2.0 +2023-03-21 02:45:15,451 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=63925.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:45:15,468 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=63925.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:45:16,389 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.149e+02 2.046e+02 2.411e+02 2.952e+02 4.779e+02, threshold=4.822e+02, percent-clipped=2.0 +2023-03-21 02:45:17,410 INFO [train.py:901] (0/2) Epoch 23, batch 1800, loss[loss=0.1377, simple_loss=0.2267, pruned_loss=0.02437, over 7350.00 frames. ], tot_loss[loss=0.147, simple_loss=0.2263, pruned_loss=0.03391, over 1444928.56 frames. ], batch size: 54, lr: 7.16e-03, grad_scale: 8.0 +2023-03-21 02:45:24,045 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=63942.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:45:25,477 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 02:45:37,616 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 02:45:42,976 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=63977.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:45:43,830 INFO [train.py:901] (0/2) Epoch 23, batch 1850, loss[loss=0.1462, simple_loss=0.2329, pruned_loss=0.0298, over 7238.00 frames. ], tot_loss[loss=0.147, simple_loss=0.2264, pruned_loss=0.03379, over 1445398.44 frames. ], batch size: 89, lr: 7.16e-03, grad_scale: 8.0 +2023-03-21 02:45:48,607 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 02:45:54,979 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-64000.pt +2023-03-21 02:46:01,272 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.75 vs. limit=5.0 +2023-03-21 02:46:04,053 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6876, 3.8203, 3.7552, 3.8319, 3.6373, 3.9125, 4.1317, 4.2023], + device='cuda:0'), covar=tensor([0.0243, 0.0206, 0.0207, 0.0206, 0.0341, 0.0285, 0.0251, 0.0185], + device='cuda:0'), in_proj_covar=tensor([0.0114, 0.0115, 0.0107, 0.0111, 0.0105, 0.0093, 0.0094, 0.0090], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:46:08,871 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 02:46:09,924 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=64023.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:46:11,805 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.170e+02 1.904e+02 2.192e+02 2.514e+02 3.885e+02, threshold=4.384e+02, percent-clipped=0.0 +2023-03-21 02:46:12,804 INFO [train.py:901] (0/2) Epoch 23, batch 1900, loss[loss=0.132, simple_loss=0.2097, pruned_loss=0.02715, over 7270.00 frames. ], tot_loss[loss=0.1458, simple_loss=0.2249, pruned_loss=0.03334, over 1444008.41 frames. ], batch size: 57, lr: 7.16e-03, grad_scale: 8.0 +2023-03-21 02:46:15,500 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2555, 1.5218, 1.2655, 1.3636, 1.3741, 1.4079, 1.1689, 1.1580], + device='cuda:0'), covar=tensor([0.0149, 0.0102, 0.0199, 0.0106, 0.0091, 0.0139, 0.0130, 0.0120], + device='cuda:0'), in_proj_covar=tensor([0.0029, 0.0027, 0.0027, 0.0028, 0.0026, 0.0027, 0.0029, 0.0036], + device='cuda:0'), out_proj_covar=tensor([3.3826e-05, 3.0541e-05, 3.1381e-05, 3.1298e-05, 3.0179e-05, 3.0071e-05, + 3.2837e-05, 4.1401e-05], device='cuda:0') +2023-03-21 02:46:20,049 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6937, 3.7748, 2.6813, 4.2035, 3.0788, 3.5153, 2.0015, 2.4077], + device='cuda:0'), covar=tensor([0.0284, 0.0790, 0.2281, 0.0412, 0.0442, 0.1059, 0.3024, 0.1781], + device='cuda:0'), in_proj_covar=tensor([0.0270, 0.0252, 0.0294, 0.0264, 0.0269, 0.0260, 0.0253, 0.0273], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 02:46:24,093 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.68 vs. limit=5.0 +2023-03-21 02:46:26,338 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7672, 3.0948, 3.7436, 3.8145, 3.6964, 3.7269, 3.6041, 3.6475], + device='cuda:0'), covar=tensor([0.0026, 0.0098, 0.0030, 0.0027, 0.0033, 0.0030, 0.0038, 0.0046], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0057, 0.0050, 0.0047, 0.0047, 0.0051, 0.0044, 0.0062], + device='cuda:0'), out_proj_covar=tensor([7.7333e-05, 1.3173e-04, 1.0689e-04, 9.3680e-05, 9.1496e-05, 1.0213e-04, + 9.5515e-05, 1.2577e-04], device='cuda:0') +2023-03-21 02:46:34,046 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3063, 2.4267, 2.2299, 3.7880, 1.6566, 3.4173, 1.5138, 3.0937], + device='cuda:0'), covar=tensor([0.0099, 0.1080, 0.1597, 0.0112, 0.3458, 0.0151, 0.1100, 0.0215], + device='cuda:0'), in_proj_covar=tensor([0.0196, 0.0260, 0.0280, 0.0191, 0.0264, 0.0200, 0.0255, 0.0227], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:46:34,949 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 02:46:37,576 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=64076.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:46:38,596 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=64078.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:46:38,986 INFO [train.py:901] (0/2) Epoch 23, batch 1950, loss[loss=0.1364, simple_loss=0.2173, pruned_loss=0.02778, over 7323.00 frames. ], tot_loss[loss=0.1454, simple_loss=0.2245, pruned_loss=0.03312, over 1444078.17 frames. ], batch size: 49, lr: 7.15e-03, grad_scale: 8.0 +2023-03-21 02:46:41,766 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=64084.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:46:46,676 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 02:46:50,740 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6452, 5.1093, 5.1607, 5.1006, 4.8645, 4.6117, 5.1925, 4.9845], + device='cuda:0'), covar=tensor([0.0412, 0.0326, 0.0343, 0.0434, 0.0318, 0.0318, 0.0269, 0.0417], + device='cuda:0'), in_proj_covar=tensor([0.0200, 0.0227, 0.0175, 0.0175, 0.0139, 0.0209, 0.0179, 0.0136], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 02:46:51,166 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 02:46:51,630 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 02:46:56,243 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9583, 3.2448, 3.8925, 3.9422, 3.9385, 3.9054, 3.7982, 3.8254], + device='cuda:0'), covar=tensor([0.0023, 0.0096, 0.0031, 0.0025, 0.0032, 0.0027, 0.0035, 0.0047], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0058, 0.0051, 0.0048, 0.0048, 0.0052, 0.0045, 0.0062], + device='cuda:0'), out_proj_covar=tensor([7.8544e-05, 1.3301e-04, 1.0777e-04, 9.5153e-05, 9.3166e-05, 1.0330e-04, + 9.7450e-05, 1.2764e-04], device='cuda:0') +2023-03-21 02:47:01,019 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.39 vs. limit=5.0 +2023-03-21 02:47:01,744 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=64124.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:47:02,744 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=64126.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:47:03,156 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.380e+02 1.828e+02 2.267e+02 2.905e+02 5.773e+02, threshold=4.533e+02, percent-clipped=4.0 +2023-03-21 02:47:04,195 INFO [train.py:901] (0/2) Epoch 23, batch 2000, loss[loss=0.119, simple_loss=0.1919, pruned_loss=0.02299, over 6965.00 frames. ], tot_loss[loss=0.1458, simple_loss=0.2246, pruned_loss=0.03355, over 1443788.51 frames. ], batch size: 35, lr: 7.15e-03, grad_scale: 8.0 +2023-03-21 02:47:07,789 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 02:47:11,124 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.53 vs. limit=5.0 +2023-03-21 02:47:20,147 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 02:47:24,067 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.75 vs. limit=2.0 +2023-03-21 02:47:24,283 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=64166.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:47:28,105 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 02:47:30,615 INFO [train.py:901] (0/2) Epoch 23, batch 2050, loss[loss=0.1373, simple_loss=0.2241, pruned_loss=0.02525, over 7268.00 frames. ], tot_loss[loss=0.1458, simple_loss=0.2245, pruned_loss=0.0336, over 1444314.66 frames. ], batch size: 77, lr: 7.15e-03, grad_scale: 8.0 +2023-03-21 02:47:33,820 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9052, 2.3267, 2.8516, 2.8358, 2.7963, 2.5366, 1.9311, 2.6679], + device='cuda:0'), covar=tensor([0.1286, 0.1216, 0.0891, 0.1123, 0.1068, 0.1315, 0.3851, 0.1928], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0059, 0.0044, 0.0044, 0.0044, 0.0043, 0.0060, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:47:39,082 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.84 vs. limit=2.0 +2023-03-21 02:47:41,002 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3504, 4.4134, 3.8152, 3.8111, 3.6354, 2.6176, 1.8793, 4.3836], + device='cuda:0'), covar=tensor([0.0041, 0.0052, 0.0089, 0.0057, 0.0098, 0.0444, 0.0597, 0.0044], + device='cuda:0'), in_proj_covar=tensor([0.0098, 0.0078, 0.0099, 0.0085, 0.0114, 0.0125, 0.0123, 0.0091], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 02:47:42,272 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=64201.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:47:54,359 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=64225.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:47:55,249 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.315e+02 1.822e+02 2.255e+02 2.710e+02 4.829e+02, threshold=4.510e+02, percent-clipped=1.0 +2023-03-21 02:47:56,251 INFO [train.py:901] (0/2) Epoch 23, batch 2100, loss[loss=0.1689, simple_loss=0.2429, pruned_loss=0.04743, over 7303.00 frames. ], tot_loss[loss=0.1452, simple_loss=0.2241, pruned_loss=0.03312, over 1445645.85 frames. ], batch size: 59, lr: 7.15e-03, grad_scale: 8.0 +2023-03-21 02:47:56,929 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9223, 3.0758, 2.1645, 3.6438, 2.0752, 2.8943, 1.5331, 2.0106], + device='cuda:0'), covar=tensor([0.0319, 0.0904, 0.2307, 0.0565, 0.0373, 0.0587, 0.2933, 0.1854], + device='cuda:0'), in_proj_covar=tensor([0.0269, 0.0253, 0.0291, 0.0262, 0.0269, 0.0258, 0.0252, 0.0273], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:47:57,930 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=64232.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:48:03,058 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 02:48:04,143 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=64242.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:48:06,054 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 02:48:19,370 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=64272.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:48:19,874 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=64273.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:48:22,866 INFO [train.py:901] (0/2) Epoch 23, batch 2150, loss[loss=0.124, simple_loss=0.1949, pruned_loss=0.0265, over 6975.00 frames. ], tot_loss[loss=0.145, simple_loss=0.2237, pruned_loss=0.03318, over 1442606.87 frames. ], batch size: 35, lr: 7.14e-03, grad_scale: 8.0 +2023-03-21 02:48:26,276 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.87 vs. limit=2.0 +2023-03-21 02:48:28,558 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=64290.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:48:30,127 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=64293.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:48:47,383 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.192e+02 1.910e+02 2.130e+02 2.674e+02 6.186e+02, threshold=4.260e+02, percent-clipped=3.0 +2023-03-21 02:48:49,000 INFO [train.py:901] (0/2) Epoch 23, batch 2200, loss[loss=0.1319, simple_loss=0.2106, pruned_loss=0.02656, over 7347.00 frames. ], tot_loss[loss=0.1453, simple_loss=0.2238, pruned_loss=0.03341, over 1442470.01 frames. ], batch size: 44, lr: 7.14e-03, grad_scale: 8.0 +2023-03-21 02:48:52,012 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 02:48:57,105 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=64345.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:49:14,228 INFO [train.py:901] (0/2) Epoch 23, batch 2250, loss[loss=0.1452, simple_loss=0.2255, pruned_loss=0.03243, over 7352.00 frames. ], tot_loss[loss=0.1458, simple_loss=0.2243, pruned_loss=0.03361, over 1441826.01 frames. ], batch size: 73, lr: 7.14e-03, grad_scale: 8.0 +2023-03-21 02:49:14,303 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=64379.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:49:26,582 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 02:49:26,594 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 02:49:28,178 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=64406.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:49:39,663 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.160e+02 1.860e+02 2.261e+02 2.711e+02 5.624e+02, threshold=4.523e+02, percent-clipped=6.0 +2023-03-21 02:49:40,200 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 02:49:40,674 INFO [train.py:901] (0/2) Epoch 23, batch 2300, loss[loss=0.1137, simple_loss=0.189, pruned_loss=0.0192, over 7153.00 frames. ], tot_loss[loss=0.1455, simple_loss=0.2238, pruned_loss=0.03356, over 1442169.21 frames. ], batch size: 39, lr: 7.13e-03, grad_scale: 8.0 +2023-03-21 02:49:56,651 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.74 vs. limit=5.0 +2023-03-21 02:49:58,956 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=64465.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:49:59,484 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=64466.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:50:00,100 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.62 vs. limit=5.0 +2023-03-21 02:50:05,799 INFO [train.py:901] (0/2) Epoch 23, batch 2350, loss[loss=0.1588, simple_loss=0.2344, pruned_loss=0.04156, over 7346.00 frames. ], tot_loss[loss=0.1457, simple_loss=0.2241, pruned_loss=0.03361, over 1444102.83 frames. ], batch size: 54, lr: 7.13e-03, grad_scale: 8.0 +2023-03-21 02:50:16,973 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=64501.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:50:24,002 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=64514.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:50:26,504 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 02:50:30,202 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=64526.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:50:30,554 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.290e+02 1.893e+02 2.267e+02 2.617e+02 7.121e+02, threshold=4.533e+02, percent-clipped=2.0 +2023-03-21 02:50:31,570 INFO [train.py:901] (0/2) Epoch 23, batch 2400, loss[loss=0.1494, simple_loss=0.224, pruned_loss=0.03736, over 7314.00 frames. ], tot_loss[loss=0.145, simple_loss=0.2235, pruned_loss=0.03321, over 1441890.44 frames. ], batch size: 49, lr: 7.13e-03, grad_scale: 8.0 +2023-03-21 02:50:31,695 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0669, 3.4434, 4.2530, 4.1166, 4.3142, 4.2040, 4.1367, 3.9179], + device='cuda:0'), covar=tensor([0.0038, 0.0142, 0.0041, 0.0039, 0.0036, 0.0040, 0.0032, 0.0071], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0059, 0.0052, 0.0049, 0.0049, 0.0053, 0.0046, 0.0064], + device='cuda:0'), out_proj_covar=tensor([8.0600e-05, 1.3611e-04, 1.1195e-04, 9.7541e-05, 9.4540e-05, 1.0391e-04, + 9.8993e-05, 1.3089e-04], device='cuda:0') +2023-03-21 02:50:33,578 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 02:50:41,697 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=64549.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:50:43,223 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3552, 4.8593, 4.9839, 4.8605, 4.7716, 4.3691, 4.9918, 4.7887], + device='cuda:0'), covar=tensor([0.0503, 0.0386, 0.0355, 0.0478, 0.0329, 0.0368, 0.0311, 0.0435], + device='cuda:0'), in_proj_covar=tensor([0.0199, 0.0228, 0.0175, 0.0174, 0.0138, 0.0207, 0.0178, 0.0137], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 02:50:43,658 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 02:50:45,721 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 02:50:53,379 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=64572.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:50:54,420 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4374, 1.0921, 1.6873, 1.9142, 1.7304, 1.7620, 1.7183, 1.7563], + device='cuda:0'), covar=tensor([0.2582, 0.3942, 0.1124, 0.1366, 0.2604, 0.3746, 0.2528, 0.2806], + device='cuda:0'), in_proj_covar=tensor([0.0059, 0.0064, 0.0051, 0.0046, 0.0048, 0.0048, 0.0076, 0.0051], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:50:56,865 INFO [train.py:901] (0/2) Epoch 23, batch 2450, loss[loss=0.1469, simple_loss=0.2267, pruned_loss=0.03351, over 7281.00 frames. ], tot_loss[loss=0.1451, simple_loss=0.2239, pruned_loss=0.03312, over 1442307.85 frames. ], batch size: 77, lr: 7.13e-03, grad_scale: 8.0 +2023-03-21 02:51:02,797 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=64588.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:51:03,319 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=64589.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:51:10,862 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.67 vs. limit=2.0 +2023-03-21 02:51:11,606 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7429, 2.0676, 1.6713, 2.6253, 2.4028, 2.5042, 2.1954, 2.4378], + device='cuda:0'), covar=tensor([0.1598, 0.0790, 0.3013, 0.0672, 0.0182, 0.0193, 0.0253, 0.0262], + device='cuda:0'), in_proj_covar=tensor([0.0265, 0.0234, 0.0264, 0.0266, 0.0172, 0.0174, 0.0202, 0.0217], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:51:12,993 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 02:51:19,080 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=64620.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:51:22,544 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.432e+02 1.913e+02 2.327e+02 2.800e+02 7.811e+02, threshold=4.653e+02, percent-clipped=3.0 +2023-03-21 02:51:23,558 INFO [train.py:901] (0/2) Epoch 23, batch 2500, loss[loss=0.1612, simple_loss=0.2328, pruned_loss=0.04481, over 7272.00 frames. ], tot_loss[loss=0.1461, simple_loss=0.2247, pruned_loss=0.03372, over 1442590.05 frames. ], batch size: 55, lr: 7.12e-03, grad_scale: 8.0 +2023-03-21 02:51:33,242 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=64648.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:51:34,297 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=64650.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:51:39,250 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 02:51:49,994 INFO [train.py:901] (0/2) Epoch 23, batch 2550, loss[loss=0.1535, simple_loss=0.2317, pruned_loss=0.03763, over 7339.00 frames. ], tot_loss[loss=0.1459, simple_loss=0.2246, pruned_loss=0.03365, over 1443097.20 frames. ], batch size: 73, lr: 7.12e-03, grad_scale: 8.0 +2023-03-21 02:51:50,117 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=64679.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:51:56,525 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.01 vs. limit=2.0 +2023-03-21 02:52:01,095 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=64701.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:52:02,168 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=64703.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:52:05,184 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=64709.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:52:07,883 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.63 vs. limit=2.0 +2023-03-21 02:52:14,056 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.258e+02 1.943e+02 2.289e+02 2.838e+02 6.373e+02, threshold=4.577e+02, percent-clipped=4.0 +2023-03-21 02:52:14,134 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=64727.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:52:15,079 INFO [train.py:901] (0/2) Epoch 23, batch 2600, loss[loss=0.155, simple_loss=0.2406, pruned_loss=0.03473, over 7320.00 frames. ], tot_loss[loss=0.1456, simple_loss=0.2242, pruned_loss=0.03352, over 1442858.80 frames. ], batch size: 59, lr: 7.12e-03, grad_scale: 8.0 +2023-03-21 02:52:23,123 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=64745.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:52:26,601 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=64752.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:52:32,490 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=64764.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:52:34,715 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.89 vs. limit=2.0 +2023-03-21 02:52:38,857 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3523, 1.5761, 1.4288, 1.2407, 1.4446, 1.5439, 1.3602, 1.1481], + device='cuda:0'), covar=tensor([0.0098, 0.0094, 0.0135, 0.0126, 0.0097, 0.0139, 0.0127, 0.0130], + device='cuda:0'), in_proj_covar=tensor([0.0030, 0.0027, 0.0028, 0.0029, 0.0027, 0.0027, 0.0029, 0.0037], + device='cuda:0'), out_proj_covar=tensor([3.4465e-05, 3.0551e-05, 3.2228e-05, 3.2365e-05, 3.1182e-05, 3.0339e-05, + 3.3112e-05, 4.2497e-05], device='cuda:0') +2023-03-21 02:52:39,677 INFO [train.py:901] (0/2) Epoch 23, batch 2650, loss[loss=0.138, simple_loss=0.2216, pruned_loss=0.02717, over 7324.00 frames. ], tot_loss[loss=0.1465, simple_loss=0.2249, pruned_loss=0.03399, over 1444046.09 frames. ], batch size: 80, lr: 7.11e-03, grad_scale: 8.0 +2023-03-21 02:52:50,598 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0558, 2.6492, 1.8259, 3.3851, 3.1470, 3.4757, 2.8144, 2.8929], + device='cuda:0'), covar=tensor([0.1996, 0.0953, 0.3387, 0.0667, 0.0196, 0.0158, 0.0260, 0.0307], + device='cuda:0'), in_proj_covar=tensor([0.0264, 0.0233, 0.0264, 0.0268, 0.0173, 0.0173, 0.0200, 0.0217], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:52:53,585 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=64806.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:52:56,927 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=64813.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:53:01,350 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=64821.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:53:04,336 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.370e+02 2.063e+02 2.379e+02 2.895e+02 7.875e+02, threshold=4.758e+02, percent-clipped=6.0 +2023-03-21 02:53:05,365 INFO [train.py:901] (0/2) Epoch 23, batch 2700, loss[loss=0.1552, simple_loss=0.2343, pruned_loss=0.03807, over 7215.00 frames. ], tot_loss[loss=0.1464, simple_loss=0.225, pruned_loss=0.03391, over 1442315.99 frames. ], batch size: 93, lr: 7.11e-03, grad_scale: 8.0 +2023-03-21 02:53:11,917 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3619, 2.9639, 3.3216, 3.2681, 3.0580, 2.8960, 3.5811, 2.6870], + device='cuda:0'), covar=tensor([0.0304, 0.0367, 0.0483, 0.0477, 0.0451, 0.0681, 0.0485, 0.1497], + device='cuda:0'), in_proj_covar=tensor([0.0330, 0.0333, 0.0267, 0.0360, 0.0303, 0.0306, 0.0344, 0.0287], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:53:30,385 INFO [train.py:901] (0/2) Epoch 23, batch 2750, loss[loss=0.1556, simple_loss=0.231, pruned_loss=0.04013, over 7346.00 frames. ], tot_loss[loss=0.1458, simple_loss=0.2243, pruned_loss=0.03361, over 1441762.20 frames. ], batch size: 63, lr: 7.11e-03, grad_scale: 8.0 +2023-03-21 02:53:34,953 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=64888.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:53:39,883 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6716, 4.2401, 3.9925, 4.5951, 4.3795, 4.5942, 3.9791, 4.2101], + device='cuda:0'), covar=tensor([0.0763, 0.2163, 0.2170, 0.1108, 0.0956, 0.1087, 0.0830, 0.1006], + device='cuda:0'), in_proj_covar=tensor([0.0220, 0.0339, 0.0269, 0.0267, 0.0201, 0.0329, 0.0200, 0.0240], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:53:48,436 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.86 vs. limit=2.0 +2023-03-21 02:53:54,089 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.330e+02 1.905e+02 2.247e+02 2.674e+02 6.835e+02, threshold=4.495e+02, percent-clipped=3.0 +2023-03-21 02:53:55,111 INFO [train.py:901] (0/2) Epoch 23, batch 2800, loss[loss=0.1338, simple_loss=0.2168, pruned_loss=0.02535, over 7261.00 frames. ], tot_loss[loss=0.1455, simple_loss=0.224, pruned_loss=0.03346, over 1442485.36 frames. ], batch size: 64, lr: 7.11e-03, grad_scale: 8.0 +2023-03-21 02:53:58,577 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=64936.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:54:02,967 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=64945.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:54:07,694 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-23.pt +2023-03-21 02:54:25,841 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 02:54:25,896 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 02:54:25,922 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 02:54:25,939 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 02:54:26,127 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 02:54:26,143 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 02:54:26,154 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 02:54:26,259 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 02:54:26,335 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 02:54:26,350 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 02:54:26,367 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 02:54:26,895 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 02:54:26,901 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 02:54:26,908 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 02:54:27,049 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 02:54:27,052 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 02:54:27,111 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 02:54:29,247 INFO [train.py:901] (0/2) Epoch 24, batch 0, loss[loss=0.1146, simple_loss=0.1745, pruned_loss=0.02734, over 6403.00 frames. ], tot_loss[loss=0.1146, simple_loss=0.1745, pruned_loss=0.02734, over 6403.00 frames. ], batch size: 28, lr: 6.96e-03, grad_scale: 8.0 +2023-03-21 02:54:29,248 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 02:54:44,679 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1412, 4.1341, 3.5626, 3.5823, 3.6092, 2.4388, 2.0537, 4.1752], + device='cuda:0'), covar=tensor([0.0042, 0.0047, 0.0061, 0.0057, 0.0065, 0.0503, 0.0577, 0.0041], + device='cuda:0'), in_proj_covar=tensor([0.0097, 0.0078, 0.0098, 0.0085, 0.0113, 0.0125, 0.0122, 0.0091], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 02:54:53,805 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3846, 1.0411, 1.4429, 1.7538, 1.4491, 1.7830, 1.4820, 1.7170], + device='cuda:0'), covar=tensor([0.2122, 0.2280, 0.0752, 0.1447, 0.3094, 0.1277, 0.1538, 0.1017], + device='cuda:0'), in_proj_covar=tensor([0.0058, 0.0063, 0.0050, 0.0045, 0.0047, 0.0048, 0.0075, 0.0050], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:54:54,615 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2805, 4.5298, 4.5342, 4.5302, 4.4119, 4.2345, 4.6063, 4.4081], + device='cuda:0'), covar=tensor([0.0464, 0.0376, 0.0360, 0.0467, 0.0265, 0.0303, 0.0290, 0.0380], + device='cuda:0'), in_proj_covar=tensor([0.0201, 0.0231, 0.0177, 0.0177, 0.0140, 0.0209, 0.0178, 0.0139], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 02:54:55,323 INFO [train.py:935] (0/2) Epoch 24, validation: loss=0.1649, simple_loss=0.2534, pruned_loss=0.03824, over 1622729.00 frames. +2023-03-21 02:54:55,323 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 02:55:01,951 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 02:55:07,046 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=64976.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:55:12,419 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 02:55:17,683 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.79 vs. limit=5.0 +2023-03-21 02:55:19,296 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 02:55:19,917 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=65001.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:55:20,808 INFO [train.py:901] (0/2) Epoch 24, batch 50, loss[loss=0.1453, simple_loss=0.2275, pruned_loss=0.03149, over 7284.00 frames. ], tot_loss[loss=0.1448, simple_loss=0.2244, pruned_loss=0.03261, over 327017.09 frames. ], batch size: 57, lr: 6.96e-03, grad_scale: 8.0 +2023-03-21 02:55:21,397 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=65004.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:55:21,830 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 02:55:24,348 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 02:55:33,338 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.354e+02 1.880e+02 2.128e+02 2.472e+02 9.567e+02, threshold=4.255e+02, percent-clipped=2.0 +2023-03-21 02:55:39,094 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=65037.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 02:55:41,415 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 02:55:41,908 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 02:55:44,981 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=65049.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:55:46,889 INFO [train.py:901] (0/2) Epoch 24, batch 100, loss[loss=0.1622, simple_loss=0.2319, pruned_loss=0.04622, over 7356.00 frames. ], tot_loss[loss=0.1448, simple_loss=0.2234, pruned_loss=0.03312, over 574609.41 frames. ], batch size: 73, lr: 6.95e-03, grad_scale: 8.0 +2023-03-21 02:55:48,050 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=65055.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:55:49,978 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=65059.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:56:11,599 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=65101.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:56:12,496 INFO [train.py:901] (0/2) Epoch 24, batch 150, loss[loss=0.1521, simple_loss=0.2323, pruned_loss=0.03591, over 7362.00 frames. ], tot_loss[loss=0.1449, simple_loss=0.224, pruned_loss=0.03287, over 768154.97 frames. ], batch size: 73, lr: 6.95e-03, grad_scale: 8.0 +2023-03-21 02:56:15,067 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=65108.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:56:19,193 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=65116.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:56:22,277 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=65121.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:56:25,192 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.371e+02 2.098e+02 2.403e+02 2.683e+02 5.360e+02, threshold=4.806e+02, percent-clipped=1.0 +2023-03-21 02:56:38,256 INFO [train.py:901] (0/2) Epoch 24, batch 200, loss[loss=0.1363, simple_loss=0.2226, pruned_loss=0.02499, over 7284.00 frames. ], tot_loss[loss=0.1447, simple_loss=0.2239, pruned_loss=0.03271, over 916816.62 frames. ], batch size: 66, lr: 6.95e-03, grad_scale: 8.0 +2023-03-21 02:56:42,232 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 02:56:46,206 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 02:56:46,242 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=65169.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:56:52,514 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 02:57:03,903 INFO [train.py:901] (0/2) Epoch 24, batch 250, loss[loss=0.1292, simple_loss=0.2063, pruned_loss=0.02604, over 7347.00 frames. ], tot_loss[loss=0.1444, simple_loss=0.2234, pruned_loss=0.03272, over 1033717.41 frames. ], batch size: 44, lr: 6.95e-03, grad_scale: 8.0 +2023-03-21 02:57:06,458 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 02:57:16,458 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.295e+02 1.997e+02 2.348e+02 3.029e+02 5.559e+02, threshold=4.697e+02, percent-clipped=4.0 +2023-03-21 02:57:25,741 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=65245.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:57:27,124 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 02:57:29,607 INFO [train.py:901] (0/2) Epoch 24, batch 300, loss[loss=0.1373, simple_loss=0.2189, pruned_loss=0.02789, over 7335.00 frames. ], tot_loss[loss=0.1436, simple_loss=0.2226, pruned_loss=0.03227, over 1124148.93 frames. ], batch size: 61, lr: 6.94e-03, grad_scale: 8.0 +2023-03-21 02:57:35,561 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 02:57:36,669 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4729, 2.6532, 2.4026, 2.6616, 2.6621, 2.4157, 2.5109, 2.4820], + device='cuda:0'), covar=tensor([0.0684, 0.1225, 0.1374, 0.0838, 0.0656, 0.0572, 0.1352, 0.0883], + device='cuda:0'), in_proj_covar=tensor([0.0052, 0.0050, 0.0058, 0.0051, 0.0049, 0.0051, 0.0051, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:57:46,222 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=65285.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:57:50,774 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=65293.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:57:55,777 INFO [train.py:901] (0/2) Epoch 24, batch 350, loss[loss=0.152, simple_loss=0.2289, pruned_loss=0.03757, over 7292.00 frames. ], tot_loss[loss=0.1442, simple_loss=0.2232, pruned_loss=0.03263, over 1194013.90 frames. ], batch size: 66, lr: 6.94e-03, grad_scale: 16.0 +2023-03-21 02:57:56,396 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=65304.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:58:07,851 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.327e+02 1.918e+02 2.301e+02 2.604e+02 6.154e+02, threshold=4.603e+02, percent-clipped=3.0 +2023-03-21 02:58:10,495 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=65332.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 02:58:11,907 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 02:58:17,534 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=65346.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:58:20,429 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=65352.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:58:20,875 INFO [train.py:901] (0/2) Epoch 24, batch 400, loss[loss=0.1432, simple_loss=0.2221, pruned_loss=0.03215, over 7273.00 frames. ], tot_loss[loss=0.144, simple_loss=0.2228, pruned_loss=0.03263, over 1248537.28 frames. ], batch size: 77, lr: 6.94e-03, grad_scale: 16.0 +2023-03-21 02:58:23,999 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=65359.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:58:26,971 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9610, 2.5378, 3.0737, 2.8498, 3.1487, 2.8226, 2.5219, 3.0355], + device='cuda:0'), covar=tensor([0.1606, 0.0951, 0.1179, 0.2160, 0.0870, 0.1044, 0.1854, 0.1513], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0059, 0.0044, 0.0045, 0.0044, 0.0042, 0.0060, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:58:36,489 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0005, 3.6674, 3.7172, 3.6715, 3.5835, 3.5064, 3.9315, 3.5553], + device='cuda:0'), covar=tensor([0.0114, 0.0162, 0.0128, 0.0157, 0.0420, 0.0127, 0.0141, 0.0154], + device='cuda:0'), in_proj_covar=tensor([0.0082, 0.0090, 0.0089, 0.0077, 0.0157, 0.0097, 0.0092, 0.0096], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 02:58:46,584 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=65401.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:58:47,498 INFO [train.py:901] (0/2) Epoch 24, batch 450, loss[loss=0.1435, simple_loss=0.2152, pruned_loss=0.03592, over 7326.00 frames. ], tot_loss[loss=0.1429, simple_loss=0.2213, pruned_loss=0.03224, over 1289556.79 frames. ], batch size: 44, lr: 6.94e-03, grad_scale: 16.0 +2023-03-21 02:58:48,858 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.34 vs. limit=2.0 +2023-03-21 02:58:49,537 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=65407.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:58:50,066 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=65408.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:58:51,566 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=65411.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:58:54,007 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 02:58:54,498 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 02:58:59,516 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.265e+02 1.901e+02 2.269e+02 2.647e+02 4.135e+02, threshold=4.538e+02, percent-clipped=0.0 +2023-03-21 02:59:11,153 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=65449.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:59:13,091 INFO [train.py:901] (0/2) Epoch 24, batch 500, loss[loss=0.1166, simple_loss=0.1962, pruned_loss=0.01854, over 7199.00 frames. ], tot_loss[loss=0.1436, simple_loss=0.2222, pruned_loss=0.03256, over 1322850.13 frames. ], batch size: 39, lr: 6.93e-03, grad_scale: 16.0 +2023-03-21 02:59:14,665 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=65456.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 02:59:23,356 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0666, 2.9199, 3.0338, 3.0426, 2.7661, 2.6940, 3.3138, 2.4702], + device='cuda:0'), covar=tensor([0.0361, 0.0418, 0.0495, 0.0530, 0.0485, 0.0715, 0.0514, 0.1400], + device='cuda:0'), in_proj_covar=tensor([0.0327, 0.0331, 0.0267, 0.0359, 0.0302, 0.0302, 0.0339, 0.0283], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:59:26,246 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 02:59:26,421 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3565, 3.1086, 3.6052, 3.2578, 3.2371, 3.0757, 3.5080, 2.8919], + device='cuda:0'), covar=tensor([0.0245, 0.0287, 0.0417, 0.0390, 0.0457, 0.0723, 0.0334, 0.1240], + device='cuda:0'), in_proj_covar=tensor([0.0326, 0.0331, 0.0267, 0.0359, 0.0301, 0.0302, 0.0339, 0.0283], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 02:59:27,781 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 02:59:28,278 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 02:59:30,218 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 02:59:31,782 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8891, 3.9087, 3.1018, 3.4023, 2.9161, 2.1459, 1.7857, 3.9290], + device='cuda:0'), covar=tensor([0.0041, 0.0052, 0.0120, 0.0061, 0.0122, 0.0479, 0.0564, 0.0041], + device='cuda:0'), in_proj_covar=tensor([0.0096, 0.0077, 0.0097, 0.0084, 0.0112, 0.0122, 0.0120, 0.0090], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 02:59:34,620 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 02:59:38,605 INFO [train.py:901] (0/2) Epoch 24, batch 550, loss[loss=0.1173, simple_loss=0.1778, pruned_loss=0.02838, over 5819.00 frames. ], tot_loss[loss=0.1438, simple_loss=0.2222, pruned_loss=0.03271, over 1348013.44 frames. ], batch size: 25, lr: 6.93e-03, grad_scale: 16.0 +2023-03-21 02:59:45,340 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.83 vs. limit=2.0 +2023-03-21 02:59:46,583 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 02:59:50,481 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.349e+02 1.926e+02 2.261e+02 2.748e+02 6.198e+02, threshold=4.522e+02, percent-clipped=2.0 +2023-03-21 02:59:53,180 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=65532.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 02:59:54,594 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 02:59:57,585 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 03:00:04,069 INFO [train.py:901] (0/2) Epoch 24, batch 600, loss[loss=0.1476, simple_loss=0.228, pruned_loss=0.03356, over 7373.00 frames. ], tot_loss[loss=0.1438, simple_loss=0.2223, pruned_loss=0.0326, over 1370144.52 frames. ], batch size: 65, lr: 6.93e-03, grad_scale: 16.0 +2023-03-21 03:00:04,244 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2223, 3.3101, 2.3573, 3.7168, 2.6014, 3.1033, 1.7282, 2.1733], + device='cuda:0'), covar=tensor([0.0356, 0.0522, 0.2192, 0.0467, 0.0419, 0.0591, 0.2977, 0.1696], + device='cuda:0'), in_proj_covar=tensor([0.0270, 0.0251, 0.0288, 0.0260, 0.0265, 0.0259, 0.0252, 0.0273], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:00:05,593 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 03:00:07,414 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.31 vs. limit=2.0 +2023-03-21 03:00:22,166 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 03:00:24,858 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=65593.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 03:00:30,018 INFO [train.py:901] (0/2) Epoch 24, batch 650, loss[loss=0.1504, simple_loss=0.2351, pruned_loss=0.03283, over 7274.00 frames. ], tot_loss[loss=0.1438, simple_loss=0.2226, pruned_loss=0.03249, over 1385655.62 frames. ], batch size: 70, lr: 6.93e-03, grad_scale: 16.0 +2023-03-21 03:00:30,564 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 03:00:42,615 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.149e+02 1.867e+02 2.152e+02 2.439e+02 4.019e+02, threshold=4.303e+02, percent-clipped=0.0 +2023-03-21 03:00:45,281 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=65632.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:00:45,840 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3926, 0.9895, 1.4799, 1.8512, 1.4095, 1.7217, 1.3147, 1.6111], + device='cuda:0'), covar=tensor([0.3036, 0.5079, 0.0889, 0.0785, 0.2644, 0.2047, 0.1527, 0.3034], + device='cuda:0'), in_proj_covar=tensor([0.0055, 0.0062, 0.0048, 0.0045, 0.0048, 0.0046, 0.0073, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:00:48,658 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 03:00:49,735 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=65641.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:00:56,271 INFO [train.py:901] (0/2) Epoch 24, batch 700, loss[loss=0.1261, simple_loss=0.2073, pruned_loss=0.02239, over 7280.00 frames. ], tot_loss[loss=0.1436, simple_loss=0.2224, pruned_loss=0.03243, over 1398772.79 frames. ], batch size: 66, lr: 6.92e-03, grad_scale: 16.0 +2023-03-21 03:00:56,786 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 03:00:57,351 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.7336, 5.2651, 5.1226, 5.6466, 5.4410, 5.6360, 5.1696, 5.2532], + device='cuda:0'), covar=tensor([0.0571, 0.1815, 0.1722, 0.0720, 0.0872, 0.0890, 0.0570, 0.0922], + device='cuda:0'), in_proj_covar=tensor([0.0217, 0.0341, 0.0269, 0.0266, 0.0200, 0.0326, 0.0197, 0.0241], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:01:09,876 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=65680.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:01:19,709 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 03:01:19,724 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 03:01:20,847 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4091, 4.9102, 4.9868, 4.8950, 4.7684, 4.4693, 4.9894, 4.8415], + device='cuda:0'), covar=tensor([0.0436, 0.0373, 0.0312, 0.0408, 0.0303, 0.0364, 0.0311, 0.0432], + device='cuda:0'), in_proj_covar=tensor([0.0202, 0.0234, 0.0178, 0.0177, 0.0140, 0.0211, 0.0182, 0.0141], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 03:01:21,788 INFO [train.py:901] (0/2) Epoch 24, batch 750, loss[loss=0.147, simple_loss=0.2309, pruned_loss=0.03153, over 7254.00 frames. ], tot_loss[loss=0.1438, simple_loss=0.2223, pruned_loss=0.0326, over 1407620.24 frames. ], batch size: 64, lr: 6.92e-03, grad_scale: 16.0 +2023-03-21 03:01:25,925 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=65711.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:01:33,724 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 03:01:34,193 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.194e+02 1.950e+02 2.225e+02 2.735e+02 4.700e+02, threshold=4.450e+02, percent-clipped=3.0 +2023-03-21 03:01:39,855 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 03:01:45,328 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 03:01:46,829 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 03:01:47,312 INFO [train.py:901] (0/2) Epoch 24, batch 800, loss[loss=0.1383, simple_loss=0.2179, pruned_loss=0.02932, over 7293.00 frames. ], tot_loss[loss=0.144, simple_loss=0.2224, pruned_loss=0.03282, over 1412642.93 frames. ], batch size: 49, lr: 6.92e-03, grad_scale: 8.0 +2023-03-21 03:01:50,363 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=65759.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:01:55,687 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.77 vs. limit=2.0 +2023-03-21 03:01:57,386 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 03:01:57,932 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7080, 5.2543, 5.3543, 5.2470, 5.0517, 4.7931, 5.3723, 5.1796], + device='cuda:0'), covar=tensor([0.0448, 0.0365, 0.0318, 0.0429, 0.0279, 0.0320, 0.0260, 0.0368], + device='cuda:0'), in_proj_covar=tensor([0.0203, 0.0235, 0.0179, 0.0178, 0.0140, 0.0212, 0.0182, 0.0142], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 03:02:13,258 INFO [train.py:901] (0/2) Epoch 24, batch 850, loss[loss=0.1514, simple_loss=0.2298, pruned_loss=0.03649, over 7342.00 frames. ], tot_loss[loss=0.1444, simple_loss=0.2229, pruned_loss=0.03291, over 1420035.89 frames. ], batch size: 54, lr: 6.91e-03, grad_scale: 8.0 +2023-03-21 03:02:15,681 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 03:02:15,685 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 03:02:21,832 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 03:02:24,880 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 03:02:26,328 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.275e+02 1.907e+02 2.307e+02 2.864e+02 5.417e+02, threshold=4.614e+02, percent-clipped=5.0 +2023-03-21 03:02:31,532 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=65838.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 03:02:38,825 INFO [train.py:901] (0/2) Epoch 24, batch 900, loss[loss=0.1331, simple_loss=0.2134, pruned_loss=0.02638, over 7294.00 frames. ], tot_loss[loss=0.1444, simple_loss=0.223, pruned_loss=0.03285, over 1426921.86 frames. ], batch size: 77, lr: 6.91e-03, grad_scale: 8.0 +2023-03-21 03:02:44,409 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.78 vs. limit=5.0 +2023-03-21 03:02:47,781 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3839, 1.6452, 1.5713, 1.6209, 1.6373, 1.5827, 1.4523, 1.2724], + device='cuda:0'), covar=tensor([0.0094, 0.0114, 0.0194, 0.0091, 0.0073, 0.0080, 0.0142, 0.0137], + device='cuda:0'), in_proj_covar=tensor([0.0031, 0.0028, 0.0028, 0.0030, 0.0028, 0.0028, 0.0030, 0.0038], + device='cuda:0'), out_proj_covar=tensor([3.5283e-05, 3.1772e-05, 3.2450e-05, 3.3613e-05, 3.2041e-05, 3.0577e-05, + 3.3741e-05, 4.2903e-05], device='cuda:0') +2023-03-21 03:02:56,794 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=65888.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 03:03:02,284 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 03:03:02,394 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=65899.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 03:03:04,862 INFO [train.py:901] (0/2) Epoch 24, batch 950, loss[loss=0.1146, simple_loss=0.1897, pruned_loss=0.01974, over 7155.00 frames. ], tot_loss[loss=0.1447, simple_loss=0.2235, pruned_loss=0.03293, over 1431680.23 frames. ], batch size: 39, lr: 6.91e-03, grad_scale: 8.0 +2023-03-21 03:03:17,336 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.335e+02 1.871e+02 2.193e+02 2.615e+02 7.204e+02, threshold=4.387e+02, percent-clipped=1.0 +2023-03-21 03:03:24,015 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=65941.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:03:25,945 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 03:03:29,899 INFO [train.py:901] (0/2) Epoch 24, batch 1000, loss[loss=0.1328, simple_loss=0.2261, pruned_loss=0.01975, over 6781.00 frames. ], tot_loss[loss=0.1441, simple_loss=0.223, pruned_loss=0.03257, over 1433400.08 frames. ], batch size: 107, lr: 6.91e-03, grad_scale: 8.0 +2023-03-21 03:03:47,310 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 03:03:48,852 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=65989.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:03:56,626 INFO [train.py:901] (0/2) Epoch 24, batch 1050, loss[loss=0.1283, simple_loss=0.2033, pruned_loss=0.02666, over 7231.00 frames. ], tot_loss[loss=0.144, simple_loss=0.2225, pruned_loss=0.03276, over 1434951.48 frames. ], batch size: 55, lr: 6.90e-03, grad_scale: 8.0 +2023-03-21 03:04:09,037 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.224e+02 1.813e+02 2.192e+02 2.509e+02 4.025e+02, threshold=4.384e+02, percent-clipped=0.0 +2023-03-21 03:04:09,582 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 03:04:13,246 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6931, 2.8020, 2.3045, 3.9233, 1.8768, 3.6133, 1.5138, 2.9893], + device='cuda:0'), covar=tensor([0.0125, 0.0921, 0.1416, 0.0123, 0.3311, 0.0182, 0.1020, 0.0290], + device='cuda:0'), in_proj_covar=tensor([0.0197, 0.0257, 0.0275, 0.0190, 0.0265, 0.0201, 0.0251, 0.0228], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:04:13,634 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 03:04:15,721 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1564, 3.8400, 3.8402, 3.7497, 3.7135, 3.7175, 4.0673, 3.5143], + device='cuda:0'), covar=tensor([0.0118, 0.0138, 0.0125, 0.0156, 0.0404, 0.0106, 0.0124, 0.0209], + device='cuda:0'), in_proj_covar=tensor([0.0082, 0.0089, 0.0091, 0.0078, 0.0156, 0.0097, 0.0093, 0.0097], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:04:22,113 INFO [train.py:901] (0/2) Epoch 24, batch 1100, loss[loss=0.1884, simple_loss=0.2514, pruned_loss=0.06268, over 7220.00 frames. ], tot_loss[loss=0.1446, simple_loss=0.2232, pruned_loss=0.03296, over 1437460.39 frames. ], batch size: 93, lr: 6.90e-03, grad_scale: 8.0 +2023-03-21 03:04:30,850 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.88 vs. limit=5.0 +2023-03-21 03:04:31,359 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-03-21 03:04:39,520 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.91 vs. limit=5.0 +2023-03-21 03:04:42,745 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 03:04:43,230 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 03:04:47,666 INFO [train.py:901] (0/2) Epoch 24, batch 1150, loss[loss=0.1566, simple_loss=0.2337, pruned_loss=0.03974, over 7348.00 frames. ], tot_loss[loss=0.1442, simple_loss=0.2227, pruned_loss=0.03289, over 1434672.22 frames. ], batch size: 63, lr: 6.90e-03, grad_scale: 8.0 +2023-03-21 03:04:55,183 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 03:04:56,173 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 03:05:00,167 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.297e+02 1.914e+02 2.178e+02 2.666e+02 4.759e+02, threshold=4.356e+02, percent-clipped=2.0 +2023-03-21 03:05:08,988 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=66144.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:05:13,310 INFO [train.py:901] (0/2) Epoch 24, batch 1200, loss[loss=0.1549, simple_loss=0.2243, pruned_loss=0.04275, over 7216.00 frames. ], tot_loss[loss=0.1441, simple_loss=0.2229, pruned_loss=0.0327, over 1437420.49 frames. ], batch size: 45, lr: 6.90e-03, grad_scale: 8.0 +2023-03-21 03:05:16,914 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=66160.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 03:05:29,121 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 03:05:31,679 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=66188.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 03:05:34,621 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=66194.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 03:05:39,349 INFO [train.py:901] (0/2) Epoch 24, batch 1250, loss[loss=0.1554, simple_loss=0.24, pruned_loss=0.03544, over 7294.00 frames. ], tot_loss[loss=0.1445, simple_loss=0.2235, pruned_loss=0.0327, over 1441401.53 frames. ], batch size: 68, lr: 6.89e-03, grad_scale: 8.0 +2023-03-21 03:05:40,530 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=66205.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 03:05:49,148 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=66221.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 03:05:51,923 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 03:05:52,401 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.147e+02 1.872e+02 2.228e+02 2.554e+02 4.921e+02, threshold=4.456e+02, percent-clipped=2.0 +2023-03-21 03:05:55,958 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 03:05:56,499 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=66236.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 03:05:56,948 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 03:06:05,562 INFO [train.py:901] (0/2) Epoch 24, batch 1300, loss[loss=0.1588, simple_loss=0.241, pruned_loss=0.03826, over 7327.00 frames. ], tot_loss[loss=0.145, simple_loss=0.2242, pruned_loss=0.03293, over 1443310.87 frames. ], batch size: 49, lr: 6.89e-03, grad_scale: 8.0 +2023-03-21 03:06:08,157 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=66258.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:06:19,637 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 03:06:21,843 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.31 vs. limit=2.0 +2023-03-21 03:06:22,065 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 03:06:25,563 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 03:06:31,026 INFO [train.py:901] (0/2) Epoch 24, batch 1350, loss[loss=0.134, simple_loss=0.2155, pruned_loss=0.02623, over 7262.00 frames. ], tot_loss[loss=0.1449, simple_loss=0.2239, pruned_loss=0.03297, over 1442991.09 frames. ], batch size: 52, lr: 6.89e-03, grad_scale: 8.0 +2023-03-21 03:06:36,006 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 03:06:39,147 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=66319.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:06:43,442 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.222e+02 1.819e+02 2.167e+02 2.669e+02 5.155e+02, threshold=4.335e+02, percent-clipped=1.0 +2023-03-21 03:06:56,410 INFO [train.py:901] (0/2) Epoch 24, batch 1400, loss[loss=0.162, simple_loss=0.2402, pruned_loss=0.04187, over 7250.00 frames. ], tot_loss[loss=0.1448, simple_loss=0.2235, pruned_loss=0.03308, over 1443994.04 frames. ], batch size: 89, lr: 6.89e-03, grad_scale: 8.0 +2023-03-21 03:07:08,496 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 03:07:18,713 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1553, 2.8196, 3.1340, 2.8160, 2.6843, 2.5130, 3.1135, 2.3590], + device='cuda:0'), covar=tensor([0.0433, 0.0496, 0.0425, 0.0532, 0.0475, 0.0731, 0.0509, 0.1336], + device='cuda:0'), in_proj_covar=tensor([0.0332, 0.0339, 0.0268, 0.0361, 0.0302, 0.0303, 0.0342, 0.0282], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:07:22,314 INFO [train.py:901] (0/2) Epoch 24, batch 1450, loss[loss=0.123, simple_loss=0.1977, pruned_loss=0.0241, over 7172.00 frames. ], tot_loss[loss=0.1456, simple_loss=0.2242, pruned_loss=0.03346, over 1443486.23 frames. ], batch size: 39, lr: 6.88e-03, grad_scale: 8.0 +2023-03-21 03:07:32,915 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 03:07:34,803 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.376e+02 1.845e+02 2.245e+02 2.730e+02 1.405e+03, threshold=4.491e+02, percent-clipped=2.0 +2023-03-21 03:07:48,007 INFO [train.py:901] (0/2) Epoch 24, batch 1500, loss[loss=0.1343, simple_loss=0.2078, pruned_loss=0.03038, over 6972.00 frames. ], tot_loss[loss=0.1451, simple_loss=0.224, pruned_loss=0.03304, over 1445683.37 frames. ], batch size: 35, lr: 6.88e-03, grad_scale: 8.0 +2023-03-21 03:07:48,966 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 03:08:09,093 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=66494.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 03:08:12,139 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=66500.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 03:08:12,663 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0209, 3.5268, 4.1095, 3.9964, 4.1068, 4.1410, 4.0380, 3.8624], + device='cuda:0'), covar=tensor([0.0030, 0.0091, 0.0027, 0.0032, 0.0028, 0.0025, 0.0031, 0.0047], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0058, 0.0051, 0.0049, 0.0047, 0.0052, 0.0044, 0.0064], + device='cuda:0'), out_proj_covar=tensor([7.8272e-05, 1.3206e-04, 1.0720e-04, 9.6429e-05, 9.0621e-05, 1.0037e-04, + 9.6193e-05, 1.3015e-04], device='cuda:0') +2023-03-21 03:08:13,075 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 03:08:13,555 INFO [train.py:901] (0/2) Epoch 24, batch 1550, loss[loss=0.1394, simple_loss=0.2126, pruned_loss=0.03307, over 7274.00 frames. ], tot_loss[loss=0.1454, simple_loss=0.2243, pruned_loss=0.03324, over 1446691.37 frames. ], batch size: 47, lr: 6.88e-03, grad_scale: 8.0 +2023-03-21 03:08:20,773 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=66516.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 03:08:23,210 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=66521.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 03:08:26,582 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.397e+02 1.920e+02 2.387e+02 2.777e+02 7.099e+02, threshold=4.774e+02, percent-clipped=1.0 +2023-03-21 03:08:33,676 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=66542.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 03:08:36,282 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5942, 2.6468, 2.2775, 3.9457, 1.7459, 3.7160, 1.4767, 3.0598], + device='cuda:0'), covar=tensor([0.0117, 0.0967, 0.1610, 0.0142, 0.3730, 0.0147, 0.1095, 0.0240], + device='cuda:0'), in_proj_covar=tensor([0.0199, 0.0258, 0.0280, 0.0192, 0.0268, 0.0202, 0.0253, 0.0231], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:08:39,092 INFO [train.py:901] (0/2) Epoch 24, batch 1600, loss[loss=0.1623, simple_loss=0.2376, pruned_loss=0.04349, over 7247.00 frames. ], tot_loss[loss=0.1455, simple_loss=0.2245, pruned_loss=0.03325, over 1444990.24 frames. ], batch size: 89, lr: 6.88e-03, grad_scale: 8.0 +2023-03-21 03:08:44,644 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 03:08:45,202 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 03:08:47,688 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 03:08:54,313 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=66582.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 03:08:57,676 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 03:08:57,802 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4216, 4.3133, 3.9264, 3.9452, 3.8436, 2.6104, 2.2450, 4.5049], + device='cuda:0'), covar=tensor([0.0040, 0.0066, 0.0070, 0.0044, 0.0084, 0.0407, 0.0456, 0.0041], + device='cuda:0'), in_proj_covar=tensor([0.0096, 0.0077, 0.0097, 0.0082, 0.0112, 0.0120, 0.0118, 0.0090], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 03:09:01,224 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 03:09:05,745 INFO [train.py:901] (0/2) Epoch 24, batch 1650, loss[loss=0.1293, simple_loss=0.2178, pruned_loss=0.02039, over 7258.00 frames. ], tot_loss[loss=0.1453, simple_loss=0.2243, pruned_loss=0.0332, over 1445110.80 frames. ], batch size: 89, lr: 6.87e-03, grad_scale: 8.0 +2023-03-21 03:09:10,913 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 03:09:11,486 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=66614.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:09:18,405 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.348e+02 1.874e+02 2.286e+02 2.733e+02 3.925e+02, threshold=4.573e+02, percent-clipped=0.0 +2023-03-21 03:09:27,990 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 03:09:31,662 INFO [train.py:901] (0/2) Epoch 24, batch 1700, loss[loss=0.1695, simple_loss=0.2452, pruned_loss=0.04686, over 7291.00 frames. ], tot_loss[loss=0.1446, simple_loss=0.2236, pruned_loss=0.03286, over 1440462.22 frames. ], batch size: 86, lr: 6.87e-03, grad_scale: 8.0 +2023-03-21 03:09:33,037 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.64 vs. limit=2.0 +2023-03-21 03:09:33,229 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 03:09:43,370 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 03:09:57,421 INFO [train.py:901] (0/2) Epoch 24, batch 1750, loss[loss=0.1619, simple_loss=0.238, pruned_loss=0.04293, over 7293.00 frames. ], tot_loss[loss=0.1446, simple_loss=0.2235, pruned_loss=0.03281, over 1440490.04 frames. ], batch size: 66, lr: 6.87e-03, grad_scale: 8.0 +2023-03-21 03:10:06,937 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 03:10:08,396 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 03:10:09,894 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.269e+02 1.848e+02 2.142e+02 2.647e+02 7.442e+02, threshold=4.285e+02, percent-clipped=1.0 +2023-03-21 03:10:12,222 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=66731.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 03:10:17,112 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.72 vs. limit=5.0 +2023-03-21 03:10:23,248 INFO [train.py:901] (0/2) Epoch 24, batch 1800, loss[loss=0.1481, simple_loss=0.227, pruned_loss=0.03456, over 7332.00 frames. ], tot_loss[loss=0.1447, simple_loss=0.224, pruned_loss=0.03266, over 1443370.36 frames. ], batch size: 75, lr: 6.87e-03, grad_scale: 8.0 +2023-03-21 03:10:28,851 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1334, 4.0890, 3.5568, 3.6309, 3.7534, 2.4307, 1.9540, 4.2519], + device='cuda:0'), covar=tensor([0.0076, 0.0075, 0.0126, 0.0087, 0.0120, 0.0541, 0.0634, 0.0063], + device='cuda:0'), in_proj_covar=tensor([0.0098, 0.0080, 0.0100, 0.0085, 0.0115, 0.0124, 0.0122, 0.0092], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 03:10:30,768 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 03:10:43,466 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=66792.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 03:10:44,879 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 03:10:47,765 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=66800.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 03:10:49,126 INFO [train.py:901] (0/2) Epoch 24, batch 1850, loss[loss=0.1196, simple_loss=0.2016, pruned_loss=0.0188, over 7315.00 frames. ], tot_loss[loss=0.1444, simple_loss=0.2233, pruned_loss=0.0327, over 1441538.77 frames. ], batch size: 44, lr: 6.86e-03, grad_scale: 8.0 +2023-03-21 03:10:53,787 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8170, 3.2216, 2.6177, 2.9864, 2.8022, 2.7426, 3.0086, 2.7927], + device='cuda:0'), covar=tensor([0.0706, 0.0546, 0.1204, 0.1005, 0.1304, 0.0709, 0.0800, 0.1069], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0049, 0.0057, 0.0051, 0.0049, 0.0051, 0.0050, 0.0046], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:10:55,716 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 03:10:55,798 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=66816.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 03:11:02,372 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.480e+02 1.896e+02 2.287e+02 2.722e+02 5.764e+02, threshold=4.573e+02, percent-clipped=2.0 +2023-03-21 03:11:12,510 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=66848.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:11:13,506 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 03:11:15,006 INFO [train.py:901] (0/2) Epoch 24, batch 1900, loss[loss=0.1378, simple_loss=0.2074, pruned_loss=0.03414, over 7137.00 frames. ], tot_loss[loss=0.1435, simple_loss=0.2224, pruned_loss=0.03234, over 1440124.50 frames. ], batch size: 41, lr: 6.86e-03, grad_scale: 8.0 +2023-03-21 03:11:20,522 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=66864.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 03:11:21,668 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-03-21 03:11:27,512 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=66877.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 03:11:37,820 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 03:11:40,320 INFO [train.py:901] (0/2) Epoch 24, batch 1950, loss[loss=0.1384, simple_loss=0.2223, pruned_loss=0.02727, over 7322.00 frames. ], tot_loss[loss=0.1438, simple_loss=0.2225, pruned_loss=0.03251, over 1441229.91 frames. ], batch size: 83, lr: 6.86e-03, grad_scale: 8.0 +2023-03-21 03:11:46,476 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=66914.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:11:48,807 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 03:11:49,983 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3677, 1.0729, 1.3597, 1.7964, 1.6255, 1.6954, 1.2202, 1.7595], + device='cuda:0'), covar=tensor([0.2598, 0.4240, 0.1628, 0.1391, 0.1561, 0.1963, 0.1416, 0.2127], + device='cuda:0'), in_proj_covar=tensor([0.0058, 0.0061, 0.0051, 0.0045, 0.0048, 0.0048, 0.0074, 0.0050], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0001, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:11:53,318 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.231e+02 1.892e+02 2.217e+02 2.546e+02 5.522e+02, threshold=4.435e+02, percent-clipped=0.0 +2023-03-21 03:11:53,343 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 03:11:53,861 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 03:11:58,085 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6316, 2.9642, 2.4405, 2.8530, 2.6655, 2.4599, 2.7703, 2.6292], + device='cuda:0'), covar=tensor([0.0890, 0.0515, 0.1102, 0.0926, 0.1164, 0.1009, 0.1141, 0.1030], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0048, 0.0056, 0.0049, 0.0048, 0.0050, 0.0049, 0.0045], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:12:01,127 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4661, 3.2283, 2.2716, 3.7727, 2.7692, 3.1995, 1.7560, 2.3277], + device='cuda:0'), covar=tensor([0.0398, 0.0772, 0.2491, 0.0563, 0.0437, 0.0845, 0.3097, 0.1801], + device='cuda:0'), in_proj_covar=tensor([0.0270, 0.0250, 0.0287, 0.0262, 0.0268, 0.0258, 0.0249, 0.0271], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:12:05,971 INFO [train.py:901] (0/2) Epoch 24, batch 2000, loss[loss=0.149, simple_loss=0.2304, pruned_loss=0.03381, over 7278.00 frames. ], tot_loss[loss=0.1435, simple_loss=0.2224, pruned_loss=0.0323, over 1442674.15 frames. ], batch size: 77, lr: 6.86e-03, grad_scale: 8.0 +2023-03-21 03:12:08,190 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=66957.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:12:09,213 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9876, 3.5194, 3.9632, 3.9740, 3.9353, 4.0137, 4.0455, 3.9344], + device='cuda:0'), covar=tensor([0.0027, 0.0092, 0.0032, 0.0029, 0.0032, 0.0026, 0.0032, 0.0043], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0059, 0.0051, 0.0049, 0.0048, 0.0053, 0.0045, 0.0065], + device='cuda:0'), out_proj_covar=tensor([8.0496e-05, 1.3379e-04, 1.0728e-04, 9.7413e-05, 9.3524e-05, 1.0219e-04, + 9.8950e-05, 1.3137e-04], device='cuda:0') +2023-03-21 03:12:10,607 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 03:12:11,300 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=66962.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:12:20,465 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.79 vs. limit=5.0 +2023-03-21 03:12:21,235 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 03:12:24,322 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0678, 3.7836, 4.1659, 4.0689, 4.1639, 4.1748, 4.1704, 4.1100], + device='cuda:0'), covar=tensor([0.0031, 0.0080, 0.0027, 0.0031, 0.0028, 0.0024, 0.0027, 0.0040], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0059, 0.0051, 0.0049, 0.0048, 0.0052, 0.0045, 0.0064], + device='cuda:0'), out_proj_covar=tensor([8.0191e-05, 1.3317e-04, 1.0635e-04, 9.6293e-05, 9.2263e-05, 1.0146e-04, + 9.8101e-05, 1.3030e-04], device='cuda:0') +2023-03-21 03:12:31,300 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 03:12:32,746 INFO [train.py:901] (0/2) Epoch 24, batch 2050, loss[loss=0.1414, simple_loss=0.2088, pruned_loss=0.03703, over 7252.00 frames. ], tot_loss[loss=0.1433, simple_loss=0.2222, pruned_loss=0.03219, over 1442391.84 frames. ], batch size: 45, lr: 6.85e-03, grad_scale: 8.0 +2023-03-21 03:12:40,301 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=67018.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 03:12:43,009 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.66 vs. limit=2.0 +2023-03-21 03:12:43,966 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.03 vs. limit=2.0 +2023-03-21 03:12:45,164 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.240e+02 1.848e+02 2.104e+02 2.679e+02 4.101e+02, threshold=4.209e+02, percent-clipped=1.0 +2023-03-21 03:12:58,213 INFO [train.py:901] (0/2) Epoch 24, batch 2100, loss[loss=0.155, simple_loss=0.2308, pruned_loss=0.03959, over 7284.00 frames. ], tot_loss[loss=0.1441, simple_loss=0.2231, pruned_loss=0.03252, over 1442710.73 frames. ], batch size: 57, lr: 6.85e-03, grad_scale: 8.0 +2023-03-21 03:13:03,657 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 03:13:06,055 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 03:13:13,797 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=67083.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:13:15,760 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=67087.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 03:13:16,359 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7990, 2.6972, 3.0557, 2.9110, 3.0758, 2.7670, 2.6318, 2.7961], + device='cuda:0'), covar=tensor([0.1930, 0.0892, 0.1282, 0.1604, 0.1394, 0.1153, 0.2079, 0.2281], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0059, 0.0045, 0.0044, 0.0044, 0.0042, 0.0060, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:13:23,615 INFO [train.py:901] (0/2) Epoch 24, batch 2150, loss[loss=0.1673, simple_loss=0.2457, pruned_loss=0.04441, over 7357.00 frames. ], tot_loss[loss=0.145, simple_loss=0.224, pruned_loss=0.033, over 1445263.31 frames. ], batch size: 63, lr: 6.85e-03, grad_scale: 8.0 +2023-03-21 03:13:26,926 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.79 vs. limit=5.0 +2023-03-21 03:13:36,126 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.281e+02 1.954e+02 2.289e+02 2.688e+02 5.693e+02, threshold=4.578e+02, percent-clipped=4.0 +2023-03-21 03:13:37,253 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8073, 3.3417, 2.8662, 3.3166, 3.1851, 2.8657, 3.3404, 3.1784], + device='cuda:0'), covar=tensor([0.0877, 0.0593, 0.1215, 0.0981, 0.1310, 0.0688, 0.0584, 0.0637], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0049, 0.0056, 0.0050, 0.0048, 0.0050, 0.0049, 0.0046], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:13:41,401 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5822, 1.2533, 1.6744, 2.0678, 1.8305, 1.9832, 1.5964, 1.9093], + device='cuda:0'), covar=tensor([0.1437, 0.3313, 0.1224, 0.0818, 0.1602, 0.1609, 0.1175, 0.2619], + device='cuda:0'), in_proj_covar=tensor([0.0058, 0.0063, 0.0052, 0.0046, 0.0048, 0.0049, 0.0075, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:13:44,957 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=67144.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:13:49,318 INFO [train.py:901] (0/2) Epoch 24, batch 2200, loss[loss=0.1188, simple_loss=0.1964, pruned_loss=0.02056, over 7137.00 frames. ], tot_loss[loss=0.1451, simple_loss=0.2242, pruned_loss=0.03306, over 1444576.81 frames. ], batch size: 41, lr: 6.85e-03, grad_scale: 8.0 +2023-03-21 03:13:51,369 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 03:14:02,154 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=67177.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 03:14:15,374 INFO [train.py:901] (0/2) Epoch 24, batch 2250, loss[loss=0.1274, simple_loss=0.2151, pruned_loss=0.0198, over 7264.00 frames. ], tot_loss[loss=0.1462, simple_loss=0.2252, pruned_loss=0.03359, over 1444335.73 frames. ], batch size: 52, lr: 6.84e-03, grad_scale: 8.0 +2023-03-21 03:14:21,880 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.1189, 4.7099, 4.5780, 5.0995, 4.9988, 5.0701, 4.3482, 4.7694], + device='cuda:0'), covar=tensor([0.0747, 0.2054, 0.2043, 0.0950, 0.0741, 0.1055, 0.0770, 0.1032], + device='cuda:0'), in_proj_covar=tensor([0.0224, 0.0346, 0.0271, 0.0268, 0.0199, 0.0333, 0.0202, 0.0247], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:14:26,006 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 03:14:26,018 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 03:14:27,121 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=67225.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 03:14:28,518 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.343e+02 1.916e+02 2.226e+02 2.655e+02 7.651e+02, threshold=4.451e+02, percent-clipped=1.0 +2023-03-21 03:14:38,508 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 03:14:41,588 INFO [train.py:901] (0/2) Epoch 24, batch 2300, loss[loss=0.1135, simple_loss=0.1848, pruned_loss=0.02109, over 7022.00 frames. ], tot_loss[loss=0.146, simple_loss=0.2249, pruned_loss=0.0335, over 1445019.66 frames. ], batch size: 35, lr: 6.84e-03, grad_scale: 8.0 +2023-03-21 03:14:55,678 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8315, 3.9326, 3.7410, 3.8670, 3.6719, 4.0014, 4.2542, 4.3120], + device='cuda:0'), covar=tensor([0.0196, 0.0154, 0.0194, 0.0161, 0.0299, 0.0249, 0.0203, 0.0153], + device='cuda:0'), in_proj_covar=tensor([0.0114, 0.0117, 0.0107, 0.0112, 0.0104, 0.0094, 0.0093, 0.0089], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:15:06,586 INFO [train.py:901] (0/2) Epoch 24, batch 2350, loss[loss=0.1319, simple_loss=0.2165, pruned_loss=0.02369, over 7286.00 frames. ], tot_loss[loss=0.1446, simple_loss=0.2238, pruned_loss=0.03275, over 1445010.65 frames. ], batch size: 68, lr: 6.84e-03, grad_scale: 8.0 +2023-03-21 03:15:12,307 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=67313.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 03:15:19,734 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.340e+02 1.854e+02 2.186e+02 2.625e+02 5.810e+02, threshold=4.373e+02, percent-clipped=2.0 +2023-03-21 03:15:25,788 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 03:15:32,495 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 03:15:32,986 INFO [train.py:901] (0/2) Epoch 24, batch 2400, loss[loss=0.1482, simple_loss=0.2347, pruned_loss=0.03087, over 7252.00 frames. ], tot_loss[loss=0.1443, simple_loss=0.2236, pruned_loss=0.03253, over 1444520.82 frames. ], batch size: 55, lr: 6.84e-03, grad_scale: 8.0 +2023-03-21 03:15:43,259 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 03:15:45,746 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 03:15:49,880 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0570, 3.0472, 2.1668, 4.0892, 1.7074, 3.8945, 1.7412, 3.2845], + device='cuda:0'), covar=tensor([0.0095, 0.0954, 0.1850, 0.0137, 0.4351, 0.0159, 0.1203, 0.0397], + device='cuda:0'), in_proj_covar=tensor([0.0198, 0.0257, 0.0280, 0.0194, 0.0265, 0.0202, 0.0250, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:15:50,319 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=67387.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 03:15:59,088 INFO [train.py:901] (0/2) Epoch 24, batch 2450, loss[loss=0.1423, simple_loss=0.2173, pruned_loss=0.0337, over 7320.00 frames. ], tot_loss[loss=0.1445, simple_loss=0.2235, pruned_loss=0.03276, over 1444452.36 frames. ], batch size: 83, lr: 6.83e-03, grad_scale: 8.0 +2023-03-21 03:16:12,254 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.300e+02 1.866e+02 2.316e+02 2.804e+02 6.661e+02, threshold=4.633e+02, percent-clipped=5.0 +2023-03-21 03:16:13,304 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 03:16:15,822 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=67435.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 03:16:17,735 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=67439.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:16:24,070 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.62 vs. limit=2.0 +2023-03-21 03:16:24,681 INFO [train.py:901] (0/2) Epoch 24, batch 2500, loss[loss=0.165, simple_loss=0.2398, pruned_loss=0.04513, over 7314.00 frames. ], tot_loss[loss=0.1441, simple_loss=0.2234, pruned_loss=0.03246, over 1444639.26 frames. ], batch size: 80, lr: 6.83e-03, grad_scale: 8.0 +2023-03-21 03:16:31,185 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0140, 4.4767, 4.5572, 4.4631, 4.4942, 4.0775, 4.6142, 4.5039], + device='cuda:0'), covar=tensor([0.0465, 0.0388, 0.0348, 0.0447, 0.0292, 0.0394, 0.0285, 0.0373], + device='cuda:0'), in_proj_covar=tensor([0.0205, 0.0236, 0.0181, 0.0182, 0.0144, 0.0212, 0.0185, 0.0141], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 03:16:37,625 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 03:16:50,200 INFO [train.py:901] (0/2) Epoch 24, batch 2550, loss[loss=0.1299, simple_loss=0.2036, pruned_loss=0.02812, over 7230.00 frames. ], tot_loss[loss=0.1444, simple_loss=0.2236, pruned_loss=0.0326, over 1444175.67 frames. ], batch size: 45, lr: 6.83e-03, grad_scale: 8.0 +2023-03-21 03:16:55,911 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=67513.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:17:03,299 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.348e+02 1.939e+02 2.232e+02 2.615e+02 5.776e+02, threshold=4.464e+02, percent-clipped=2.0 +2023-03-21 03:17:15,861 INFO [train.py:901] (0/2) Epoch 24, batch 2600, loss[loss=0.1266, simple_loss=0.2068, pruned_loss=0.02325, over 7155.00 frames. ], tot_loss[loss=0.1442, simple_loss=0.2235, pruned_loss=0.03248, over 1440734.60 frames. ], batch size: 41, lr: 6.83e-03, grad_scale: 8.0 +2023-03-21 03:17:26,317 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=67574.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 03:17:38,481 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8480, 2.0713, 2.1497, 2.2039, 2.1943, 1.8725, 1.8280, 1.6471], + device='cuda:0'), covar=tensor([0.0488, 0.0410, 0.0235, 0.0205, 0.0618, 0.0522, 0.0232, 0.0331], + device='cuda:0'), in_proj_covar=tensor([0.0031, 0.0030, 0.0030, 0.0029, 0.0030, 0.0029, 0.0033, 0.0031], + device='cuda:0'), out_proj_covar=tensor([7.8207e-05, 7.8700e-05, 7.4159e-05, 7.3819e-05, 7.6529e-05, 7.4549e-05, + 8.0711e-05, 8.1117e-05], device='cuda:0') +2023-03-21 03:17:40,612 INFO [train.py:901] (0/2) Epoch 24, batch 2650, loss[loss=0.1331, simple_loss=0.2145, pruned_loss=0.02584, over 7261.00 frames. ], tot_loss[loss=0.1432, simple_loss=0.2225, pruned_loss=0.03195, over 1438229.86 frames. ], batch size: 89, lr: 6.82e-03, grad_scale: 8.0 +2023-03-21 03:17:45,604 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=67613.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:17:52,510 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9542, 2.4930, 2.8823, 2.8844, 3.0668, 2.7396, 2.5287, 3.0118], + device='cuda:0'), covar=tensor([0.1478, 0.0925, 0.1912, 0.1400, 0.1088, 0.1261, 0.2599, 0.1405], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0060, 0.0046, 0.0045, 0.0045, 0.0042, 0.0061, 0.0048], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:17:52,862 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.362e+02 1.855e+02 2.123e+02 2.589e+02 4.353e+02, threshold=4.245e+02, percent-clipped=1.0 +2023-03-21 03:17:55,365 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0641, 3.5628, 4.0263, 3.9494, 3.8858, 3.8782, 4.0945, 3.7944], + device='cuda:0'), covar=tensor([0.0028, 0.0084, 0.0028, 0.0032, 0.0039, 0.0035, 0.0027, 0.0050], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0060, 0.0052, 0.0049, 0.0048, 0.0053, 0.0045, 0.0066], + device='cuda:0'), out_proj_covar=tensor([8.1475e-05, 1.3550e-04, 1.0733e-04, 9.7001e-05, 9.3049e-05, 1.0319e-04, + 9.7823e-05, 1.3269e-04], device='cuda:0') +2023-03-21 03:18:05,657 INFO [train.py:901] (0/2) Epoch 24, batch 2700, loss[loss=0.1145, simple_loss=0.182, pruned_loss=0.02351, over 7010.00 frames. ], tot_loss[loss=0.1429, simple_loss=0.222, pruned_loss=0.03186, over 1438591.01 frames. ], batch size: 35, lr: 6.82e-03, grad_scale: 8.0 +2023-03-21 03:18:09,620 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=67661.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:18:18,962 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=67680.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:18:24,879 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=67692.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:18:30,276 INFO [train.py:901] (0/2) Epoch 24, batch 2750, loss[loss=0.1421, simple_loss=0.2322, pruned_loss=0.026, over 7297.00 frames. ], tot_loss[loss=0.1432, simple_loss=0.2225, pruned_loss=0.03197, over 1439469.02 frames. ], batch size: 80, lr: 6.82e-03, grad_scale: 8.0 +2023-03-21 03:18:34,374 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.89 vs. limit=2.0 +2023-03-21 03:18:42,803 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.352e+02 1.781e+02 2.129e+02 2.566e+02 3.925e+02, threshold=4.258e+02, percent-clipped=0.0 +2023-03-21 03:18:48,277 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=67739.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:18:49,294 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=67741.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:18:54,794 INFO [train.py:901] (0/2) Epoch 24, batch 2800, loss[loss=0.1475, simple_loss=0.233, pruned_loss=0.03101, over 7388.00 frames. ], tot_loss[loss=0.1441, simple_loss=0.2235, pruned_loss=0.0323, over 1442723.88 frames. ], batch size: 56, lr: 6.81e-03, grad_scale: 16.0 +2023-03-21 03:18:54,944 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=67753.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 03:19:07,445 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-24.pt +2023-03-21 03:19:22,974 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 03:19:23,030 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 03:19:23,056 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 03:19:23,073 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 03:19:23,248 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 03:19:23,264 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 03:19:23,275 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 03:19:23,375 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 03:19:23,451 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 03:19:23,466 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 03:19:23,831 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 03:19:23,994 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 03:19:24,000 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 03:19:24,008 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 03:19:24,148 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 03:19:24,151 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 03:19:24,210 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 03:19:26,438 INFO [train.py:901] (0/2) Epoch 25, batch 0, loss[loss=0.1619, simple_loss=0.2368, pruned_loss=0.04346, over 7307.00 frames. ], tot_loss[loss=0.1619, simple_loss=0.2368, pruned_loss=0.04346, over 7307.00 frames. ], batch size: 49, lr: 6.68e-03, grad_scale: 16.0 +2023-03-21 03:19:26,440 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 03:19:32,138 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.4118, 5.1936, 4.9245, 5.4315, 5.2826, 5.5486, 5.0670, 5.1357], + device='cuda:0'), covar=tensor([0.0458, 0.1342, 0.1246, 0.0807, 0.0527, 0.0625, 0.0349, 0.0509], + device='cuda:0'), in_proj_covar=tensor([0.0230, 0.0356, 0.0273, 0.0275, 0.0203, 0.0344, 0.0207, 0.0252], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:19:35,099 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0130, 2.7007, 2.0648, 3.2521, 1.9514, 2.6776, 1.5262, 1.9793], + device='cuda:0'), covar=tensor([0.0427, 0.0660, 0.2504, 0.0717, 0.0485, 0.0668, 0.3696, 0.1715], + device='cuda:0'), in_proj_covar=tensor([0.0270, 0.0248, 0.0289, 0.0260, 0.0268, 0.0258, 0.0247, 0.0271], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:19:46,739 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4209, 1.5906, 1.4122, 1.5089, 1.5938, 1.5708, 1.4659, 1.3129], + device='cuda:0'), covar=tensor([0.0164, 0.0153, 0.0173, 0.0167, 0.0081, 0.0090, 0.0150, 0.0113], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0026, 0.0026, 0.0028, 0.0025, 0.0025, 0.0028, 0.0035], + device='cuda:0'), out_proj_covar=tensor([3.2567e-05, 2.9842e-05, 2.9985e-05, 3.0858e-05, 2.9073e-05, 2.8247e-05, + 3.1645e-05, 4.0120e-05], device='cuda:0') +2023-03-21 03:19:52,672 INFO [train.py:935] (0/2) Epoch 25, validation: loss=0.1652, simple_loss=0.2538, pruned_loss=0.03834, over 1622729.00 frames. +2023-03-21 03:19:52,673 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 03:19:57,702 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=67787.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:19:59,661 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 03:20:02,982 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.78 vs. limit=5.0 +2023-03-21 03:20:10,371 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 03:20:16,817 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 03:20:17,733 INFO [train.py:901] (0/2) Epoch 25, batch 50, loss[loss=0.1406, simple_loss=0.221, pruned_loss=0.03007, over 7281.00 frames. ], tot_loss[loss=0.1455, simple_loss=0.2246, pruned_loss=0.03318, over 327348.10 frames. ], batch size: 77, lr: 6.68e-03, grad_scale: 16.0 +2023-03-21 03:20:18,206 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.338e+02 1.908e+02 2.288e+02 2.779e+02 7.073e+02, threshold=4.575e+02, percent-clipped=8.0 +2023-03-21 03:20:19,238 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 03:20:22,161 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 03:20:40,063 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=67869.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 03:20:40,491 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 03:20:40,991 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 03:20:43,889 INFO [train.py:901] (0/2) Epoch 25, batch 100, loss[loss=0.1329, simple_loss=0.2223, pruned_loss=0.02168, over 7327.00 frames. ], tot_loss[loss=0.1452, simple_loss=0.2237, pruned_loss=0.03335, over 573967.55 frames. ], batch size: 61, lr: 6.67e-03, grad_scale: 16.0 +2023-03-21 03:20:52,188 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.18 vs. limit=5.0 +2023-03-21 03:20:58,497 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2139, 2.9568, 2.0876, 3.4496, 3.2140, 3.3265, 2.9304, 2.8177], + device='cuda:0'), covar=tensor([0.1882, 0.0641, 0.3231, 0.0603, 0.0174, 0.0171, 0.0325, 0.0350], + device='cuda:0'), in_proj_covar=tensor([0.0266, 0.0237, 0.0269, 0.0268, 0.0179, 0.0177, 0.0206, 0.0220], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:21:05,482 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6070, 3.8164, 3.5272, 3.7754, 3.4757, 3.7625, 4.0402, 4.1150], + device='cuda:0'), covar=tensor([0.0237, 0.0177, 0.0231, 0.0190, 0.0338, 0.0406, 0.0247, 0.0204], + device='cuda:0'), in_proj_covar=tensor([0.0115, 0.0118, 0.0108, 0.0112, 0.0105, 0.0095, 0.0095, 0.0091], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:21:08,882 INFO [train.py:901] (0/2) Epoch 25, batch 150, loss[loss=0.149, simple_loss=0.2245, pruned_loss=0.03672, over 7227.00 frames. ], tot_loss[loss=0.1452, simple_loss=0.224, pruned_loss=0.03317, over 767232.77 frames. ], batch size: 45, lr: 6.67e-03, grad_scale: 16.0 +2023-03-21 03:21:09,365 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.311e+02 1.796e+02 2.230e+02 2.756e+02 5.619e+02, threshold=4.460e+02, percent-clipped=1.0 +2023-03-21 03:21:35,041 INFO [train.py:901] (0/2) Epoch 25, batch 200, loss[loss=0.1455, simple_loss=0.2355, pruned_loss=0.02782, over 7359.00 frames. ], tot_loss[loss=0.144, simple_loss=0.2233, pruned_loss=0.03236, over 917789.03 frames. ], batch size: 54, lr: 6.67e-03, grad_scale: 16.0 +2023-03-21 03:21:37,203 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=67981.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:21:41,629 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 03:21:46,146 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 03:21:46,950 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-68000.pt +2023-03-21 03:21:56,621 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 03:22:04,536 INFO [train.py:901] (0/2) Epoch 25, batch 250, loss[loss=0.1384, simple_loss=0.2213, pruned_loss=0.02776, over 7293.00 frames. ], tot_loss[loss=0.1439, simple_loss=0.2228, pruned_loss=0.03247, over 1034783.55 frames. ], batch size: 86, lr: 6.67e-03, grad_scale: 16.0 +2023-03-21 03:22:04,989 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.332e+02 1.830e+02 2.222e+02 2.557e+02 4.530e+02, threshold=4.443e+02, percent-clipped=1.0 +2023-03-21 03:22:09,626 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=68036.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:22:10,556 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 03:22:12,602 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=68042.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:22:15,492 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=68048.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 03:22:29,838 INFO [train.py:901] (0/2) Epoch 25, batch 300, loss[loss=0.1332, simple_loss=0.2163, pruned_loss=0.02504, over 7309.00 frames. ], tot_loss[loss=0.1436, simple_loss=0.2225, pruned_loss=0.03234, over 1125558.23 frames. ], batch size: 86, lr: 6.66e-03, grad_scale: 16.0 +2023-03-21 03:22:29,863 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 03:22:35,013 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=68087.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 03:22:39,367 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 03:22:55,967 INFO [train.py:901] (0/2) Epoch 25, batch 350, loss[loss=0.1291, simple_loss=0.2061, pruned_loss=0.02605, over 7327.00 frames. ], tot_loss[loss=0.143, simple_loss=0.2223, pruned_loss=0.0319, over 1195169.15 frames. ], batch size: 75, lr: 6.66e-03, grad_scale: 16.0 +2023-03-21 03:22:56,441 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.041e+02 1.796e+02 2.023e+02 2.447e+02 4.467e+02, threshold=4.045e+02, percent-clipped=1.0 +2023-03-21 03:23:06,612 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=68148.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 03:23:11,849 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-03-21 03:23:13,498 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 03:23:17,039 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=68169.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:23:20,915 INFO [train.py:901] (0/2) Epoch 25, batch 400, loss[loss=0.1472, simple_loss=0.2239, pruned_loss=0.03523, over 7293.00 frames. ], tot_loss[loss=0.1425, simple_loss=0.2217, pruned_loss=0.03164, over 1250314.32 frames. ], batch size: 83, lr: 6.66e-03, grad_scale: 16.0 +2023-03-21 03:23:41,435 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9533, 1.9053, 2.3192, 2.3120, 2.0830, 2.0118, 1.7454, 1.6545], + device='cuda:0'), covar=tensor([0.0425, 0.0609, 0.0211, 0.0150, 0.0579, 0.0680, 0.0430, 0.0436], + device='cuda:0'), in_proj_covar=tensor([0.0030, 0.0031, 0.0029, 0.0029, 0.0029, 0.0029, 0.0032, 0.0031], + device='cuda:0'), out_proj_covar=tensor([7.7102e-05, 7.9083e-05, 7.3169e-05, 7.3027e-05, 7.5621e-05, 7.3696e-05, + 7.9850e-05, 8.0630e-05], device='cuda:0') +2023-03-21 03:23:42,356 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=68217.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:23:47,347 INFO [train.py:901] (0/2) Epoch 25, batch 450, loss[loss=0.1474, simple_loss=0.2274, pruned_loss=0.03373, over 7335.00 frames. ], tot_loss[loss=0.1429, simple_loss=0.2223, pruned_loss=0.03177, over 1293893.45 frames. ], batch size: 61, lr: 6.66e-03, grad_scale: 16.0 +2023-03-21 03:23:47,815 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.300e+02 1.902e+02 2.293e+02 2.811e+02 5.974e+02, threshold=4.586e+02, percent-clipped=4.0 +2023-03-21 03:23:53,307 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 03:23:53,824 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 03:24:11,381 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.89 vs. limit=2.0 +2023-03-21 03:24:12,138 INFO [train.py:901] (0/2) Epoch 25, batch 500, loss[loss=0.1505, simple_loss=0.2304, pruned_loss=0.03533, over 7241.00 frames. ], tot_loss[loss=0.1425, simple_loss=0.2217, pruned_loss=0.03169, over 1326943.23 frames. ], batch size: 55, lr: 6.65e-03, grad_scale: 16.0 +2023-03-21 03:24:27,221 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 03:24:28,741 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 03:24:29,235 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 03:24:31,736 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 03:24:33,309 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1326, 4.5994, 4.6236, 4.5775, 4.5585, 4.1370, 4.6946, 4.5604], + device='cuda:0'), covar=tensor([0.0466, 0.0439, 0.0447, 0.0468, 0.0306, 0.0405, 0.0371, 0.0422], + device='cuda:0'), in_proj_covar=tensor([0.0203, 0.0235, 0.0179, 0.0181, 0.0143, 0.0208, 0.0186, 0.0140], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 03:24:35,746 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 03:24:38,182 INFO [train.py:901] (0/2) Epoch 25, batch 550, loss[loss=0.1463, simple_loss=0.2289, pruned_loss=0.03182, over 7279.00 frames. ], tot_loss[loss=0.1421, simple_loss=0.2211, pruned_loss=0.03158, over 1351143.17 frames. ], batch size: 57, lr: 6.65e-03, grad_scale: 16.0 +2023-03-21 03:24:38,651 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.401e+02 1.821e+02 2.169e+02 2.559e+02 5.307e+02, threshold=4.339e+02, percent-clipped=1.0 +2023-03-21 03:24:42,776 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=68336.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:24:43,250 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=68337.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:24:47,100 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 03:24:48,680 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=68348.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:24:55,029 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 03:24:57,954 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 03:25:04,174 INFO [train.py:901] (0/2) Epoch 25, batch 600, loss[loss=0.123, simple_loss=0.2009, pruned_loss=0.02259, over 7328.00 frames. ], tot_loss[loss=0.1424, simple_loss=0.2212, pruned_loss=0.03174, over 1368838.69 frames. ], batch size: 44, lr: 6.65e-03, grad_scale: 16.0 +2023-03-21 03:25:06,217 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 03:25:07,718 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=68384.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:25:13,679 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=68396.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:25:15,214 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=68399.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:25:22,888 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 03:25:28,695 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.02 vs. limit=2.0 +2023-03-21 03:25:29,366 INFO [train.py:901] (0/2) Epoch 25, batch 650, loss[loss=0.1551, simple_loss=0.2411, pruned_loss=0.0346, over 7321.00 frames. ], tot_loss[loss=0.1423, simple_loss=0.2215, pruned_loss=0.03157, over 1387040.23 frames. ], batch size: 83, lr: 6.65e-03, grad_scale: 16.0 +2023-03-21 03:25:29,821 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.426e+02 1.836e+02 2.195e+02 2.668e+02 4.895e+02, threshold=4.391e+02, percent-clipped=2.0 +2023-03-21 03:25:31,799 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 03:25:37,328 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=68443.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 03:25:38,414 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9021, 2.4383, 1.8566, 2.9432, 2.7512, 2.9007, 2.5470, 2.7885], + device='cuda:0'), covar=tensor([0.1935, 0.0889, 0.3243, 0.0828, 0.0234, 0.0278, 0.0281, 0.0369], + device='cuda:0'), in_proj_covar=tensor([0.0263, 0.0234, 0.0263, 0.0265, 0.0179, 0.0175, 0.0205, 0.0217], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:25:46,402 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=68460.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:25:48,264 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 03:25:55,638 INFO [train.py:901] (0/2) Epoch 25, batch 700, loss[loss=0.1622, simple_loss=0.2321, pruned_loss=0.04614, over 7246.00 frames. ], tot_loss[loss=0.1424, simple_loss=0.2215, pruned_loss=0.03164, over 1399606.90 frames. ], batch size: 64, lr: 6.64e-03, grad_scale: 16.0 +2023-03-21 03:25:57,657 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 03:26:03,778 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5377, 2.6331, 2.2218, 3.8300, 1.7201, 3.5512, 1.5171, 2.9627], + device='cuda:0'), covar=tensor([0.0136, 0.0980, 0.1575, 0.0143, 0.3608, 0.0230, 0.1063, 0.0348], + device='cuda:0'), in_proj_covar=tensor([0.0196, 0.0257, 0.0274, 0.0192, 0.0263, 0.0204, 0.0250, 0.0227], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:26:05,534 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.83 vs. limit=5.0 +2023-03-21 03:26:12,220 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-21 03:26:15,918 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8668, 2.0712, 1.7375, 2.7285, 2.4399, 2.4831, 2.1040, 2.5475], + device='cuda:0'), covar=tensor([0.1874, 0.0879, 0.3280, 0.0785, 0.0255, 0.0214, 0.0291, 0.0395], + device='cuda:0'), in_proj_covar=tensor([0.0262, 0.0232, 0.0262, 0.0262, 0.0178, 0.0174, 0.0203, 0.0216], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:26:20,651 INFO [train.py:901] (0/2) Epoch 25, batch 750, loss[loss=0.1502, simple_loss=0.2279, pruned_loss=0.03628, over 7129.00 frames. ], tot_loss[loss=0.1432, simple_loss=0.2223, pruned_loss=0.03203, over 1409129.67 frames. ], batch size: 41, lr: 6.64e-03, grad_scale: 16.0 +2023-03-21 03:26:21,134 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.241e+02 1.945e+02 2.256e+02 2.748e+02 5.442e+02, threshold=4.512e+02, percent-clipped=1.0 +2023-03-21 03:26:21,178 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 03:26:22,157 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 03:26:36,930 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 03:26:40,095 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3899, 4.4528, 3.8484, 3.8832, 3.5281, 2.3029, 2.1578, 4.5421], + device='cuda:0'), covar=tensor([0.0038, 0.0029, 0.0072, 0.0055, 0.0090, 0.0431, 0.0471, 0.0031], + device='cuda:0'), in_proj_covar=tensor([0.0098, 0.0078, 0.0100, 0.0087, 0.0115, 0.0123, 0.0120, 0.0091], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 03:26:41,012 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 03:26:46,922 INFO [train.py:901] (0/2) Epoch 25, batch 800, loss[loss=0.1354, simple_loss=0.2243, pruned_loss=0.02323, over 7131.00 frames. ], tot_loss[loss=0.1443, simple_loss=0.2234, pruned_loss=0.0326, over 1414710.77 frames. ], batch size: 98, lr: 6.64e-03, grad_scale: 16.0 +2023-03-21 03:26:46,931 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 03:26:47,551 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=68578.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:26:47,963 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 03:26:58,793 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 03:27:12,627 INFO [train.py:901] (0/2) Epoch 25, batch 850, loss[loss=0.1398, simple_loss=0.2296, pruned_loss=0.02498, over 7328.00 frames. ], tot_loss[loss=0.1446, simple_loss=0.2238, pruned_loss=0.03269, over 1423010.15 frames. ], batch size: 83, lr: 6.64e-03, grad_scale: 16.0 +2023-03-21 03:27:13,101 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.204e+02 1.927e+02 2.218e+02 2.601e+02 4.733e+02, threshold=4.436e+02, percent-clipped=1.0 +2023-03-21 03:27:15,728 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0684, 3.6343, 4.0415, 3.8086, 4.0548, 4.0918, 3.9765, 3.8914], + device='cuda:0'), covar=tensor([0.0037, 0.0096, 0.0042, 0.0045, 0.0039, 0.0033, 0.0040, 0.0051], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0061, 0.0053, 0.0051, 0.0049, 0.0054, 0.0047, 0.0066], + device='cuda:0'), out_proj_covar=tensor([8.3461e-05, 1.3906e-04, 1.0815e-04, 9.9367e-05, 9.3924e-05, 1.0404e-04, + 1.0098e-04, 1.3292e-04], device='cuda:0') +2023-03-21 03:27:17,601 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 03:27:17,610 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 03:27:17,706 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=68637.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:27:19,349 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=68639.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:27:23,698 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 03:27:24,229 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8228, 4.3623, 4.2582, 4.8168, 4.6384, 4.7555, 4.0607, 4.3506], + device='cuda:0'), covar=tensor([0.0774, 0.2257, 0.2213, 0.0876, 0.0780, 0.1052, 0.0733, 0.0977], + device='cuda:0'), in_proj_covar=tensor([0.0226, 0.0348, 0.0272, 0.0270, 0.0200, 0.0331, 0.0203, 0.0247], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:27:26,628 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 03:27:33,488 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.67 vs. limit=2.0 +2023-03-21 03:27:38,146 INFO [train.py:901] (0/2) Epoch 25, batch 900, loss[loss=0.1378, simple_loss=0.2162, pruned_loss=0.02972, over 7309.00 frames. ], tot_loss[loss=0.1442, simple_loss=0.2237, pruned_loss=0.03237, over 1426574.60 frames. ], batch size: 83, lr: 6.64e-03, grad_scale: 16.0 +2023-03-21 03:27:42,086 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=68685.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:27:42,150 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=68685.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:27:46,404 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.90 vs. limit=2.0 +2023-03-21 03:28:00,154 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3739, 1.6227, 1.5100, 1.4005, 1.5988, 1.4762, 1.4006, 1.1762], + device='cuda:0'), covar=tensor([0.0131, 0.0115, 0.0171, 0.0150, 0.0085, 0.0093, 0.0157, 0.0169], + device='cuda:0'), in_proj_covar=tensor([0.0028, 0.0026, 0.0026, 0.0028, 0.0026, 0.0025, 0.0027, 0.0035], + device='cuda:0'), out_proj_covar=tensor([3.2423e-05, 2.9312e-05, 2.9449e-05, 3.1397e-05, 2.9682e-05, 2.8081e-05, + 3.1375e-05, 3.9529e-05], device='cuda:0') +2023-03-21 03:28:04,028 INFO [train.py:901] (0/2) Epoch 25, batch 950, loss[loss=0.1392, simple_loss=0.2216, pruned_loss=0.02834, over 7338.00 frames. ], tot_loss[loss=0.1437, simple_loss=0.2228, pruned_loss=0.03224, over 1427841.63 frames. ], batch size: 63, lr: 6.63e-03, grad_scale: 16.0 +2023-03-21 03:28:04,490 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.356e+02 1.864e+02 2.097e+02 2.567e+02 3.634e+02, threshold=4.194e+02, percent-clipped=0.0 +2023-03-21 03:28:04,553 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 03:28:12,051 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=68743.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 03:28:13,572 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=68746.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:28:17,949 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=68755.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:28:23,233 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.77 vs. limit=2.0 +2023-03-21 03:28:26,768 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 03:28:28,767 INFO [train.py:901] (0/2) Epoch 25, batch 1000, loss[loss=0.1414, simple_loss=0.2261, pruned_loss=0.02835, over 7333.00 frames. ], tot_loss[loss=0.1442, simple_loss=0.2233, pruned_loss=0.03251, over 1432587.93 frames. ], batch size: 61, lr: 6.63e-03, grad_scale: 16.0 +2023-03-21 03:28:35,771 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=68791.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 03:28:48,260 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 03:28:48,425 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9601, 2.9935, 2.0277, 3.5055, 2.3894, 2.7997, 1.3792, 2.1055], + device='cuda:0'), covar=tensor([0.0363, 0.0794, 0.2416, 0.0607, 0.0330, 0.0464, 0.3301, 0.1666], + device='cuda:0'), in_proj_covar=tensor([0.0277, 0.0254, 0.0294, 0.0269, 0.0273, 0.0265, 0.0253, 0.0275], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 03:28:55,270 INFO [train.py:901] (0/2) Epoch 25, batch 1050, loss[loss=0.1554, simple_loss=0.2395, pruned_loss=0.03565, over 6737.00 frames. ], tot_loss[loss=0.144, simple_loss=0.2234, pruned_loss=0.03231, over 1432840.57 frames. ], batch size: 106, lr: 6.63e-03, grad_scale: 16.0 +2023-03-21 03:28:55,738 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.223e+02 1.902e+02 2.261e+02 2.596e+02 5.077e+02, threshold=4.521e+02, percent-clipped=1.0 +2023-03-21 03:29:09,928 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 03:29:14,050 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 03:29:20,593 INFO [train.py:901] (0/2) Epoch 25, batch 1100, loss[loss=0.1609, simple_loss=0.2365, pruned_loss=0.04265, over 7351.00 frames. ], tot_loss[loss=0.144, simple_loss=0.2233, pruned_loss=0.03232, over 1434064.85 frames. ], batch size: 73, lr: 6.63e-03, grad_scale: 16.0 +2023-03-21 03:29:43,294 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 03:29:43,802 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 03:29:46,775 INFO [train.py:901] (0/2) Epoch 25, batch 1150, loss[loss=0.158, simple_loss=0.2355, pruned_loss=0.04022, over 7273.00 frames. ], tot_loss[loss=0.1434, simple_loss=0.2228, pruned_loss=0.032, over 1436314.12 frames. ], batch size: 52, lr: 6.62e-03, grad_scale: 16.0 +2023-03-21 03:29:47,272 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.304e+02 1.925e+02 2.200e+02 2.576e+02 4.325e+02, threshold=4.400e+02, percent-clipped=0.0 +2023-03-21 03:29:50,310 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=68934.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:29:56,123 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 03:29:56,625 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 03:30:12,254 INFO [train.py:901] (0/2) Epoch 25, batch 1200, loss[loss=0.1567, simple_loss=0.2471, pruned_loss=0.03313, over 7337.00 frames. ], tot_loss[loss=0.1433, simple_loss=0.223, pruned_loss=0.03186, over 1438997.19 frames. ], batch size: 61, lr: 6.62e-03, grad_scale: 16.0 +2023-03-21 03:30:13,343 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0532, 3.6829, 3.8082, 3.7662, 3.4570, 3.6034, 3.9453, 3.5373], + device='cuda:0'), covar=tensor([0.0173, 0.0189, 0.0168, 0.0194, 0.0583, 0.0159, 0.0201, 0.0199], + device='cuda:0'), in_proj_covar=tensor([0.0084, 0.0093, 0.0091, 0.0080, 0.0157, 0.0099, 0.0095, 0.0099], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:30:25,214 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1531, 3.4020, 3.0079, 3.5437, 3.5209, 3.1894, 3.1010, 3.0519], + device='cuda:0'), covar=tensor([0.1019, 0.1133, 0.1077, 0.0618, 0.0817, 0.0567, 0.1438, 0.1191], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0049, 0.0057, 0.0051, 0.0048, 0.0052, 0.0051, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:30:27,707 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1059, 1.4392, 1.3990, 1.3192, 1.4052, 1.3382, 1.2735, 1.1007], + device='cuda:0'), covar=tensor([0.0162, 0.0106, 0.0165, 0.0116, 0.0107, 0.0088, 0.0125, 0.0134], + device='cuda:0'), in_proj_covar=tensor([0.0029, 0.0026, 0.0026, 0.0029, 0.0026, 0.0025, 0.0028, 0.0035], + device='cuda:0'), out_proj_covar=tensor([3.3008e-05, 2.9445e-05, 2.9639e-05, 3.2172e-05, 3.0127e-05, 2.7910e-05, + 3.1597e-05, 3.9695e-05], device='cuda:0') +2023-03-21 03:30:30,050 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 03:30:37,944 INFO [train.py:901] (0/2) Epoch 25, batch 1250, loss[loss=0.146, simple_loss=0.2227, pruned_loss=0.03467, over 7366.00 frames. ], tot_loss[loss=0.1436, simple_loss=0.2229, pruned_loss=0.03211, over 1438281.97 frames. ], batch size: 51, lr: 6.62e-03, grad_scale: 16.0 +2023-03-21 03:30:38,399 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.226e+02 1.819e+02 2.207e+02 2.633e+02 5.960e+02, threshold=4.414e+02, percent-clipped=3.0 +2023-03-21 03:30:44,993 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69041.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:30:46,062 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=69043.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:30:52,010 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69055.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:30:52,388 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 03:30:57,052 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 03:30:58,056 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 03:31:04,029 INFO [train.py:901] (0/2) Epoch 25, batch 1300, loss[loss=0.1618, simple_loss=0.2381, pruned_loss=0.04278, over 7251.00 frames. ], tot_loss[loss=0.1444, simple_loss=0.2236, pruned_loss=0.03257, over 1438933.75 frames. ], batch size: 47, lr: 6.62e-03, grad_scale: 16.0 +2023-03-21 03:31:17,168 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=69103.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:31:17,721 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=69104.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:31:21,511 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 03:31:23,518 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 03:31:27,043 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 03:31:29,076 INFO [train.py:901] (0/2) Epoch 25, batch 1350, loss[loss=0.1532, simple_loss=0.2285, pruned_loss=0.03898, over 7242.00 frames. ], tot_loss[loss=0.144, simple_loss=0.2235, pruned_loss=0.03228, over 1441208.25 frames. ], batch size: 55, lr: 6.61e-03, grad_scale: 16.0 +2023-03-21 03:31:29,537 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.299e+02 1.957e+02 2.209e+02 2.682e+02 1.203e+03, threshold=4.418e+02, percent-clipped=5.0 +2023-03-21 03:31:36,548 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8520, 1.8927, 2.1897, 2.1209, 2.0766, 2.1959, 1.7872, 1.6261], + device='cuda:0'), covar=tensor([0.0478, 0.0506, 0.0318, 0.0225, 0.0312, 0.0240, 0.0323, 0.0330], + device='cuda:0'), in_proj_covar=tensor([0.0031, 0.0030, 0.0029, 0.0029, 0.0030, 0.0029, 0.0032, 0.0032], + device='cuda:0'), out_proj_covar=tensor([7.8275e-05, 7.8321e-05, 7.3814e-05, 7.3849e-05, 7.6265e-05, 7.4068e-05, + 8.0149e-05, 8.1991e-05], device='cuda:0') +2023-03-21 03:31:37,447 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 03:31:47,264 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=69162.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:31:49,889 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6085, 1.2840, 1.6574, 1.9376, 1.6367, 1.9516, 1.5713, 1.9424], + device='cuda:0'), covar=tensor([0.2432, 0.4278, 0.1402, 0.1262, 0.1790, 0.1630, 0.2582, 0.2438], + device='cuda:0'), in_proj_covar=tensor([0.0060, 0.0064, 0.0051, 0.0046, 0.0050, 0.0049, 0.0078, 0.0050], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:31:50,916 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=69168.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:31:51,349 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5193, 5.0311, 5.0924, 5.0108, 4.9393, 4.4897, 5.0990, 4.9709], + device='cuda:0'), covar=tensor([0.0492, 0.0445, 0.0469, 0.0572, 0.0312, 0.0343, 0.0355, 0.0438], + device='cuda:0'), in_proj_covar=tensor([0.0206, 0.0239, 0.0181, 0.0182, 0.0143, 0.0211, 0.0185, 0.0140], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 03:31:55,236 INFO [train.py:901] (0/2) Epoch 25, batch 1400, loss[loss=0.1495, simple_loss=0.2371, pruned_loss=0.03091, over 7356.00 frames. ], tot_loss[loss=0.1438, simple_loss=0.2236, pruned_loss=0.03194, over 1442139.53 frames. ], batch size: 73, lr: 6.61e-03, grad_scale: 16.0 +2023-03-21 03:32:10,658 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 03:32:18,764 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=69223.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:32:19,765 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=69225.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:32:20,622 INFO [train.py:901] (0/2) Epoch 25, batch 1450, loss[loss=0.1446, simple_loss=0.2261, pruned_loss=0.03156, over 7278.00 frames. ], tot_loss[loss=0.1438, simple_loss=0.2239, pruned_loss=0.03186, over 1443447.09 frames. ], batch size: 70, lr: 6.61e-03, grad_scale: 16.0 +2023-03-21 03:32:21,091 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.249e+02 1.877e+02 2.365e+02 2.787e+02 5.081e+02, threshold=4.731e+02, percent-clipped=1.0 +2023-03-21 03:32:21,744 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=69229.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:32:24,190 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69234.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:32:30,314 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=69245.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 03:32:35,260 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 03:32:44,809 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=69273.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:32:46,637 INFO [train.py:901] (0/2) Epoch 25, batch 1500, loss[loss=0.1408, simple_loss=0.2219, pruned_loss=0.02987, over 7230.00 frames. ], tot_loss[loss=0.1435, simple_loss=0.2234, pruned_loss=0.03181, over 1442276.32 frames. ], batch size: 55, lr: 6.61e-03, grad_scale: 16.0 +2023-03-21 03:32:49,139 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=69282.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:32:51,200 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=69286.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:32:51,569 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 03:33:01,192 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=69306.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 03:33:08,151 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=69320.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:33:12,044 INFO [train.py:901] (0/2) Epoch 25, batch 1550, loss[loss=0.1446, simple_loss=0.2259, pruned_loss=0.03168, over 7311.00 frames. ], tot_loss[loss=0.143, simple_loss=0.2229, pruned_loss=0.03152, over 1443094.59 frames. ], batch size: 80, lr: 6.60e-03, grad_scale: 16.0 +2023-03-21 03:33:12,527 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.197e+02 1.813e+02 2.080e+02 2.419e+02 4.351e+02, threshold=4.160e+02, percent-clipped=0.0 +2023-03-21 03:33:14,521 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 03:33:15,630 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=69334.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 03:33:19,683 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69341.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:33:30,274 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-03-21 03:33:33,160 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=69368.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:33:37,603 INFO [train.py:901] (0/2) Epoch 25, batch 1600, loss[loss=0.1226, simple_loss=0.2029, pruned_loss=0.02118, over 7338.00 frames. ], tot_loss[loss=0.1437, simple_loss=0.2238, pruned_loss=0.03181, over 1445117.63 frames. ], batch size: 44, lr: 6.60e-03, grad_scale: 16.0 +2023-03-21 03:33:39,747 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=69381.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:33:40,240 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6642, 1.3512, 1.7067, 2.0472, 1.6837, 2.0381, 1.8694, 2.0672], + device='cuda:0'), covar=tensor([0.1655, 0.4231, 0.1678, 0.1426, 0.3524, 0.2828, 0.1861, 0.2686], + device='cuda:0'), in_proj_covar=tensor([0.0061, 0.0066, 0.0053, 0.0048, 0.0051, 0.0051, 0.0080, 0.0051], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:33:40,729 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6644, 2.8963, 2.4117, 2.7732, 2.8952, 2.6034, 2.7892, 2.8876], + device='cuda:0'), covar=tensor([0.0731, 0.0954, 0.1144, 0.1340, 0.0632, 0.0697, 0.0810, 0.0547], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0050, 0.0058, 0.0051, 0.0049, 0.0052, 0.0051, 0.0046], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:33:43,700 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=69389.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:33:44,667 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 03:33:45,181 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 03:33:46,275 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=69394.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:33:48,642 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 03:33:48,706 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69399.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:33:57,522 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.95 vs. limit=2.0 +2023-03-21 03:33:59,195 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 03:33:59,830 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2535, 2.3721, 2.1294, 3.5840, 1.5749, 3.1936, 1.4553, 3.0139], + device='cuda:0'), covar=tensor([0.0143, 0.1058, 0.1547, 0.0152, 0.3619, 0.0167, 0.1039, 0.0325], + device='cuda:0'), in_proj_covar=tensor([0.0201, 0.0258, 0.0277, 0.0197, 0.0268, 0.0207, 0.0253, 0.0231], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:34:03,649 INFO [train.py:901] (0/2) Epoch 25, batch 1650, loss[loss=0.1581, simple_loss=0.2332, pruned_loss=0.04151, over 7242.00 frames. ], tot_loss[loss=0.1428, simple_loss=0.2228, pruned_loss=0.03139, over 1443875.03 frames. ], batch size: 55, lr: 6.60e-03, grad_scale: 16.0 +2023-03-21 03:34:04,321 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 03:34:04,764 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.215e+02 1.921e+02 2.303e+02 2.853e+02 4.440e+02, threshold=4.606e+02, percent-clipped=2.0 +2023-03-21 03:34:05,441 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=69429.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:34:12,337 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 03:34:18,467 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=69455.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:34:19,406 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4551, 4.0509, 4.1714, 4.1948, 4.1530, 4.0478, 4.4622, 3.7832], + device='cuda:0'), covar=tensor([0.0111, 0.0133, 0.0113, 0.0110, 0.0332, 0.0104, 0.0119, 0.0178], + device='cuda:0'), in_proj_covar=tensor([0.0083, 0.0091, 0.0091, 0.0079, 0.0156, 0.0099, 0.0094, 0.0098], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:34:28,779 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 03:34:29,257 INFO [train.py:901] (0/2) Epoch 25, batch 1700, loss[loss=0.1343, simple_loss=0.2216, pruned_loss=0.02346, over 7309.00 frames. ], tot_loss[loss=0.1427, simple_loss=0.2228, pruned_loss=0.03131, over 1445086.59 frames. ], batch size: 80, lr: 6.60e-03, grad_scale: 16.0 +2023-03-21 03:34:32,729 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 03:34:43,761 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 03:34:50,934 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69518.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:34:53,833 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69524.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:34:55,269 INFO [train.py:901] (0/2) Epoch 25, batch 1750, loss[loss=0.1628, simple_loss=0.2417, pruned_loss=0.04198, over 7328.00 frames. ], tot_loss[loss=0.1433, simple_loss=0.2233, pruned_loss=0.03171, over 1443473.58 frames. ], batch size: 59, lr: 6.59e-03, grad_scale: 16.0 +2023-03-21 03:34:55,737 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.346e+02 1.951e+02 2.380e+02 2.655e+02 3.975e+02, threshold=4.760e+02, percent-clipped=0.0 +2023-03-21 03:35:08,784 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 03:35:09,762 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 03:35:20,426 INFO [train.py:901] (0/2) Epoch 25, batch 1800, loss[loss=0.14, simple_loss=0.2275, pruned_loss=0.02627, over 7354.00 frames. ], tot_loss[loss=0.1429, simple_loss=0.2231, pruned_loss=0.03131, over 1444799.23 frames. ], batch size: 73, lr: 6.59e-03, grad_scale: 16.0 +2023-03-21 03:35:20,549 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4920, 4.1962, 4.4777, 4.6264, 4.7131, 4.6302, 4.5567, 4.5223], + device='cuda:0'), covar=tensor([0.0024, 0.0068, 0.0030, 0.0024, 0.0023, 0.0027, 0.0025, 0.0045], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0061, 0.0053, 0.0050, 0.0050, 0.0054, 0.0047, 0.0067], + device='cuda:0'), out_proj_covar=tensor([8.1872e-05, 1.3786e-04, 1.0786e-04, 9.8385e-05, 9.4725e-05, 1.0463e-04, + 1.0055e-04, 1.3560e-04], device='cuda:0') +2023-03-21 03:35:22,570 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69581.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:35:32,003 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 03:35:34,065 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69601.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 03:35:46,420 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 03:35:46,926 INFO [train.py:901] (0/2) Epoch 25, batch 1850, loss[loss=0.1574, simple_loss=0.2412, pruned_loss=0.03676, over 7307.00 frames. ], tot_loss[loss=0.1429, simple_loss=0.2225, pruned_loss=0.03162, over 1440040.53 frames. ], batch size: 59, lr: 6.59e-03, grad_scale: 16.0 +2023-03-21 03:35:47,415 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.142e+02 1.867e+02 2.216e+02 2.571e+02 4.186e+02, threshold=4.433e+02, percent-clipped=0.0 +2023-03-21 03:35:47,974 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69629.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 03:35:55,959 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 03:36:11,086 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=69674.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 03:36:12,010 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69676.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:36:12,407 INFO [train.py:901] (0/2) Epoch 25, batch 1900, loss[loss=0.1654, simple_loss=0.2338, pruned_loss=0.04854, over 7287.00 frames. ], tot_loss[loss=0.1432, simple_loss=0.2228, pruned_loss=0.03175, over 1442105.08 frames. ], batch size: 47, lr: 6.59e-03, grad_scale: 16.0 +2023-03-21 03:36:12,419 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 03:36:23,988 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69699.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:36:36,422 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69724.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:36:36,854 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 03:36:37,854 INFO [train.py:901] (0/2) Epoch 25, batch 1950, loss[loss=0.1337, simple_loss=0.2196, pruned_loss=0.02389, over 7319.00 frames. ], tot_loss[loss=0.1431, simple_loss=0.2226, pruned_loss=0.03178, over 1441103.26 frames. ], batch size: 83, lr: 6.59e-03, grad_scale: 16.0 +2023-03-21 03:36:38,847 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.292e+02 1.839e+02 2.249e+02 2.694e+02 4.374e+02, threshold=4.499e+02, percent-clipped=0.0 +2023-03-21 03:36:41,962 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=69735.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 03:36:46,006 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2361, 3.0003, 2.1636, 3.5554, 2.5571, 2.9318, 1.5761, 2.0243], + device='cuda:0'), covar=tensor([0.0440, 0.0691, 0.2523, 0.0638, 0.0424, 0.0516, 0.3243, 0.1765], + device='cuda:0'), in_proj_covar=tensor([0.0273, 0.0255, 0.0293, 0.0267, 0.0272, 0.0264, 0.0253, 0.0272], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:36:47,857 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 03:36:47,892 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=69747.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:36:49,441 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=69750.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:36:52,845 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 03:36:53,852 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 03:37:04,176 INFO [train.py:901] (0/2) Epoch 25, batch 2000, loss[loss=0.1179, simple_loss=0.1934, pruned_loss=0.02127, over 7139.00 frames. ], tot_loss[loss=0.1432, simple_loss=0.2228, pruned_loss=0.03178, over 1441329.93 frames. ], batch size: 41, lr: 6.58e-03, grad_scale: 16.0 +2023-03-21 03:37:11,636 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 03:37:22,436 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 03:37:24,966 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69818.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:37:26,666 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.91 vs. limit=2.0 +2023-03-21 03:37:27,947 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69824.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:37:29,322 INFO [train.py:901] (0/2) Epoch 25, batch 2050, loss[loss=0.152, simple_loss=0.2324, pruned_loss=0.03573, over 7250.00 frames. ], tot_loss[loss=0.1431, simple_loss=0.2228, pruned_loss=0.0317, over 1442688.07 frames. ], batch size: 89, lr: 6.58e-03, grad_scale: 16.0 +2023-03-21 03:37:29,840 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 03:37:30,799 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.280e+02 1.910e+02 2.230e+02 2.522e+02 4.297e+02, threshold=4.460e+02, percent-clipped=0.0 +2023-03-21 03:37:32,508 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8651, 2.3137, 1.8066, 2.6443, 2.6577, 2.5021, 2.4316, 2.2763], + device='cuda:0'), covar=tensor([0.1944, 0.0828, 0.3366, 0.0594, 0.0173, 0.0173, 0.0221, 0.0286], + device='cuda:0'), in_proj_covar=tensor([0.0262, 0.0235, 0.0264, 0.0264, 0.0177, 0.0179, 0.0206, 0.0219], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:37:45,169 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7286, 3.0225, 3.4577, 3.7499, 3.6952, 3.7126, 3.6286, 3.5322], + device='cuda:0'), covar=tensor([0.0026, 0.0112, 0.0038, 0.0029, 0.0032, 0.0028, 0.0053, 0.0049], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0061, 0.0052, 0.0050, 0.0049, 0.0053, 0.0046, 0.0066], + device='cuda:0'), out_proj_covar=tensor([8.0788e-05, 1.3618e-04, 1.0592e-04, 9.6646e-05, 9.3832e-05, 1.0158e-04, + 9.8818e-05, 1.3299e-04], device='cuda:0') +2023-03-21 03:37:50,231 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=69866.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:37:53,200 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=69872.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:37:55,681 INFO [train.py:901] (0/2) Epoch 25, batch 2100, loss[loss=0.1136, simple_loss=0.1772, pruned_loss=0.02496, over 5821.00 frames. ], tot_loss[loss=0.1428, simple_loss=0.2223, pruned_loss=0.03161, over 1438319.41 frames. ], batch size: 25, lr: 6.58e-03, grad_scale: 8.0 +2023-03-21 03:37:57,784 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69881.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:38:02,768 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8814, 1.9990, 1.9995, 2.0079, 2.0643, 1.8246, 1.4564, 1.6967], + device='cuda:0'), covar=tensor([0.0426, 0.0369, 0.0254, 0.0198, 0.0427, 0.0500, 0.0558, 0.0257], + device='cuda:0'), in_proj_covar=tensor([0.0030, 0.0029, 0.0030, 0.0028, 0.0030, 0.0030, 0.0033, 0.0032], + device='cuda:0'), out_proj_covar=tensor([7.7755e-05, 7.7417e-05, 7.5121e-05, 7.3318e-05, 7.6578e-05, 7.5245e-05, + 8.1378e-05, 8.2496e-05], device='cuda:0') +2023-03-21 03:38:03,138 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 03:38:06,560 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 03:38:07,570 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69901.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 03:38:20,447 INFO [train.py:901] (0/2) Epoch 25, batch 2150, loss[loss=0.1257, simple_loss=0.2068, pruned_loss=0.02231, over 7322.00 frames. ], tot_loss[loss=0.1429, simple_loss=0.2223, pruned_loss=0.03176, over 1440641.71 frames. ], batch size: 44, lr: 6.58e-03, grad_scale: 8.0 +2023-03-21 03:38:21,546 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=69929.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:38:21,609 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69929.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 03:38:21,978 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.326e+02 1.939e+02 2.247e+02 2.652e+02 6.279e+02, threshold=4.495e+02, percent-clipped=1.0 +2023-03-21 03:38:33,007 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=69949.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 03:38:46,516 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=69976.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:38:46,920 INFO [train.py:901] (0/2) Epoch 25, batch 2200, loss[loss=0.1641, simple_loss=0.2363, pruned_loss=0.04594, over 7272.00 frames. ], tot_loss[loss=0.1428, simple_loss=0.2223, pruned_loss=0.03162, over 1440149.45 frames. ], batch size: 64, lr: 6.57e-03, grad_scale: 8.0 +2023-03-21 03:38:46,990 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=69977.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:38:51,748 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.69 vs. limit=2.0 +2023-03-21 03:38:51,969 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 03:39:11,479 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=70024.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:39:11,544 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=70024.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:39:12,921 INFO [train.py:901] (0/2) Epoch 25, batch 2250, loss[loss=0.1158, simple_loss=0.2012, pruned_loss=0.0152, over 7335.00 frames. ], tot_loss[loss=0.1429, simple_loss=0.2228, pruned_loss=0.03151, over 1443325.70 frames. ], batch size: 44, lr: 6.57e-03, grad_scale: 8.0 +2023-03-21 03:39:14,428 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.235e+02 1.881e+02 2.155e+02 2.675e+02 4.570e+02, threshold=4.310e+02, percent-clipped=1.0 +2023-03-21 03:39:14,521 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=70030.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 03:39:25,356 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=70050.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:39:26,790 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 03:39:27,270 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 03:39:36,198 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=70072.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:39:38,657 INFO [train.py:901] (0/2) Epoch 25, batch 2300, loss[loss=0.1311, simple_loss=0.2128, pruned_loss=0.02469, over 7282.00 frames. ], tot_loss[loss=0.1424, simple_loss=0.222, pruned_loss=0.0314, over 1441193.36 frames. ], batch size: 68, lr: 6.57e-03, grad_scale: 8.0 +2023-03-21 03:39:39,134 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 03:39:49,218 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=70098.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:39:55,770 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=70110.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:39:55,821 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8618, 2.7688, 2.8082, 2.9174, 2.6744, 2.4964, 3.0282, 2.1819], + device='cuda:0'), covar=tensor([0.0455, 0.0589, 0.0456, 0.0561, 0.0558, 0.0736, 0.0691, 0.1530], + device='cuda:0'), in_proj_covar=tensor([0.0334, 0.0336, 0.0268, 0.0357, 0.0304, 0.0300, 0.0347, 0.0277], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:40:04,778 INFO [train.py:901] (0/2) Epoch 25, batch 2350, loss[loss=0.1216, simple_loss=0.193, pruned_loss=0.02511, over 6966.00 frames. ], tot_loss[loss=0.1425, simple_loss=0.222, pruned_loss=0.0315, over 1440614.06 frames. ], batch size: 35, lr: 6.57e-03, grad_scale: 8.0 +2023-03-21 03:40:06,252 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.272e+02 1.829e+02 2.172e+02 2.649e+02 5.270e+02, threshold=4.344e+02, percent-clipped=1.0 +2023-03-21 03:40:18,079 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1440, 2.1461, 2.5038, 2.0671, 1.9017, 1.8492, 2.0748, 1.7611], + device='cuda:0'), covar=tensor([0.0290, 0.0491, 0.0155, 0.0277, 0.0685, 0.0505, 0.0288, 0.0455], + device='cuda:0'), in_proj_covar=tensor([0.0030, 0.0029, 0.0030, 0.0029, 0.0030, 0.0029, 0.0033, 0.0032], + device='cuda:0'), out_proj_covar=tensor([7.7498e-05, 7.7043e-05, 7.5114e-05, 7.3453e-05, 7.5872e-05, 7.4222e-05, + 8.0532e-05, 8.2318e-05], device='cuda:0') +2023-03-21 03:40:26,032 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 03:40:27,154 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=70171.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:40:29,605 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6443, 4.0853, 4.3512, 4.2362, 4.2001, 4.0852, 4.4447, 3.9422], + device='cuda:0'), covar=tensor([0.0097, 0.0136, 0.0095, 0.0115, 0.0332, 0.0104, 0.0123, 0.0172], + device='cuda:0'), in_proj_covar=tensor([0.0085, 0.0091, 0.0091, 0.0080, 0.0157, 0.0100, 0.0095, 0.0100], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:40:29,994 INFO [train.py:901] (0/2) Epoch 25, batch 2400, loss[loss=0.1308, simple_loss=0.2045, pruned_loss=0.02861, over 7134.00 frames. ], tot_loss[loss=0.1422, simple_loss=0.2214, pruned_loss=0.03149, over 1441839.14 frames. ], batch size: 41, lr: 6.56e-03, grad_scale: 8.0 +2023-03-21 03:40:32,455 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 03:40:42,827 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 03:40:45,379 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 03:40:56,558 INFO [train.py:901] (0/2) Epoch 25, batch 2450, loss[loss=0.1463, simple_loss=0.2173, pruned_loss=0.03771, over 7274.00 frames. ], tot_loss[loss=0.1427, simple_loss=0.222, pruned_loss=0.03165, over 1443506.08 frames. ], batch size: 47, lr: 6.56e-03, grad_scale: 8.0 +2023-03-21 03:40:58,018 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.170e+02 1.945e+02 2.299e+02 2.753e+02 4.532e+02, threshold=4.598e+02, percent-clipped=1.0 +2023-03-21 03:41:13,000 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=70260.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:41:13,412 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 03:41:15,010 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1274, 4.1631, 3.5429, 3.5618, 2.9133, 2.2399, 1.7641, 4.1476], + device='cuda:0'), covar=tensor([0.0039, 0.0031, 0.0086, 0.0065, 0.0145, 0.0490, 0.0574, 0.0043], + device='cuda:0'), in_proj_covar=tensor([0.0099, 0.0081, 0.0103, 0.0088, 0.0117, 0.0124, 0.0122, 0.0094], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 03:41:17,699 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-21 03:41:21,398 INFO [train.py:901] (0/2) Epoch 25, batch 2500, loss[loss=0.143, simple_loss=0.2256, pruned_loss=0.03021, over 7306.00 frames. ], tot_loss[loss=0.1423, simple_loss=0.2218, pruned_loss=0.0314, over 1445072.81 frames. ], batch size: 80, lr: 6.56e-03, grad_scale: 8.0 +2023-03-21 03:41:39,738 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 03:41:44,853 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=70321.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:41:47,617 INFO [train.py:901] (0/2) Epoch 25, batch 2550, loss[loss=0.1503, simple_loss=0.2365, pruned_loss=0.03208, over 7331.00 frames. ], tot_loss[loss=0.1423, simple_loss=0.2218, pruned_loss=0.0314, over 1442799.34 frames. ], batch size: 83, lr: 6.56e-03, grad_scale: 8.0 +2023-03-21 03:41:49,156 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.118e+02 1.766e+02 2.062e+02 2.405e+02 6.552e+02, threshold=4.124e+02, percent-clipped=2.0 +2023-03-21 03:41:49,289 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=70330.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 03:42:13,034 INFO [train.py:901] (0/2) Epoch 25, batch 2600, loss[loss=0.153, simple_loss=0.2347, pruned_loss=0.03567, over 7252.00 frames. ], tot_loss[loss=0.1429, simple_loss=0.2226, pruned_loss=0.03156, over 1444166.77 frames. ], batch size: 55, lr: 6.56e-03, grad_scale: 8.0 +2023-03-21 03:42:13,582 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=70378.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 03:42:37,925 INFO [train.py:901] (0/2) Epoch 25, batch 2650, loss[loss=0.1372, simple_loss=0.2136, pruned_loss=0.03038, over 7254.00 frames. ], tot_loss[loss=0.1435, simple_loss=0.2234, pruned_loss=0.03185, over 1444870.80 frames. ], batch size: 47, lr: 6.55e-03, grad_scale: 8.0 +2023-03-21 03:42:39,416 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.240e+02 1.969e+02 2.211e+02 2.540e+02 3.994e+02, threshold=4.422e+02, percent-clipped=0.0 +2023-03-21 03:42:58,040 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=70466.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:43:03,336 INFO [train.py:901] (0/2) Epoch 25, batch 2700, loss[loss=0.1553, simple_loss=0.2258, pruned_loss=0.04237, over 7269.00 frames. ], tot_loss[loss=0.1433, simple_loss=0.223, pruned_loss=0.03174, over 1445605.04 frames. ], batch size: 77, lr: 6.55e-03, grad_scale: 8.0 +2023-03-21 03:43:28,020 INFO [train.py:901] (0/2) Epoch 25, batch 2750, loss[loss=0.1638, simple_loss=0.2476, pruned_loss=0.03996, over 6802.00 frames. ], tot_loss[loss=0.1434, simple_loss=0.2231, pruned_loss=0.03187, over 1445189.81 frames. ], batch size: 106, lr: 6.55e-03, grad_scale: 8.0 +2023-03-21 03:43:29,592 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.221e+02 1.852e+02 2.282e+02 2.643e+02 6.214e+02, threshold=4.563e+02, percent-clipped=1.0 +2023-03-21 03:43:39,506 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-21 03:43:52,391 INFO [train.py:901] (0/2) Epoch 25, batch 2800, loss[loss=0.1145, simple_loss=0.1845, pruned_loss=0.02224, over 7012.00 frames. ], tot_loss[loss=0.1432, simple_loss=0.2229, pruned_loss=0.03174, over 1444368.34 frames. ], batch size: 35, lr: 6.55e-03, grad_scale: 8.0 +2023-03-21 03:43:56,789 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4642, 3.9907, 3.9707, 4.4167, 4.3712, 4.3530, 3.7379, 3.9704], + device='cuda:0'), covar=tensor([0.0980, 0.3048, 0.2560, 0.1323, 0.0875, 0.1503, 0.1103, 0.1323], + device='cuda:0'), in_proj_covar=tensor([0.0234, 0.0362, 0.0283, 0.0279, 0.0205, 0.0351, 0.0212, 0.0254], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:44:05,333 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-25.pt +2023-03-21 03:44:19,402 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 03:44:19,455 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 03:44:19,481 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 03:44:19,500 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 03:44:19,679 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 03:44:19,695 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 03:44:19,706 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 03:44:20,155 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 03:44:20,229 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 03:44:20,245 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 03:44:20,262 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 03:44:20,432 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 03:44:20,437 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 03:44:20,445 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 03:44:20,589 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 03:44:20,592 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 03:44:20,651 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 03:44:22,833 INFO [train.py:901] (0/2) Epoch 26, batch 0, loss[loss=0.1243, simple_loss=0.201, pruned_loss=0.02376, over 7286.00 frames. ], tot_loss[loss=0.1243, simple_loss=0.201, pruned_loss=0.02376, over 7286.00 frames. ], batch size: 66, lr: 6.42e-03, grad_scale: 8.0 +2023-03-21 03:44:22,834 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 03:44:27,802 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9361, 2.0791, 2.2979, 1.9872, 1.7966, 1.9850, 1.7999, 1.7072], + device='cuda:0'), covar=tensor([0.0390, 0.0283, 0.0185, 0.0184, 0.0388, 0.0365, 0.0236, 0.0250], + device='cuda:0'), in_proj_covar=tensor([0.0030, 0.0029, 0.0030, 0.0028, 0.0029, 0.0029, 0.0032, 0.0032], + device='cuda:0'), out_proj_covar=tensor([7.7333e-05, 7.6813e-05, 7.4536e-05, 7.2799e-05, 7.5776e-05, 7.4223e-05, + 7.9874e-05, 8.1938e-05], device='cuda:0') +2023-03-21 03:44:31,039 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5288, 1.0399, 1.5291, 1.8860, 1.4700, 1.7914, 1.4252, 1.9478], + device='cuda:0'), covar=tensor([0.1549, 0.2883, 0.1173, 0.0844, 0.1180, 0.1237, 0.2403, 0.1464], + device='cuda:0'), in_proj_covar=tensor([0.0062, 0.0065, 0.0052, 0.0049, 0.0053, 0.0051, 0.0081, 0.0051], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:44:36,032 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8112, 3.7909, 2.8609, 4.1239, 3.3799, 3.5581, 2.1817, 2.8445], + device='cuda:0'), covar=tensor([0.0358, 0.0747, 0.2181, 0.0437, 0.0412, 0.0595, 0.2804, 0.1729], + device='cuda:0'), in_proj_covar=tensor([0.0272, 0.0251, 0.0294, 0.0270, 0.0273, 0.0264, 0.0251, 0.0272], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:44:43,375 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0733, 2.1977, 2.5793, 2.1420, 2.0394, 1.7266, 1.9819, 1.9013], + device='cuda:0'), covar=tensor([0.0534, 0.0413, 0.0207, 0.0204, 0.0474, 0.0822, 0.0278, 0.0315], + device='cuda:0'), in_proj_covar=tensor([0.0030, 0.0029, 0.0030, 0.0028, 0.0029, 0.0029, 0.0032, 0.0032], + device='cuda:0'), out_proj_covar=tensor([7.7333e-05, 7.6813e-05, 7.4536e-05, 7.2799e-05, 7.5776e-05, 7.4223e-05, + 7.9874e-05, 8.1938e-05], device='cuda:0') +2023-03-21 03:44:48,532 INFO [train.py:935] (0/2) Epoch 26, validation: loss=0.166, simple_loss=0.2547, pruned_loss=0.03864, over 1622729.00 frames. +2023-03-21 03:44:48,533 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 03:44:56,126 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 03:44:56,184 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=70616.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:45:03,812 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.284e+02 1.970e+02 2.312e+02 2.848e+02 5.222e+02, threshold=4.623e+02, percent-clipped=3.0 +2023-03-21 03:45:07,353 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 03:45:14,348 INFO [train.py:901] (0/2) Epoch 26, batch 50, loss[loss=0.1706, simple_loss=0.2559, pruned_loss=0.04259, over 6730.00 frames. ], tot_loss[loss=0.1432, simple_loss=0.2229, pruned_loss=0.03177, over 325136.70 frames. ], batch size: 106, lr: 6.42e-03, grad_scale: 8.0 +2023-03-21 03:45:14,352 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 03:45:16,350 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 03:45:19,171 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9591, 2.8208, 2.0171, 3.2058, 2.1062, 2.5387, 1.4097, 1.8927], + device='cuda:0'), covar=tensor([0.0345, 0.0667, 0.2479, 0.0553, 0.0453, 0.0535, 0.3023, 0.1840], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0252, 0.0295, 0.0271, 0.0274, 0.0263, 0.0252, 0.0273], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:45:19,505 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 03:45:20,258 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.63 vs. limit=2.0 +2023-03-21 03:45:22,541 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9454, 4.0191, 3.8252, 4.0345, 3.6597, 4.0596, 4.2630, 4.3050], + device='cuda:0'), covar=tensor([0.0166, 0.0155, 0.0198, 0.0131, 0.0333, 0.0317, 0.0254, 0.0187], + device='cuda:0'), in_proj_covar=tensor([0.0113, 0.0117, 0.0106, 0.0111, 0.0105, 0.0094, 0.0094, 0.0089], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:45:36,344 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 03:45:36,813 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 03:45:39,755 INFO [train.py:901] (0/2) Epoch 26, batch 100, loss[loss=0.1539, simple_loss=0.2335, pruned_loss=0.03714, over 7244.00 frames. ], tot_loss[loss=0.1426, simple_loss=0.2225, pruned_loss=0.03134, over 574255.47 frames. ], batch size: 93, lr: 6.42e-03, grad_scale: 8.0 +2023-03-21 03:45:42,429 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6493, 4.1397, 4.5333, 4.6491, 4.8359, 4.7340, 4.6641, 4.6792], + device='cuda:0'), covar=tensor([0.0020, 0.0065, 0.0019, 0.0019, 0.0015, 0.0019, 0.0016, 0.0034], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0061, 0.0052, 0.0049, 0.0049, 0.0054, 0.0046, 0.0067], + device='cuda:0'), out_proj_covar=tensor([8.1938e-05, 1.3642e-04, 1.0577e-04, 9.4907e-05, 9.3389e-05, 1.0158e-04, + 9.8449e-05, 1.3317e-04], device='cuda:0') +2023-03-21 03:45:54,705 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.346e+02 1.824e+02 2.219e+02 2.666e+02 5.274e+02, threshold=4.438e+02, percent-clipped=1.0 +2023-03-21 03:46:05,892 INFO [train.py:901] (0/2) Epoch 26, batch 150, loss[loss=0.144, simple_loss=0.2283, pruned_loss=0.02982, over 7274.00 frames. ], tot_loss[loss=0.1423, simple_loss=0.2225, pruned_loss=0.03112, over 763490.70 frames. ], batch size: 89, lr: 6.41e-03, grad_scale: 8.0 +2023-03-21 03:46:13,542 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=70766.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:46:16,160 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.67 vs. limit=2.0 +2023-03-21 03:46:22,533 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8170, 3.0388, 2.5342, 2.7953, 2.9848, 2.7843, 2.8737, 2.7857], + device='cuda:0'), covar=tensor([0.0516, 0.0542, 0.1215, 0.1292, 0.0893, 0.0485, 0.1349, 0.0943], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0050, 0.0058, 0.0052, 0.0049, 0.0053, 0.0051, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:46:22,762 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.98 vs. limit=2.0 +2023-03-21 03:46:31,896 INFO [train.py:901] (0/2) Epoch 26, batch 200, loss[loss=0.1509, simple_loss=0.2293, pruned_loss=0.03622, over 7289.00 frames. ], tot_loss[loss=0.1411, simple_loss=0.2214, pruned_loss=0.03044, over 916224.04 frames. ], batch size: 68, lr: 6.41e-03, grad_scale: 8.0 +2023-03-21 03:46:33,554 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2326, 2.4511, 2.2031, 2.3444, 2.5534, 2.3151, 2.4575, 2.4236], + device='cuda:0'), covar=tensor([0.1028, 0.0720, 0.0956, 0.1167, 0.0588, 0.0524, 0.0613, 0.0648], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0050, 0.0058, 0.0052, 0.0049, 0.0052, 0.0051, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:46:37,450 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 03:46:38,521 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=70814.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:46:41,435 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 03:46:47,027 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.309e+02 1.835e+02 2.165e+02 2.609e+02 3.766e+02, threshold=4.330e+02, percent-clipped=0.0 +2023-03-21 03:46:48,077 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 03:46:57,518 INFO [train.py:901] (0/2) Epoch 26, batch 250, loss[loss=0.1404, simple_loss=0.221, pruned_loss=0.02996, over 7256.00 frames. ], tot_loss[loss=0.1413, simple_loss=0.2212, pruned_loss=0.03074, over 1031037.91 frames. ], batch size: 64, lr: 6.41e-03, grad_scale: 8.0 +2023-03-21 03:47:00,601 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 03:47:09,269 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8636, 2.8723, 3.1057, 2.8237, 2.6628, 2.4895, 3.1767, 2.2993], + device='cuda:0'), covar=tensor([0.0439, 0.0423, 0.0452, 0.0523, 0.0557, 0.0793, 0.0546, 0.1576], + device='cuda:0'), in_proj_covar=tensor([0.0331, 0.0335, 0.0269, 0.0361, 0.0299, 0.0296, 0.0342, 0.0275], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:47:21,668 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 03:47:23,108 INFO [train.py:901] (0/2) Epoch 26, batch 300, loss[loss=0.1595, simple_loss=0.2392, pruned_loss=0.03991, over 6656.00 frames. ], tot_loss[loss=0.1413, simple_loss=0.221, pruned_loss=0.03079, over 1119835.36 frames. ], batch size: 107, lr: 6.41e-03, grad_scale: 8.0 +2023-03-21 03:47:25,496 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.84 vs. limit=5.0 +2023-03-21 03:47:30,713 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 03:47:30,822 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=70916.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:47:38,301 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.193e+02 1.793e+02 2.181e+02 2.612e+02 5.300e+02, threshold=4.362e+02, percent-clipped=1.0 +2023-03-21 03:47:38,588 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.22 vs. limit=2.0 +2023-03-21 03:47:41,373 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7296, 3.7998, 3.6251, 3.7652, 3.5566, 3.8789, 4.0464, 4.0601], + device='cuda:0'), covar=tensor([0.0202, 0.0185, 0.0200, 0.0168, 0.0296, 0.0230, 0.0218, 0.0179], + device='cuda:0'), in_proj_covar=tensor([0.0113, 0.0117, 0.0106, 0.0111, 0.0105, 0.0094, 0.0094, 0.0090], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:47:48,503 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9027, 2.8542, 3.2335, 2.9743, 2.8211, 2.4606, 3.2563, 2.4027], + device='cuda:0'), covar=tensor([0.0384, 0.0476, 0.0463, 0.0518, 0.0520, 0.0739, 0.0612, 0.1401], + device='cuda:0'), in_proj_covar=tensor([0.0332, 0.0336, 0.0270, 0.0362, 0.0301, 0.0297, 0.0344, 0.0275], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:47:48,824 INFO [train.py:901] (0/2) Epoch 26, batch 350, loss[loss=0.1508, simple_loss=0.2364, pruned_loss=0.03265, over 7291.00 frames. ], tot_loss[loss=0.1417, simple_loss=0.2213, pruned_loss=0.031, over 1192563.32 frames. ], batch size: 70, lr: 6.40e-03, grad_scale: 8.0 +2023-03-21 03:47:56,019 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=70964.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:48:05,460 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 03:48:10,061 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=70992.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:48:14,666 INFO [train.py:901] (0/2) Epoch 26, batch 400, loss[loss=0.1528, simple_loss=0.2286, pruned_loss=0.03852, over 7260.00 frames. ], tot_loss[loss=0.1416, simple_loss=0.2214, pruned_loss=0.03092, over 1249672.25 frames. ], batch size: 47, lr: 6.40e-03, grad_scale: 8.0 +2023-03-21 03:48:29,728 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.349e+02 1.976e+02 2.210e+02 2.576e+02 5.239e+02, threshold=4.420e+02, percent-clipped=2.0 +2023-03-21 03:48:29,905 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=71030.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:48:40,692 INFO [train.py:901] (0/2) Epoch 26, batch 450, loss[loss=0.1467, simple_loss=0.2184, pruned_loss=0.03747, over 7224.00 frames. ], tot_loss[loss=0.1415, simple_loss=0.2209, pruned_loss=0.03106, over 1291532.73 frames. ], batch size: 45, lr: 6.40e-03, grad_scale: 8.0 +2023-03-21 03:48:40,822 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4949, 2.7249, 3.2142, 3.4954, 3.4896, 3.5687, 3.3001, 3.3857], + device='cuda:0'), covar=tensor([0.0033, 0.0132, 0.0048, 0.0032, 0.0032, 0.0030, 0.0074, 0.0057], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0061, 0.0052, 0.0049, 0.0049, 0.0053, 0.0046, 0.0067], + device='cuda:0'), out_proj_covar=tensor([8.1705e-05, 1.3756e-04, 1.0575e-04, 9.4841e-05, 9.1688e-05, 1.0060e-04, + 9.7938e-05, 1.3356e-04], device='cuda:0') +2023-03-21 03:48:41,875 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=71053.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:48:46,201 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 03:48:46,215 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 03:49:01,670 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=71091.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:49:05,857 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.76 vs. limit=5.0 +2023-03-21 03:49:06,538 INFO [train.py:901] (0/2) Epoch 26, batch 500, loss[loss=0.1633, simple_loss=0.2393, pruned_loss=0.04363, over 7256.00 frames. ], tot_loss[loss=0.1422, simple_loss=0.2214, pruned_loss=0.03145, over 1324673.23 frames. ], batch size: 55, lr: 6.40e-03, grad_scale: 8.0 +2023-03-21 03:49:19,603 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 03:49:20,247 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5797, 2.8255, 2.5212, 2.8517, 2.9057, 2.5703, 2.8615, 2.7011], + device='cuda:0'), covar=tensor([0.0821, 0.0751, 0.1242, 0.0989, 0.0962, 0.0646, 0.1129, 0.1269], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0051, 0.0060, 0.0053, 0.0051, 0.0054, 0.0053, 0.0048], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:49:21,079 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.306e+02 1.818e+02 2.239e+02 2.782e+02 4.854e+02, threshold=4.477e+02, percent-clipped=3.0 +2023-03-21 03:49:21,134 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 03:49:21,628 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 03:49:23,670 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 03:49:28,800 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 03:49:32,341 INFO [train.py:901] (0/2) Epoch 26, batch 550, loss[loss=0.1327, simple_loss=0.2152, pruned_loss=0.02507, over 7279.00 frames. ], tot_loss[loss=0.1415, simple_loss=0.2208, pruned_loss=0.03114, over 1349228.36 frames. ], batch size: 52, lr: 6.40e-03, grad_scale: 8.0 +2023-03-21 03:49:41,180 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 03:49:41,274 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4979, 2.5792, 3.5189, 3.5601, 3.5722, 3.5620, 3.3837, 3.1684], + device='cuda:0'), covar=tensor([0.0084, 0.0278, 0.0080, 0.0071, 0.0071, 0.0075, 0.0123, 0.0130], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0061, 0.0052, 0.0049, 0.0049, 0.0053, 0.0046, 0.0067], + device='cuda:0'), out_proj_covar=tensor([8.2244e-05, 1.3723e-04, 1.0629e-04, 9.4901e-05, 9.1979e-05, 9.9862e-05, + 9.7413e-05, 1.3358e-04], device='cuda:0') +2023-03-21 03:49:46,388 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3446, 2.9657, 2.2883, 3.7508, 2.6898, 3.1505, 1.6427, 2.2703], + device='cuda:0'), covar=tensor([0.0426, 0.0784, 0.2522, 0.0562, 0.0559, 0.0624, 0.3605, 0.1975], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0252, 0.0296, 0.0268, 0.0272, 0.0266, 0.0252, 0.0271], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:49:49,356 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 03:49:52,900 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 03:49:58,799 INFO [train.py:901] (0/2) Epoch 26, batch 600, loss[loss=0.129, simple_loss=0.2043, pruned_loss=0.02684, over 7235.00 frames. ], tot_loss[loss=0.1417, simple_loss=0.2212, pruned_loss=0.03111, over 1373186.87 frames. ], batch size: 45, lr: 6.39e-03, grad_scale: 8.0 +2023-03-21 03:50:00,838 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 03:50:07,519 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6483, 3.7128, 2.8945, 3.2629, 2.3823, 2.2173, 1.9589, 3.6932], + device='cuda:0'), covar=tensor([0.0041, 0.0033, 0.0118, 0.0058, 0.0180, 0.0415, 0.0461, 0.0043], + device='cuda:0'), in_proj_covar=tensor([0.0097, 0.0079, 0.0101, 0.0086, 0.0115, 0.0122, 0.0119, 0.0092], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 03:50:14,064 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.205e+02 1.854e+02 2.159e+02 2.545e+02 4.866e+02, threshold=4.318e+02, percent-clipped=2.0 +2023-03-21 03:50:17,078 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 03:50:24,644 INFO [train.py:901] (0/2) Epoch 26, batch 650, loss[loss=0.1327, simple_loss=0.2143, pruned_loss=0.02556, over 7313.00 frames. ], tot_loss[loss=0.1411, simple_loss=0.2205, pruned_loss=0.03089, over 1387690.92 frames. ], batch size: 49, lr: 6.39e-03, grad_scale: 8.0 +2023-03-21 03:50:25,647 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 03:50:43,920 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 03:50:50,418 INFO [train.py:901] (0/2) Epoch 26, batch 700, loss[loss=0.15, simple_loss=0.2298, pruned_loss=0.0351, over 7264.00 frames. ], tot_loss[loss=0.141, simple_loss=0.2202, pruned_loss=0.0309, over 1399186.58 frames. ], batch size: 77, lr: 6.39e-03, grad_scale: 8.0 +2023-03-21 03:50:52,875 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 03:50:52,938 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.0117, 4.4959, 4.4665, 5.0238, 4.8603, 4.9124, 4.2668, 4.4803], + device='cuda:0'), covar=tensor([0.0737, 0.2546, 0.2309, 0.0957, 0.0881, 0.1226, 0.0809, 0.1092], + device='cuda:0'), in_proj_covar=tensor([0.0233, 0.0358, 0.0279, 0.0277, 0.0206, 0.0343, 0.0211, 0.0252], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:51:05,619 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.388e+02 1.831e+02 2.185e+02 2.577e+02 5.485e+02, threshold=4.371e+02, percent-clipped=2.0 +2023-03-21 03:51:14,677 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=71348.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:51:16,110 INFO [train.py:901] (0/2) Epoch 26, batch 750, loss[loss=0.1099, simple_loss=0.1784, pruned_loss=0.02067, over 7055.00 frames. ], tot_loss[loss=0.1411, simple_loss=0.2205, pruned_loss=0.03086, over 1409147.73 frames. ], batch size: 35, lr: 6.39e-03, grad_scale: 8.0 +2023-03-21 03:51:16,127 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 03:51:16,598 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 03:51:31,862 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 03:51:34,392 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=71386.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:51:36,348 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 03:51:42,846 INFO [train.py:901] (0/2) Epoch 26, batch 800, loss[loss=0.1449, simple_loss=0.2267, pruned_loss=0.03157, over 7324.00 frames. ], tot_loss[loss=0.1415, simple_loss=0.2212, pruned_loss=0.03096, over 1417090.96 frames. ], batch size: 61, lr: 6.38e-03, grad_scale: 8.0 +2023-03-21 03:51:42,863 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 03:51:44,412 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 03:51:55,050 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 03:51:57,556 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.309e+02 1.848e+02 2.128e+02 2.521e+02 4.378e+02, threshold=4.256e+02, percent-clipped=1.0 +2023-03-21 03:52:08,880 INFO [train.py:901] (0/2) Epoch 26, batch 850, loss[loss=0.1425, simple_loss=0.2276, pruned_loss=0.02874, over 7273.00 frames. ], tot_loss[loss=0.1422, simple_loss=0.2218, pruned_loss=0.03128, over 1422158.36 frames. ], batch size: 66, lr: 6.38e-03, grad_scale: 8.0 +2023-03-21 03:52:14,493 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 03:52:14,497 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 03:52:20,084 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 03:52:24,043 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 03:52:34,604 INFO [train.py:901] (0/2) Epoch 26, batch 900, loss[loss=0.144, simple_loss=0.2219, pruned_loss=0.03299, over 7288.00 frames. ], tot_loss[loss=0.1419, simple_loss=0.2216, pruned_loss=0.0311, over 1427929.93 frames. ], batch size: 57, lr: 6.38e-03, grad_scale: 8.0 +2023-03-21 03:52:38,187 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3681, 1.1531, 1.4936, 1.7009, 1.4482, 1.7377, 1.3944, 1.7729], + device='cuda:0'), covar=tensor([0.1169, 0.3522, 0.1108, 0.0762, 0.1202, 0.1211, 0.1836, 0.1335], + device='cuda:0'), in_proj_covar=tensor([0.0063, 0.0068, 0.0054, 0.0050, 0.0054, 0.0053, 0.0083, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:52:49,102 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.291e+02 1.869e+02 2.202e+02 2.598e+02 5.253e+02, threshold=4.404e+02, percent-clipped=1.0 +2023-03-21 03:52:51,459 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4111, 2.2892, 1.9791, 3.5234, 1.4938, 3.2436, 1.3177, 3.0916], + device='cuda:0'), covar=tensor([0.0176, 0.1306, 0.1908, 0.0181, 0.4769, 0.0217, 0.1317, 0.0350], + device='cuda:0'), in_proj_covar=tensor([0.0196, 0.0256, 0.0272, 0.0199, 0.0262, 0.0205, 0.0250, 0.0228], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:52:57,927 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8477, 2.5369, 2.9655, 2.8544, 2.7526, 2.7903, 2.3714, 2.8012], + device='cuda:0'), covar=tensor([0.1683, 0.0932, 0.1366, 0.1312, 0.1199, 0.1146, 0.2358, 0.1893], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0061, 0.0045, 0.0045, 0.0045, 0.0042, 0.0062, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:53:00,349 INFO [train.py:901] (0/2) Epoch 26, batch 950, loss[loss=0.1371, simple_loss=0.2215, pruned_loss=0.02634, over 7338.00 frames. ], tot_loss[loss=0.1427, simple_loss=0.2221, pruned_loss=0.03167, over 1432091.19 frames. ], batch size: 59, lr: 6.38e-03, grad_scale: 8.0 +2023-03-21 03:53:02,394 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 03:53:06,547 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6298, 3.0536, 2.6874, 2.9184, 2.8553, 2.4607, 3.0785, 2.8320], + device='cuda:0'), covar=tensor([0.1221, 0.0535, 0.0999, 0.1059, 0.1690, 0.0971, 0.0984, 0.0759], + device='cuda:0'), in_proj_covar=tensor([0.0052, 0.0051, 0.0059, 0.0052, 0.0050, 0.0053, 0.0051, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:53:24,408 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 03:53:26,268 INFO [train.py:901] (0/2) Epoch 26, batch 1000, loss[loss=0.1396, simple_loss=0.2187, pruned_loss=0.03024, over 7323.00 frames. ], tot_loss[loss=0.1428, simple_loss=0.2227, pruned_loss=0.03147, over 1436581.32 frames. ], batch size: 59, lr: 6.38e-03, grad_scale: 8.0 +2023-03-21 03:53:34,066 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=71616.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:53:41,505 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.311e+02 1.757e+02 2.158e+02 2.500e+02 4.041e+02, threshold=4.316e+02, percent-clipped=0.0 +2023-03-21 03:53:45,556 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 03:53:50,665 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=71648.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:53:52,001 INFO [train.py:901] (0/2) Epoch 26, batch 1050, loss[loss=0.1407, simple_loss=0.2316, pruned_loss=0.02492, over 6701.00 frames. ], tot_loss[loss=0.1428, simple_loss=0.2227, pruned_loss=0.03143, over 1438428.78 frames. ], batch size: 107, lr: 6.37e-03, grad_scale: 8.0 +2023-03-21 03:54:05,926 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=71677.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:54:08,276 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 03:54:10,375 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=71686.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:54:12,817 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 03:54:15,292 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=71696.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:54:17,541 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.29 vs. limit=2.0 +2023-03-21 03:54:17,755 INFO [train.py:901] (0/2) Epoch 26, batch 1100, loss[loss=0.1334, simple_loss=0.2113, pruned_loss=0.02779, over 7343.00 frames. ], tot_loss[loss=0.143, simple_loss=0.2228, pruned_loss=0.03163, over 1439085.23 frames. ], batch size: 54, lr: 6.37e-03, grad_scale: 8.0 +2023-03-21 03:54:27,686 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0485, 3.9434, 3.8510, 3.8992, 3.7509, 3.7213, 3.8819, 4.0742], + device='cuda:0'), covar=tensor([0.0322, 0.0320, 0.0337, 0.0364, 0.0525, 0.0570, 0.0694, 0.0509], + device='cuda:0'), in_proj_covar=tensor([0.0115, 0.0120, 0.0108, 0.0114, 0.0106, 0.0096, 0.0096, 0.0091], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:54:33,010 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.196e+02 1.836e+02 2.092e+02 2.603e+02 5.313e+02, threshold=4.183e+02, percent-clipped=2.0 +2023-03-21 03:54:34,653 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7767, 3.7516, 3.6536, 3.7945, 3.5401, 3.5705, 3.8157, 3.9352], + device='cuda:0'), covar=tensor([0.0309, 0.0275, 0.0300, 0.0271, 0.0408, 0.0456, 0.0424, 0.0354], + device='cuda:0'), in_proj_covar=tensor([0.0114, 0.0119, 0.0107, 0.0113, 0.0106, 0.0096, 0.0095, 0.0090], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:54:35,103 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=71734.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:54:42,631 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 03:54:42,644 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 03:54:44,134 INFO [train.py:901] (0/2) Epoch 26, batch 1150, loss[loss=0.1316, simple_loss=0.2179, pruned_loss=0.0226, over 7289.00 frames. ], tot_loss[loss=0.143, simple_loss=0.2227, pruned_loss=0.03169, over 1437943.70 frames. ], batch size: 86, lr: 6.37e-03, grad_scale: 8.0 +2023-03-21 03:54:54,746 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 03:54:55,260 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 03:55:10,340 INFO [train.py:901] (0/2) Epoch 26, batch 1200, loss[loss=0.1437, simple_loss=0.2256, pruned_loss=0.0309, over 7301.00 frames. ], tot_loss[loss=0.1424, simple_loss=0.2221, pruned_loss=0.03131, over 1438758.06 frames. ], batch size: 80, lr: 6.37e-03, grad_scale: 8.0 +2023-03-21 03:55:24,900 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.304e+02 1.827e+02 2.248e+02 2.679e+02 4.033e+02, threshold=4.496e+02, percent-clipped=0.0 +2023-03-21 03:55:27,477 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 03:55:35,983 INFO [train.py:901] (0/2) Epoch 26, batch 1250, loss[loss=0.1127, simple_loss=0.1936, pruned_loss=0.01595, over 7240.00 frames. ], tot_loss[loss=0.1431, simple_loss=0.2228, pruned_loss=0.03166, over 1438681.90 frames. ], batch size: 39, lr: 6.36e-03, grad_scale: 16.0 +2023-03-21 03:55:47,062 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9318, 3.4770, 3.8644, 4.0256, 3.8709, 3.9542, 3.9624, 3.8614], + device='cuda:0'), covar=tensor([0.0033, 0.0083, 0.0030, 0.0025, 0.0030, 0.0026, 0.0029, 0.0042], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0060, 0.0051, 0.0049, 0.0048, 0.0052, 0.0045, 0.0066], + device='cuda:0'), out_proj_covar=tensor([8.0018e-05, 1.3341e-04, 1.0406e-04, 9.3459e-05, 9.1177e-05, 9.8063e-05, + 9.5243e-05, 1.3113e-04], device='cuda:0') +2023-03-21 03:55:51,145 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 03:55:54,723 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 03:55:56,253 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 03:56:01,834 INFO [train.py:901] (0/2) Epoch 26, batch 1300, loss[loss=0.1475, simple_loss=0.2318, pruned_loss=0.0316, over 7337.00 frames. ], tot_loss[loss=0.143, simple_loss=0.2228, pruned_loss=0.03158, over 1439949.51 frames. ], batch size: 61, lr: 6.36e-03, grad_scale: 16.0 +2023-03-21 03:56:11,682 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=71919.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:56:17,018 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.296e+02 1.780e+02 2.070e+02 2.538e+02 4.671e+02, threshold=4.140e+02, percent-clipped=1.0 +2023-03-21 03:56:19,534 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 03:56:21,567 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 03:56:24,988 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 03:56:27,525 INFO [train.py:901] (0/2) Epoch 26, batch 1350, loss[loss=0.1507, simple_loss=0.2303, pruned_loss=0.03556, over 7249.00 frames. ], tot_loss[loss=0.1425, simple_loss=0.2224, pruned_loss=0.03134, over 1441472.27 frames. ], batch size: 55, lr: 6.36e-03, grad_scale: 16.0 +2023-03-21 03:56:36,295 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 03:56:38,782 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=71972.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:56:42,922 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=71980.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:56:53,152 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-72000.pt +2023-03-21 03:56:57,018 INFO [train.py:901] (0/2) Epoch 26, batch 1400, loss[loss=0.1493, simple_loss=0.2319, pruned_loss=0.03332, over 7379.00 frames. ], tot_loss[loss=0.1422, simple_loss=0.2219, pruned_loss=0.0313, over 1442534.69 frames. ], batch size: 65, lr: 6.36e-03, grad_scale: 16.0 +2023-03-21 03:57:11,281 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 03:57:12,290 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.375e+02 1.905e+02 2.148e+02 2.727e+02 3.705e+02, threshold=4.295e+02, percent-clipped=0.0 +2023-03-21 03:57:16,530 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0406, 3.0215, 2.1735, 3.4812, 2.5918, 3.0888, 1.4829, 2.1454], + device='cuda:0'), covar=tensor([0.0406, 0.0803, 0.2229, 0.0589, 0.0500, 0.0627, 0.3125, 0.1622], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0253, 0.0293, 0.0268, 0.0269, 0.0266, 0.0250, 0.0268], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:57:23,438 INFO [train.py:901] (0/2) Epoch 26, batch 1450, loss[loss=0.1543, simple_loss=0.2317, pruned_loss=0.03841, over 7271.00 frames. ], tot_loss[loss=0.1421, simple_loss=0.2218, pruned_loss=0.03119, over 1441028.28 frames. ], batch size: 64, lr: 6.36e-03, grad_scale: 16.0 +2023-03-21 03:57:35,585 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0456, 2.9223, 3.3580, 3.1177, 3.4914, 3.1876, 2.7565, 3.3522], + device='cuda:0'), covar=tensor([0.2198, 0.0762, 0.1252, 0.1488, 0.0623, 0.1299, 0.2148, 0.1373], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0060, 0.0045, 0.0045, 0.0044, 0.0042, 0.0062, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 03:57:36,123 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5929, 3.2451, 2.7319, 3.2038, 3.0531, 2.7865, 3.0864, 2.8438], + device='cuda:0'), covar=tensor([0.0901, 0.0875, 0.1336, 0.0737, 0.1247, 0.0937, 0.1003, 0.1133], + device='cuda:0'), in_proj_covar=tensor([0.0052, 0.0050, 0.0059, 0.0052, 0.0051, 0.0053, 0.0051, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 03:57:36,992 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 03:57:49,181 INFO [train.py:901] (0/2) Epoch 26, batch 1500, loss[loss=0.1506, simple_loss=0.2264, pruned_loss=0.03736, over 7305.00 frames. ], tot_loss[loss=0.1426, simple_loss=0.2222, pruned_loss=0.03151, over 1440687.58 frames. ], batch size: 80, lr: 6.35e-03, grad_scale: 16.0 +2023-03-21 03:57:52,237 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 03:58:01,180 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.95 vs. limit=2.0 +2023-03-21 03:58:03,440 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=72129.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:58:03,794 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.319e+02 1.929e+02 2.352e+02 2.805e+02 7.015e+02, threshold=4.703e+02, percent-clipped=5.0 +2023-03-21 03:58:15,072 INFO [train.py:901] (0/2) Epoch 26, batch 1550, loss[loss=0.1318, simple_loss=0.2067, pruned_loss=0.02847, over 7265.00 frames. ], tot_loss[loss=0.1419, simple_loss=0.2216, pruned_loss=0.03115, over 1440090.63 frames. ], batch size: 47, lr: 6.35e-03, grad_scale: 16.0 +2023-03-21 03:58:17,061 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 03:58:35,466 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=72190.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:58:41,186 INFO [train.py:901] (0/2) Epoch 26, batch 1600, loss[loss=0.1119, simple_loss=0.1788, pruned_loss=0.02244, over 7010.00 frames. ], tot_loss[loss=0.1425, simple_loss=0.222, pruned_loss=0.03156, over 1439915.39 frames. ], batch size: 35, lr: 6.35e-03, grad_scale: 16.0 +2023-03-21 03:58:43,462 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.75 vs. limit=2.0 +2023-03-21 03:58:47,731 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 03:58:48,736 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 03:58:52,868 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 03:58:56,361 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.220e+02 1.815e+02 2.080e+02 2.642e+02 5.040e+02, threshold=4.161e+02, percent-clipped=1.0 +2023-03-21 03:59:02,419 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 03:59:06,425 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 03:59:06,852 INFO [train.py:901] (0/2) Epoch 26, batch 1650, loss[loss=0.1784, simple_loss=0.2472, pruned_loss=0.05484, over 7211.00 frames. ], tot_loss[loss=0.1425, simple_loss=0.2218, pruned_loss=0.03166, over 1438718.81 frames. ], batch size: 93, lr: 6.35e-03, grad_scale: 16.0 +2023-03-21 03:59:08,328 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.62 vs. limit=2.0 +2023-03-21 03:59:15,622 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 03:59:18,183 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=72272.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:59:19,639 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=72275.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:59:31,532 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 03:59:32,565 INFO [train.py:901] (0/2) Epoch 26, batch 1700, loss[loss=0.1538, simple_loss=0.2257, pruned_loss=0.04089, over 7271.00 frames. ], tot_loss[loss=0.1423, simple_loss=0.2218, pruned_loss=0.03142, over 1440317.36 frames. ], batch size: 52, lr: 6.34e-03, grad_scale: 16.0 +2023-03-21 03:59:35,109 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 03:59:42,996 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=72320.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:59:45,987 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 03:59:46,086 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=72326.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 03:59:47,968 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.249e+02 1.865e+02 2.256e+02 2.622e+02 4.805e+02, threshold=4.513e+02, percent-clipped=4.0 +2023-03-21 03:59:59,234 INFO [train.py:901] (0/2) Epoch 26, batch 1750, loss[loss=0.1473, simple_loss=0.2324, pruned_loss=0.03114, over 7309.00 frames. ], tot_loss[loss=0.1423, simple_loss=0.2219, pruned_loss=0.03135, over 1441676.40 frames. ], batch size: 83, lr: 6.34e-03, grad_scale: 16.0 +2023-03-21 04:00:01,310 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4596, 4.2221, 4.5312, 4.5411, 4.4553, 4.5274, 4.4283, 4.4717], + device='cuda:0'), covar=tensor([0.0029, 0.0060, 0.0024, 0.0026, 0.0027, 0.0029, 0.0025, 0.0045], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0061, 0.0052, 0.0050, 0.0049, 0.0054, 0.0045, 0.0067], + device='cuda:0'), out_proj_covar=tensor([8.1798e-05, 1.3569e-04, 1.0479e-04, 9.6719e-05, 9.3318e-05, 1.0170e-04, + 9.5662e-05, 1.3482e-04], device='cuda:0') +2023-03-21 04:00:09,736 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 04:00:10,719 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 04:00:16,181 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.80 vs. limit=2.0 +2023-03-21 04:00:17,608 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=72387.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:00:24,534 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.08 vs. limit=5.0 +2023-03-21 04:00:25,534 INFO [train.py:901] (0/2) Epoch 26, batch 1800, loss[loss=0.155, simple_loss=0.2368, pruned_loss=0.03659, over 7270.00 frames. ], tot_loss[loss=0.1424, simple_loss=0.2222, pruned_loss=0.03135, over 1443381.10 frames. ], batch size: 89, lr: 6.34e-03, grad_scale: 16.0 +2023-03-21 04:00:33,751 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 04:00:40,212 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.401e+02 1.894e+02 2.228e+02 2.651e+02 4.777e+02, threshold=4.456e+02, percent-clipped=1.0 +2023-03-21 04:00:47,438 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 04:00:50,535 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4842, 4.9587, 5.1018, 4.9919, 4.8365, 4.4855, 5.1094, 4.8701], + device='cuda:0'), covar=tensor([0.0412, 0.0360, 0.0285, 0.0412, 0.0320, 0.0355, 0.0253, 0.0461], + device='cuda:0'), in_proj_covar=tensor([0.0199, 0.0230, 0.0176, 0.0175, 0.0141, 0.0207, 0.0179, 0.0135], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 04:00:51,409 INFO [train.py:901] (0/2) Epoch 26, batch 1850, loss[loss=0.1323, simple_loss=0.2175, pruned_loss=0.02359, over 7307.00 frames. ], tot_loss[loss=0.1422, simple_loss=0.2221, pruned_loss=0.03113, over 1442292.13 frames. ], batch size: 49, lr: 6.34e-03, grad_scale: 16.0 +2023-03-21 04:00:56,890 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 04:01:09,214 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=72485.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:01:14,189 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 04:01:17,130 INFO [train.py:901] (0/2) Epoch 26, batch 1900, loss[loss=0.1347, simple_loss=0.2026, pruned_loss=0.03336, over 7315.00 frames. ], tot_loss[loss=0.1413, simple_loss=0.2211, pruned_loss=0.03071, over 1441756.16 frames. ], batch size: 49, lr: 6.34e-03, grad_scale: 16.0 +2023-03-21 04:01:32,368 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.186e+02 1.886e+02 2.210e+02 2.551e+02 4.752e+02, threshold=4.419e+02, percent-clipped=1.0 +2023-03-21 04:01:40,017 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 04:01:41,718 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9239, 2.5721, 2.8941, 2.7889, 2.4706, 2.5067, 3.0169, 2.1800], + device='cuda:0'), covar=tensor([0.0504, 0.0543, 0.0561, 0.0537, 0.0649, 0.0799, 0.0601, 0.1621], + device='cuda:0'), in_proj_covar=tensor([0.0336, 0.0337, 0.0269, 0.0362, 0.0305, 0.0298, 0.0344, 0.0277], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:01:42,962 INFO [train.py:901] (0/2) Epoch 26, batch 1950, loss[loss=0.1533, simple_loss=0.2394, pruned_loss=0.03364, over 7328.00 frames. ], tot_loss[loss=0.141, simple_loss=0.2209, pruned_loss=0.03048, over 1442938.64 frames. ], batch size: 75, lr: 6.33e-03, grad_scale: 16.0 +2023-03-21 04:01:48,773 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3340, 2.6277, 1.9495, 3.2787, 2.9638, 2.8177, 3.0673, 2.6948], + device='cuda:0'), covar=tensor([0.1798, 0.0821, 0.3606, 0.0517, 0.0225, 0.0197, 0.0359, 0.0399], + device='cuda:0'), in_proj_covar=tensor([0.0260, 0.0234, 0.0262, 0.0262, 0.0180, 0.0184, 0.0206, 0.0216], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:01:51,036 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 04:01:55,843 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=72575.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:01:56,245 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 04:01:56,787 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 04:01:56,887 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8782, 3.8975, 3.1307, 3.3783, 2.7911, 2.2502, 1.8042, 3.8714], + device='cuda:0'), covar=tensor([0.0046, 0.0044, 0.0113, 0.0075, 0.0149, 0.0443, 0.0517, 0.0057], + device='cuda:0'), in_proj_covar=tensor([0.0096, 0.0078, 0.0099, 0.0085, 0.0113, 0.0119, 0.0118, 0.0092], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 04:01:56,901 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8223, 2.4178, 2.9300, 2.6472, 2.7872, 2.7025, 2.3465, 2.9115], + device='cuda:0'), covar=tensor([0.1433, 0.0757, 0.1109, 0.1612, 0.0983, 0.1000, 0.1908, 0.1152], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0061, 0.0046, 0.0046, 0.0045, 0.0042, 0.0062, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:02:09,058 INFO [train.py:901] (0/2) Epoch 26, batch 2000, loss[loss=0.144, simple_loss=0.2256, pruned_loss=0.03124, over 7371.00 frames. ], tot_loss[loss=0.1407, simple_loss=0.2207, pruned_loss=0.03031, over 1442417.08 frames. ], batch size: 65, lr: 6.33e-03, grad_scale: 16.0 +2023-03-21 04:02:09,682 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0891, 3.9575, 3.9078, 3.8759, 3.8493, 3.6767, 4.0521, 3.5311], + device='cuda:0'), covar=tensor([0.0182, 0.0124, 0.0129, 0.0175, 0.0387, 0.0140, 0.0150, 0.0193], + device='cuda:0'), in_proj_covar=tensor([0.0083, 0.0090, 0.0089, 0.0080, 0.0155, 0.0100, 0.0094, 0.0099], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:02:13,734 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 04:02:20,889 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=72623.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:02:24,305 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.326e+02 1.824e+02 2.165e+02 2.550e+02 4.522e+02, threshold=4.329e+02, percent-clipped=1.0 +2023-03-21 04:02:24,348 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 04:02:26,011 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5565, 3.0076, 3.4124, 3.2699, 2.9623, 2.9183, 3.4345, 2.5916], + device='cuda:0'), covar=tensor([0.0341, 0.0386, 0.0489, 0.0478, 0.0559, 0.0774, 0.0551, 0.1628], + device='cuda:0'), in_proj_covar=tensor([0.0333, 0.0334, 0.0268, 0.0359, 0.0302, 0.0296, 0.0340, 0.0275], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:02:31,901 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 04:02:34,900 INFO [train.py:901] (0/2) Epoch 26, batch 2050, loss[loss=0.1434, simple_loss=0.2255, pruned_loss=0.03071, over 7352.00 frames. ], tot_loss[loss=0.1412, simple_loss=0.2211, pruned_loss=0.0306, over 1444112.20 frames. ], batch size: 61, lr: 6.33e-03, grad_scale: 16.0 +2023-03-21 04:02:36,509 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3312, 4.0987, 4.1673, 4.1741, 4.1514, 4.0242, 4.3190, 3.8268], + device='cuda:0'), covar=tensor([0.0145, 0.0125, 0.0102, 0.0146, 0.0299, 0.0107, 0.0132, 0.0177], + device='cuda:0'), in_proj_covar=tensor([0.0083, 0.0090, 0.0089, 0.0080, 0.0155, 0.0100, 0.0094, 0.0100], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:02:40,034 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.81 vs. limit=5.0 +2023-03-21 04:02:51,394 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=72682.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:03:01,490 INFO [train.py:901] (0/2) Epoch 26, batch 2100, loss[loss=0.1549, simple_loss=0.2314, pruned_loss=0.03926, over 7306.00 frames. ], tot_loss[loss=0.1406, simple_loss=0.2206, pruned_loss=0.03031, over 1443788.97 frames. ], batch size: 83, lr: 6.33e-03, grad_scale: 8.0 +2023-03-21 04:03:06,499 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 04:03:09,538 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 04:03:16,557 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.317e+02 1.881e+02 2.177e+02 2.542e+02 4.877e+02, threshold=4.354e+02, percent-clipped=2.0 +2023-03-21 04:03:27,375 INFO [train.py:901] (0/2) Epoch 26, batch 2150, loss[loss=0.1338, simple_loss=0.2181, pruned_loss=0.02473, over 7309.00 frames. ], tot_loss[loss=0.1408, simple_loss=0.2208, pruned_loss=0.03041, over 1443948.95 frames. ], batch size: 80, lr: 6.32e-03, grad_scale: 8.0 +2023-03-21 04:03:27,652 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-03-21 04:03:45,265 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=72785.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:03:53,545 INFO [train.py:901] (0/2) Epoch 26, batch 2200, loss[loss=0.1741, simple_loss=0.2443, pruned_loss=0.05192, over 7339.00 frames. ], tot_loss[loss=0.1401, simple_loss=0.2197, pruned_loss=0.03026, over 1439802.48 frames. ], batch size: 75, lr: 6.32e-03, grad_scale: 8.0 +2023-03-21 04:03:56,590 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 04:04:09,309 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.170e+02 1.820e+02 2.213e+02 2.627e+02 5.046e+02, threshold=4.426e+02, percent-clipped=1.0 +2023-03-21 04:04:10,392 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=72833.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:04:19,359 INFO [train.py:901] (0/2) Epoch 26, batch 2250, loss[loss=0.1421, simple_loss=0.2247, pruned_loss=0.02972, over 7336.00 frames. ], tot_loss[loss=0.1406, simple_loss=0.2203, pruned_loss=0.03039, over 1439856.50 frames. ], batch size: 54, lr: 6.32e-03, grad_scale: 8.0 +2023-03-21 04:04:24,768 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.69 vs. limit=2.0 +2023-03-21 04:04:30,577 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 04:04:31,040 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 04:04:41,832 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=72894.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:04:43,181 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 04:04:45,204 INFO [train.py:901] (0/2) Epoch 26, batch 2300, loss[loss=0.1182, simple_loss=0.1934, pruned_loss=0.02145, over 7177.00 frames. ], tot_loss[loss=0.1409, simple_loss=0.2206, pruned_loss=0.03058, over 1441454.46 frames. ], batch size: 39, lr: 6.32e-03, grad_scale: 8.0 +2023-03-21 04:05:01,078 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.198e+02 1.738e+02 1.984e+02 2.359e+02 4.663e+02, threshold=3.968e+02, percent-clipped=1.0 +2023-03-21 04:05:11,640 INFO [train.py:901] (0/2) Epoch 26, batch 2350, loss[loss=0.123, simple_loss=0.2029, pruned_loss=0.02156, over 7134.00 frames. ], tot_loss[loss=0.1408, simple_loss=0.2204, pruned_loss=0.03062, over 1441882.84 frames. ], batch size: 41, lr: 6.32e-03, grad_scale: 8.0 +2023-03-21 04:05:13,827 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=72955.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:05:18,416 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=72964.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 04:05:22,284 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8235, 3.4524, 2.9208, 3.2640, 3.2512, 2.7965, 3.2420, 2.9574], + device='cuda:0'), covar=tensor([0.1276, 0.0402, 0.0866, 0.0789, 0.0827, 0.0911, 0.0956, 0.1601], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0052, 0.0060, 0.0053, 0.0051, 0.0054, 0.0052, 0.0048], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:05:27,430 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=72982.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:05:30,297 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 04:05:31,474 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6018, 2.6002, 2.2365, 3.5409, 3.4931, 3.0340, 3.1684, 3.1311], + device='cuda:0'), covar=tensor([0.1860, 0.0499, 0.3546, 0.0630, 0.0225, 0.0151, 0.0303, 0.0409], + device='cuda:0'), in_proj_covar=tensor([0.0262, 0.0235, 0.0261, 0.0264, 0.0179, 0.0182, 0.0207, 0.0216], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:05:36,672 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 04:05:37,116 INFO [train.py:901] (0/2) Epoch 26, batch 2400, loss[loss=0.14, simple_loss=0.2198, pruned_loss=0.03014, over 7291.00 frames. ], tot_loss[loss=0.1414, simple_loss=0.2211, pruned_loss=0.0309, over 1442561.88 frames. ], batch size: 86, lr: 6.31e-03, grad_scale: 8.0 +2023-03-21 04:05:43,442 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5066, 2.6834, 3.3839, 3.3710, 3.4595, 3.6304, 3.3741, 3.4156], + device='cuda:0'), covar=tensor([0.0030, 0.0134, 0.0040, 0.0044, 0.0038, 0.0028, 0.0059, 0.0052], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0062, 0.0052, 0.0050, 0.0049, 0.0054, 0.0046, 0.0067], + device='cuda:0'), out_proj_covar=tensor([8.1429e-05, 1.3705e-04, 1.0594e-04, 9.5183e-05, 9.3142e-05, 1.0108e-04, + 9.7261e-05, 1.3460e-04], device='cuda:0') +2023-03-21 04:05:45,954 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5027, 2.4048, 2.1836, 3.5986, 1.6565, 3.4864, 1.2828, 3.1336], + device='cuda:0'), covar=tensor([0.0168, 0.1180, 0.1751, 0.0130, 0.3663, 0.0218, 0.1206, 0.0352], + device='cuda:0'), in_proj_covar=tensor([0.0196, 0.0253, 0.0270, 0.0199, 0.0260, 0.0206, 0.0244, 0.0228], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:05:46,827 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 04:05:49,368 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 04:05:49,970 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=73025.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 04:05:52,983 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=73030.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:05:53,418 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.197e+02 1.959e+02 2.204e+02 2.693e+02 9.099e+02, threshold=4.408e+02, percent-clipped=4.0 +2023-03-21 04:06:03,425 INFO [train.py:901] (0/2) Epoch 26, batch 2450, loss[loss=0.1075, simple_loss=0.1825, pruned_loss=0.01624, over 7049.00 frames. ], tot_loss[loss=0.1413, simple_loss=0.2209, pruned_loss=0.03083, over 1442302.65 frames. ], batch size: 35, lr: 6.31e-03, grad_scale: 8.0 +2023-03-21 04:06:16,149 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 04:06:29,492 INFO [train.py:901] (0/2) Epoch 26, batch 2500, loss[loss=0.1615, simple_loss=0.2299, pruned_loss=0.04654, over 7351.00 frames. ], tot_loss[loss=0.1416, simple_loss=0.2213, pruned_loss=0.031, over 1443058.96 frames. ], batch size: 63, lr: 6.31e-03, grad_scale: 8.0 +2023-03-21 04:06:37,231 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3003, 2.9989, 3.0874, 3.0544, 2.6590, 2.6614, 3.0176, 2.3148], + device='cuda:0'), covar=tensor([0.0460, 0.0462, 0.0473, 0.0520, 0.0568, 0.0731, 0.0409, 0.1533], + device='cuda:0'), in_proj_covar=tensor([0.0332, 0.0335, 0.0269, 0.0359, 0.0300, 0.0297, 0.0339, 0.0273], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:06:42,102 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 04:06:45,139 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.135e+02 1.825e+02 2.142e+02 2.392e+02 4.308e+02, threshold=4.285e+02, percent-clipped=0.0 +2023-03-21 04:06:53,772 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-03-21 04:06:55,367 INFO [train.py:901] (0/2) Epoch 26, batch 2550, loss[loss=0.1516, simple_loss=0.2314, pruned_loss=0.03592, over 7281.00 frames. ], tot_loss[loss=0.1411, simple_loss=0.2207, pruned_loss=0.03072, over 1442104.92 frames. ], batch size: 77, lr: 6.31e-03, grad_scale: 8.0 +2023-03-21 04:07:21,557 INFO [train.py:901] (0/2) Epoch 26, batch 2600, loss[loss=0.1351, simple_loss=0.2173, pruned_loss=0.02646, over 7279.00 frames. ], tot_loss[loss=0.1407, simple_loss=0.2203, pruned_loss=0.03058, over 1444215.66 frames. ], batch size: 57, lr: 6.31e-03, grad_scale: 8.0 +2023-03-21 04:07:34,168 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8274, 1.9619, 1.5131, 1.8762, 2.0514, 1.6886, 1.9878, 1.5370], + device='cuda:0'), covar=tensor([0.0085, 0.0186, 0.0258, 0.0120, 0.0086, 0.0128, 0.0094, 0.0129], + device='cuda:0'), in_proj_covar=tensor([0.0032, 0.0029, 0.0029, 0.0031, 0.0029, 0.0028, 0.0030, 0.0039], + device='cuda:0'), out_proj_covar=tensor([3.6282e-05, 3.2589e-05, 3.3088e-05, 3.5147e-05, 3.2935e-05, 3.1814e-05, + 3.4386e-05, 4.4619e-05], device='cuda:0') +2023-03-21 04:07:36,914 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.372e+02 1.805e+02 2.103e+02 2.646e+02 6.004e+02, threshold=4.206e+02, percent-clipped=1.0 +2023-03-21 04:07:46,231 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=73250.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:07:46,660 INFO [train.py:901] (0/2) Epoch 26, batch 2650, loss[loss=0.1115, simple_loss=0.1805, pruned_loss=0.02127, over 6979.00 frames. ], tot_loss[loss=0.141, simple_loss=0.2202, pruned_loss=0.03092, over 1442108.40 frames. ], batch size: 35, lr: 6.30e-03, grad_scale: 8.0 +2023-03-21 04:07:49,255 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9585, 2.7634, 3.4665, 3.0728, 3.0729, 3.0159, 2.5190, 3.1291], + device='cuda:0'), covar=tensor([0.2140, 0.0706, 0.0705, 0.1240, 0.0966, 0.1052, 0.1989, 0.1447], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0060, 0.0046, 0.0046, 0.0046, 0.0044, 0.0063, 0.0048], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:07:59,906 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.76 vs. limit=5.0 +2023-03-21 04:08:07,521 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=73293.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:08:11,334 INFO [train.py:901] (0/2) Epoch 26, batch 2700, loss[loss=0.1594, simple_loss=0.2459, pruned_loss=0.03642, over 7330.00 frames. ], tot_loss[loss=0.1417, simple_loss=0.2209, pruned_loss=0.03122, over 1439700.11 frames. ], batch size: 61, lr: 6.30e-03, grad_scale: 8.0 +2023-03-21 04:08:20,754 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=73320.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 04:08:26,060 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.284e+02 1.768e+02 2.025e+02 2.508e+02 4.489e+02, threshold=4.051e+02, percent-clipped=2.0 +2023-03-21 04:08:29,180 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0116, 3.9494, 3.2833, 3.4912, 2.8468, 2.2401, 1.8611, 3.9908], + device='cuda:0'), covar=tensor([0.0038, 0.0053, 0.0106, 0.0070, 0.0138, 0.0450, 0.0584, 0.0044], + device='cuda:0'), in_proj_covar=tensor([0.0098, 0.0080, 0.0102, 0.0086, 0.0115, 0.0123, 0.0121, 0.0093], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 04:08:36,034 INFO [train.py:901] (0/2) Epoch 26, batch 2750, loss[loss=0.1377, simple_loss=0.2187, pruned_loss=0.02837, over 7291.00 frames. ], tot_loss[loss=0.1416, simple_loss=0.2211, pruned_loss=0.031, over 1441711.50 frames. ], batch size: 57, lr: 6.30e-03, grad_scale: 8.0 +2023-03-21 04:08:37,626 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=73354.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:09:01,000 INFO [train.py:901] (0/2) Epoch 26, batch 2800, loss[loss=0.164, simple_loss=0.2381, pruned_loss=0.04495, over 7332.00 frames. ], tot_loss[loss=0.1419, simple_loss=0.2214, pruned_loss=0.0312, over 1442351.20 frames. ], batch size: 61, lr: 6.30e-03, grad_scale: 8.0 +2023-03-21 04:09:13,604 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-26.pt +2023-03-21 04:09:28,600 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 04:09:28,660 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 04:09:29,009 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 04:09:29,025 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 04:09:29,207 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 04:09:29,223 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 04:09:29,233 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 04:09:29,335 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 04:09:29,410 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 04:09:29,430 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 04:09:29,447 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 04:09:29,614 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 04:09:29,620 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 04:09:29,627 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 04:09:30,133 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 04:09:30,136 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 04:09:30,197 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 04:09:32,507 INFO [train.py:901] (0/2) Epoch 27, batch 0, loss[loss=0.1388, simple_loss=0.2173, pruned_loss=0.03012, over 7260.00 frames. ], tot_loss[loss=0.1388, simple_loss=0.2173, pruned_loss=0.03012, over 7260.00 frames. ], batch size: 47, lr: 6.18e-03, grad_scale: 8.0 +2023-03-21 04:09:32,508 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 04:09:39,377 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8225, 1.5641, 2.0985, 2.0594, 1.8847, 2.2023, 2.0421, 2.1621], + device='cuda:0'), covar=tensor([0.2394, 0.2723, 0.1433, 0.1116, 0.3313, 0.1669, 0.1567, 0.2991], + device='cuda:0'), in_proj_covar=tensor([0.0065, 0.0069, 0.0056, 0.0051, 0.0053, 0.0054, 0.0084, 0.0054], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:09:42,482 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.4706, 5.1050, 4.9058, 5.4297, 5.2071, 5.4789, 5.0447, 5.2497], + device='cuda:0'), covar=tensor([0.0393, 0.1716, 0.1487, 0.1010, 0.0744, 0.0699, 0.0448, 0.0581], + device='cuda:0'), in_proj_covar=tensor([0.0233, 0.0359, 0.0277, 0.0277, 0.0205, 0.0345, 0.0208, 0.0255], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:09:57,909 INFO [train.py:935] (0/2) Epoch 27, validation: loss=0.1647, simple_loss=0.2543, pruned_loss=0.03752, over 1622729.00 frames. +2023-03-21 04:09:57,910 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 04:10:00,981 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.354e+02 1.845e+02 2.089e+02 2.471e+02 4.051e+02, threshold=4.178e+02, percent-clipped=1.0 +2023-03-21 04:10:05,686 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 04:10:08,412 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-03-21 04:10:16,108 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 04:10:23,318 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 04:10:24,288 INFO [train.py:901] (0/2) Epoch 27, batch 50, loss[loss=0.1442, simple_loss=0.2278, pruned_loss=0.03031, over 7282.00 frames. ], tot_loss[loss=0.1429, simple_loss=0.2232, pruned_loss=0.03131, over 327377.48 frames. ], batch size: 57, lr: 6.18e-03, grad_scale: 8.0 +2023-03-21 04:10:25,317 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 04:10:28,330 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 04:10:32,433 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6340, 2.9792, 2.5592, 2.8702, 2.8722, 2.6268, 2.9071, 2.9270], + device='cuda:0'), covar=tensor([0.0734, 0.0702, 0.1316, 0.1052, 0.1713, 0.0554, 0.0729, 0.0853], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0051, 0.0059, 0.0052, 0.0050, 0.0053, 0.0050, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:10:44,387 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 04:10:44,885 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 04:10:45,195 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.88 vs. limit=2.0 +2023-03-21 04:10:50,029 INFO [train.py:901] (0/2) Epoch 27, batch 100, loss[loss=0.1305, simple_loss=0.209, pruned_loss=0.02605, over 7328.00 frames. ], tot_loss[loss=0.1423, simple_loss=0.2226, pruned_loss=0.03098, over 574021.78 frames. ], batch size: 54, lr: 6.18e-03, grad_scale: 8.0 +2023-03-21 04:10:52,946 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.130e+02 1.895e+02 2.298e+02 2.662e+02 5.511e+02, threshold=4.595e+02, percent-clipped=3.0 +2023-03-21 04:11:02,462 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=73550.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:11:13,048 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4238, 4.5215, 4.3448, 4.4782, 4.2150, 4.4792, 4.8004, 4.8534], + device='cuda:0'), covar=tensor([0.0147, 0.0126, 0.0171, 0.0150, 0.0252, 0.0196, 0.0164, 0.0116], + device='cuda:0'), in_proj_covar=tensor([0.0111, 0.0116, 0.0106, 0.0114, 0.0105, 0.0093, 0.0093, 0.0091], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:11:15,578 INFO [train.py:901] (0/2) Epoch 27, batch 150, loss[loss=0.1249, simple_loss=0.2102, pruned_loss=0.01979, over 7287.00 frames. ], tot_loss[loss=0.1416, simple_loss=0.2222, pruned_loss=0.03052, over 767971.54 frames. ], batch size: 70, lr: 6.17e-03, grad_scale: 8.0 +2023-03-21 04:11:23,268 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=73590.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:11:27,199 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=73598.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:11:39,091 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=73620.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 04:11:41,456 INFO [train.py:901] (0/2) Epoch 27, batch 200, loss[loss=0.1454, simple_loss=0.2291, pruned_loss=0.03079, over 7340.00 frames. ], tot_loss[loss=0.142, simple_loss=0.2219, pruned_loss=0.03099, over 919270.83 frames. ], batch size: 61, lr: 6.17e-03, grad_scale: 8.0 +2023-03-21 04:11:44,512 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.178e+02 1.801e+02 2.024e+02 2.555e+02 5.461e+02, threshold=4.047e+02, percent-clipped=3.0 +2023-03-21 04:11:45,360 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-03-21 04:11:46,088 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 04:11:50,642 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 04:11:54,248 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=73649.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:11:55,306 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=73651.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:11:55,778 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9506, 4.1681, 3.9835, 4.1491, 3.8708, 4.0368, 4.3524, 4.4021], + device='cuda:0'), covar=tensor([0.0240, 0.0160, 0.0186, 0.0170, 0.0292, 0.0263, 0.0271, 0.0213], + device='cuda:0'), in_proj_covar=tensor([0.0112, 0.0118, 0.0106, 0.0115, 0.0106, 0.0093, 0.0094, 0.0091], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:11:57,674 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 04:12:03,825 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=73668.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 04:12:07,321 INFO [train.py:901] (0/2) Epoch 27, batch 250, loss[loss=0.1564, simple_loss=0.2394, pruned_loss=0.03675, over 6610.00 frames. ], tot_loss[loss=0.142, simple_loss=0.2219, pruned_loss=0.03101, over 1036168.20 frames. ], batch size: 106, lr: 6.17e-03, grad_scale: 8.0 +2023-03-21 04:12:10,284 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 04:12:31,138 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 04:12:32,611 INFO [train.py:901] (0/2) Epoch 27, batch 300, loss[loss=0.1378, simple_loss=0.2211, pruned_loss=0.02724, over 7221.00 frames. ], tot_loss[loss=0.1412, simple_loss=0.2209, pruned_loss=0.03075, over 1126421.07 frames. ], batch size: 93, lr: 6.17e-03, grad_scale: 8.0 +2023-03-21 04:12:35,674 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.259e+02 1.939e+02 2.170e+02 2.578e+02 5.056e+02, threshold=4.340e+02, percent-clipped=4.0 +2023-03-21 04:12:36,298 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=73732.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:12:40,673 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 04:12:56,326 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8231, 2.6745, 2.8951, 2.9656, 2.4824, 2.5748, 2.9634, 2.3023], + device='cuda:0'), covar=tensor([0.0454, 0.0519, 0.0545, 0.0593, 0.0551, 0.0745, 0.0509, 0.1474], + device='cuda:0'), in_proj_covar=tensor([0.0338, 0.0342, 0.0273, 0.0363, 0.0305, 0.0299, 0.0345, 0.0274], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:12:58,082 INFO [train.py:901] (0/2) Epoch 27, batch 350, loss[loss=0.1365, simple_loss=0.2087, pruned_loss=0.03215, over 7220.00 frames. ], tot_loss[loss=0.1412, simple_loss=0.2209, pruned_loss=0.03078, over 1198459.15 frames. ], batch size: 45, lr: 6.17e-03, grad_scale: 8.0 +2023-03-21 04:13:07,872 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=73793.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:13:13,498 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 04:13:24,628 INFO [train.py:901] (0/2) Epoch 27, batch 400, loss[loss=0.154, simple_loss=0.2359, pruned_loss=0.03605, over 7125.00 frames. ], tot_loss[loss=0.1413, simple_loss=0.2212, pruned_loss=0.03067, over 1253055.21 frames. ], batch size: 98, lr: 6.16e-03, grad_scale: 8.0 +2023-03-21 04:13:27,544 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.146e+02 1.915e+02 2.214e+02 2.597e+02 4.451e+02, threshold=4.429e+02, percent-clipped=1.0 +2023-03-21 04:13:37,147 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7906, 3.5096, 2.7007, 4.0237, 2.0575, 4.0616, 1.7995, 3.1288], + device='cuda:0'), covar=tensor([0.0141, 0.0700, 0.1814, 0.0190, 0.4322, 0.0196, 0.1286, 0.0349], + device='cuda:0'), in_proj_covar=tensor([0.0199, 0.0254, 0.0276, 0.0199, 0.0262, 0.0208, 0.0248, 0.0233], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:13:50,223 INFO [train.py:901] (0/2) Epoch 27, batch 450, loss[loss=0.136, simple_loss=0.2159, pruned_loss=0.02808, over 7292.00 frames. ], tot_loss[loss=0.1411, simple_loss=0.2208, pruned_loss=0.03069, over 1293730.81 frames. ], batch size: 66, lr: 6.16e-03, grad_scale: 8.0 +2023-03-21 04:13:54,664 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 04:13:55,146 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 04:14:15,705 INFO [train.py:901] (0/2) Epoch 27, batch 500, loss[loss=0.147, simple_loss=0.2358, pruned_loss=0.02913, over 7240.00 frames. ], tot_loss[loss=0.1409, simple_loss=0.2207, pruned_loss=0.03054, over 1328548.70 frames. ], batch size: 55, lr: 6.16e-03, grad_scale: 8.0 +2023-03-21 04:14:18,672 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.251e+02 1.829e+02 2.169e+02 2.599e+02 4.226e+02, threshold=4.339e+02, percent-clipped=0.0 +2023-03-21 04:14:26,334 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=73946.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:14:27,803 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 04:14:27,886 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=73949.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:14:29,289 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 04:14:30,245 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 04:14:32,889 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 04:14:36,979 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 04:14:41,390 INFO [train.py:901] (0/2) Epoch 27, batch 550, loss[loss=0.1403, simple_loss=0.2175, pruned_loss=0.03154, over 7281.00 frames. ], tot_loss[loss=0.1407, simple_loss=0.2207, pruned_loss=0.03035, over 1353523.13 frames. ], batch size: 70, lr: 6.16e-03, grad_scale: 8.0 +2023-03-21 04:14:49,397 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 04:14:53,075 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=73997.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:14:58,402 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 04:15:01,403 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 04:15:05,503 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.0957, 4.5121, 4.4959, 4.9997, 4.9291, 4.9682, 4.3349, 4.6848], + device='cuda:0'), covar=tensor([0.0735, 0.2675, 0.2336, 0.1061, 0.0828, 0.1211, 0.0756, 0.1016], + device='cuda:0'), in_proj_covar=tensor([0.0228, 0.0354, 0.0274, 0.0275, 0.0203, 0.0339, 0.0206, 0.0253], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:15:07,480 INFO [train.py:901] (0/2) Epoch 27, batch 600, loss[loss=0.1429, simple_loss=0.2282, pruned_loss=0.02876, over 7222.00 frames. ], tot_loss[loss=0.1409, simple_loss=0.2208, pruned_loss=0.03047, over 1373974.72 frames. ], batch size: 93, lr: 6.15e-03, grad_scale: 8.0 +2023-03-21 04:15:08,032 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 04:15:10,482 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.222e+02 1.761e+02 2.044e+02 2.479e+02 4.557e+02, threshold=4.088e+02, percent-clipped=1.0 +2023-03-21 04:15:25,824 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 04:15:27,404 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=74064.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:15:32,775 INFO [train.py:901] (0/2) Epoch 27, batch 650, loss[loss=0.1303, simple_loss=0.207, pruned_loss=0.02679, over 7145.00 frames. ], tot_loss[loss=0.1402, simple_loss=0.2195, pruned_loss=0.03045, over 1386852.10 frames. ], batch size: 41, lr: 6.15e-03, grad_scale: 8.0 +2023-03-21 04:15:34,315 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 04:15:39,858 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=74088.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:15:44,837 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1129, 2.5045, 1.8744, 2.5342, 2.9521, 2.5362, 2.4470, 2.3279], + device='cuda:0'), covar=tensor([0.2114, 0.0895, 0.3474, 0.0614, 0.0238, 0.0199, 0.0363, 0.0345], + device='cuda:0'), in_proj_covar=tensor([0.0261, 0.0233, 0.0258, 0.0264, 0.0180, 0.0180, 0.0206, 0.0215], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:15:45,292 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4959, 2.9234, 2.6045, 2.8249, 3.0288, 2.5105, 2.8052, 2.8091], + device='cuda:0'), covar=tensor([0.1171, 0.1020, 0.0792, 0.1071, 0.0601, 0.0757, 0.0640, 0.0918], + device='cuda:0'), in_proj_covar=tensor([0.0052, 0.0051, 0.0059, 0.0053, 0.0050, 0.0055, 0.0051, 0.0048], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:15:51,680 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 04:15:58,160 INFO [train.py:901] (0/2) Epoch 27, batch 700, loss[loss=0.1596, simple_loss=0.2377, pruned_loss=0.04071, over 6689.00 frames. ], tot_loss[loss=0.1405, simple_loss=0.2198, pruned_loss=0.03054, over 1399204.97 frames. ], batch size: 106, lr: 6.15e-03, grad_scale: 8.0 +2023-03-21 04:15:58,329 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=74125.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:16:00,309 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 04:16:00,948 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4328, 1.6858, 1.3475, 1.6268, 1.7173, 1.5339, 1.6234, 1.2921], + device='cuda:0'), covar=tensor([0.0159, 0.0140, 0.0292, 0.0110, 0.0092, 0.0121, 0.0135, 0.0158], + device='cuda:0'), in_proj_covar=tensor([0.0031, 0.0028, 0.0029, 0.0030, 0.0028, 0.0028, 0.0030, 0.0038], + device='cuda:0'), out_proj_covar=tensor([3.6043e-05, 3.1874e-05, 3.2786e-05, 3.4020e-05, 3.1980e-05, 3.1087e-05, + 3.4154e-05, 4.3416e-05], device='cuda:0') +2023-03-21 04:16:01,735 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.376e+02 1.824e+02 2.144e+02 2.665e+02 4.628e+02, threshold=4.287e+02, percent-clipped=3.0 +2023-03-21 04:16:08,822 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4346, 4.0275, 4.2119, 4.1635, 4.1691, 4.0991, 4.4284, 3.8565], + device='cuda:0'), covar=tensor([0.0158, 0.0143, 0.0111, 0.0135, 0.0373, 0.0107, 0.0134, 0.0161], + device='cuda:0'), in_proj_covar=tensor([0.0085, 0.0093, 0.0091, 0.0082, 0.0159, 0.0101, 0.0095, 0.0101], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:16:24,404 INFO [train.py:901] (0/2) Epoch 27, batch 750, loss[loss=0.1428, simple_loss=0.224, pruned_loss=0.03077, over 7265.00 frames. ], tot_loss[loss=0.1408, simple_loss=0.2206, pruned_loss=0.03051, over 1411684.74 frames. ], batch size: 70, lr: 6.15e-03, grad_scale: 8.0 +2023-03-21 04:16:24,919 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 04:16:25,448 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 04:16:40,823 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 04:16:45,314 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 04:16:50,414 INFO [train.py:901] (0/2) Epoch 27, batch 800, loss[loss=0.15, simple_loss=0.233, pruned_loss=0.03347, over 7319.00 frames. ], tot_loss[loss=0.1403, simple_loss=0.2201, pruned_loss=0.03028, over 1420421.59 frames. ], batch size: 59, lr: 6.15e-03, grad_scale: 8.0 +2023-03-21 04:16:51,915 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 04:16:52,949 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 04:16:53,434 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.079e+02 1.936e+02 2.182e+02 2.596e+02 4.514e+02, threshold=4.364e+02, percent-clipped=1.0 +2023-03-21 04:16:58,057 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=74240.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:17:01,055 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=74246.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:17:03,994 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 04:17:16,441 INFO [train.py:901] (0/2) Epoch 27, batch 850, loss[loss=0.1563, simple_loss=0.2424, pruned_loss=0.03507, over 6741.00 frames. ], tot_loss[loss=0.1407, simple_loss=0.2205, pruned_loss=0.03047, over 1422979.41 frames. ], batch size: 107, lr: 6.14e-03, grad_scale: 8.0 +2023-03-21 04:17:23,557 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 04:17:23,566 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 04:17:26,185 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=74294.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:17:28,753 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 04:17:29,918 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=74301.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:17:31,638 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.65 vs. limit=5.0 +2023-03-21 04:17:32,870 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 04:17:38,025 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1270, 4.5499, 4.6201, 4.5153, 4.5283, 4.0778, 4.6452, 4.4790], + device='cuda:0'), covar=tensor([0.0456, 0.0433, 0.0358, 0.0480, 0.0362, 0.0452, 0.0327, 0.0461], + device='cuda:0'), in_proj_covar=tensor([0.0207, 0.0238, 0.0182, 0.0183, 0.0147, 0.0214, 0.0189, 0.0139], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 04:17:42,425 INFO [train.py:901] (0/2) Epoch 27, batch 900, loss[loss=0.1485, simple_loss=0.2246, pruned_loss=0.03623, over 7293.00 frames. ], tot_loss[loss=0.141, simple_loss=0.2205, pruned_loss=0.03076, over 1425571.84 frames. ], batch size: 49, lr: 6.14e-03, grad_scale: 8.0 +2023-03-21 04:17:45,466 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.281e+02 1.810e+02 2.047e+02 2.519e+02 5.499e+02, threshold=4.094e+02, percent-clipped=2.0 +2023-03-21 04:18:00,905 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-03-21 04:18:08,098 INFO [train.py:901] (0/2) Epoch 27, batch 950, loss[loss=0.1258, simple_loss=0.2117, pruned_loss=0.01998, over 7278.00 frames. ], tot_loss[loss=0.1407, simple_loss=0.2202, pruned_loss=0.03059, over 1427752.11 frames. ], batch size: 70, lr: 6.14e-03, grad_scale: 8.0 +2023-03-21 04:18:09,108 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 04:18:09,688 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2985, 4.7399, 4.8637, 4.7426, 4.7077, 4.2266, 4.8622, 4.6498], + device='cuda:0'), covar=tensor([0.0438, 0.0430, 0.0353, 0.0498, 0.0347, 0.0422, 0.0345, 0.0462], + device='cuda:0'), in_proj_covar=tensor([0.0207, 0.0239, 0.0182, 0.0184, 0.0146, 0.0214, 0.0189, 0.0139], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 04:18:14,607 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=74388.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:18:31,605 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=74420.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:18:33,520 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 04:18:33,998 INFO [train.py:901] (0/2) Epoch 27, batch 1000, loss[loss=0.1398, simple_loss=0.2196, pruned_loss=0.03001, over 7266.00 frames. ], tot_loss[loss=0.1402, simple_loss=0.2199, pruned_loss=0.03031, over 1429297.80 frames. ], batch size: 52, lr: 6.14e-03, grad_scale: 8.0 +2023-03-21 04:18:37,002 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.410e+02 1.886e+02 2.186e+02 2.588e+02 3.946e+02, threshold=4.372e+02, percent-clipped=0.0 +2023-03-21 04:18:40,169 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=74436.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:18:55,233 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 04:18:58,955 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.43 vs. limit=5.0 +2023-03-21 04:18:59,726 INFO [train.py:901] (0/2) Epoch 27, batch 1050, loss[loss=0.1441, simple_loss=0.2247, pruned_loss=0.03176, over 7282.00 frames. ], tot_loss[loss=0.1399, simple_loss=0.2196, pruned_loss=0.03009, over 1432783.02 frames. ], batch size: 68, lr: 6.14e-03, grad_scale: 8.0 +2023-03-21 04:19:16,231 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 04:19:20,851 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 04:19:25,982 INFO [train.py:901] (0/2) Epoch 27, batch 1100, loss[loss=0.1273, simple_loss=0.2155, pruned_loss=0.01955, over 7148.00 frames. ], tot_loss[loss=0.1399, simple_loss=0.2195, pruned_loss=0.03013, over 1436604.56 frames. ], batch size: 41, lr: 6.13e-03, grad_scale: 8.0 +2023-03-21 04:19:29,002 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.409e+02 1.926e+02 2.261e+02 2.649e+02 8.248e+02, threshold=4.522e+02, percent-clipped=2.0 +2023-03-21 04:19:41,997 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.91 vs. limit=5.0 +2023-03-21 04:19:50,621 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 04:19:51,084 INFO [train.py:901] (0/2) Epoch 27, batch 1150, loss[loss=0.1242, simple_loss=0.2011, pruned_loss=0.02363, over 7148.00 frames. ], tot_loss[loss=0.1399, simple_loss=0.2194, pruned_loss=0.0302, over 1438669.74 frames. ], batch size: 41, lr: 6.13e-03, grad_scale: 8.0 +2023-03-21 04:19:51,096 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 04:19:55,943 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-03-21 04:20:01,843 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=74595.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:20:02,271 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=74596.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:20:03,754 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 04:20:04,551 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 04:20:13,267 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3236, 3.0237, 2.3946, 3.7098, 2.8888, 3.2262, 1.7077, 2.3874], + device='cuda:0'), covar=tensor([0.0357, 0.0675, 0.2212, 0.0351, 0.0410, 0.0545, 0.3236, 0.1760], + device='cuda:0'), in_proj_covar=tensor([0.0277, 0.0253, 0.0294, 0.0268, 0.0272, 0.0266, 0.0249, 0.0271], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:20:17,609 INFO [train.py:901] (0/2) Epoch 27, batch 1200, loss[loss=0.1066, simple_loss=0.1732, pruned_loss=0.02001, over 6238.00 frames. ], tot_loss[loss=0.1395, simple_loss=0.219, pruned_loss=0.02997, over 1437568.29 frames. ], batch size: 26, lr: 6.13e-03, grad_scale: 8.0 +2023-03-21 04:20:20,653 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.288e+02 1.763e+02 2.138e+02 2.575e+02 4.166e+02, threshold=4.276e+02, percent-clipped=0.0 +2023-03-21 04:20:33,531 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=74656.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:20:37,443 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 04:20:38,190 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.29 vs. limit=2.0 +2023-03-21 04:20:43,648 INFO [train.py:901] (0/2) Epoch 27, batch 1250, loss[loss=0.1446, simple_loss=0.2247, pruned_loss=0.0322, over 7356.00 frames. ], tot_loss[loss=0.1397, simple_loss=0.2193, pruned_loss=0.03001, over 1435862.84 frames. ], batch size: 73, lr: 6.13e-03, grad_scale: 16.0 +2023-03-21 04:20:48,972 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.71 vs. limit=2.0 +2023-03-21 04:20:58,927 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.1960, 1.4705, 1.1999, 1.3142, 1.3969, 1.3417, 1.3589, 1.1857], + device='cuda:0'), covar=tensor([0.0117, 0.0104, 0.0137, 0.0115, 0.0082, 0.0118, 0.0088, 0.0112], + device='cuda:0'), in_proj_covar=tensor([0.0031, 0.0028, 0.0029, 0.0030, 0.0028, 0.0028, 0.0030, 0.0038], + device='cuda:0'), out_proj_covar=tensor([3.5801e-05, 3.2149e-05, 3.2759e-05, 3.3804e-05, 3.2175e-05, 3.0967e-05, + 3.3721e-05, 4.3141e-05], device='cuda:0') +2023-03-21 04:21:02,149 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 04:21:06,139 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 04:21:06,733 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=74720.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:21:07,178 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 04:21:09,129 INFO [train.py:901] (0/2) Epoch 27, batch 1300, loss[loss=0.1404, simple_loss=0.2205, pruned_loss=0.03009, over 7276.00 frames. ], tot_loss[loss=0.1402, simple_loss=0.2201, pruned_loss=0.03021, over 1438274.41 frames. ], batch size: 52, lr: 6.13e-03, grad_scale: 16.0 +2023-03-21 04:21:09,251 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8309, 3.9436, 3.7688, 3.9316, 3.6719, 4.0117, 4.1349, 4.2426], + device='cuda:0'), covar=tensor([0.0216, 0.0173, 0.0210, 0.0231, 0.0395, 0.0327, 0.0294, 0.0197], + device='cuda:0'), in_proj_covar=tensor([0.0115, 0.0119, 0.0109, 0.0117, 0.0108, 0.0094, 0.0096, 0.0093], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:21:12,184 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.400e+02 1.865e+02 2.147e+02 2.548e+02 5.558e+02, threshold=4.294e+02, percent-clipped=5.0 +2023-03-21 04:21:29,425 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 04:21:31,478 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=74768.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:21:31,931 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 04:21:33,619 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0604, 2.4728, 2.4812, 2.2697, 2.5657, 2.2423, 2.1731, 2.0072], + device='cuda:0'), covar=tensor([0.0359, 0.0409, 0.0285, 0.0255, 0.0305, 0.0693, 0.0213, 0.0263], + device='cuda:0'), in_proj_covar=tensor([0.0033, 0.0031, 0.0032, 0.0030, 0.0029, 0.0030, 0.0034, 0.0033], + device='cuda:0'), out_proj_covar=tensor([8.2390e-05, 8.1526e-05, 7.9364e-05, 7.6780e-05, 7.8179e-05, 7.7877e-05, + 8.3713e-05, 8.5018e-05], device='cuda:0') +2023-03-21 04:21:35,000 INFO [train.py:901] (0/2) Epoch 27, batch 1350, loss[loss=0.1435, simple_loss=0.2256, pruned_loss=0.03068, over 7290.00 frames. ], tot_loss[loss=0.1399, simple_loss=0.2201, pruned_loss=0.02986, over 1439073.89 frames. ], batch size: 77, lr: 6.12e-03, grad_scale: 16.0 +2023-03-21 04:21:35,013 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 04:21:45,677 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 04:22:00,960 INFO [train.py:901] (0/2) Epoch 27, batch 1400, loss[loss=0.1344, simple_loss=0.2177, pruned_loss=0.02552, over 7318.00 frames. ], tot_loss[loss=0.1398, simple_loss=0.22, pruned_loss=0.02978, over 1438929.11 frames. ], batch size: 59, lr: 6.12e-03, grad_scale: 16.0 +2023-03-21 04:22:03,884 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.131e+02 1.860e+02 2.142e+02 2.591e+02 4.616e+02, threshold=4.284e+02, percent-clipped=2.0 +2023-03-21 04:22:17,623 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 04:22:27,221 INFO [train.py:901] (0/2) Epoch 27, batch 1450, loss[loss=0.1488, simple_loss=0.2384, pruned_loss=0.02964, over 7294.00 frames. ], tot_loss[loss=0.1404, simple_loss=0.2206, pruned_loss=0.0301, over 1441073.16 frames. ], batch size: 86, lr: 6.12e-03, grad_scale: 16.0 +2023-03-21 04:22:37,829 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=74896.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:22:41,753 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 04:22:44,387 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=74909.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:22:52,256 INFO [train.py:901] (0/2) Epoch 27, batch 1500, loss[loss=0.1479, simple_loss=0.2327, pruned_loss=0.0316, over 7320.00 frames. ], tot_loss[loss=0.1402, simple_loss=0.2203, pruned_loss=0.03005, over 1441967.70 frames. ], batch size: 75, lr: 6.12e-03, grad_scale: 16.0 +2023-03-21 04:22:54,921 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2383, 1.4053, 1.2433, 1.2646, 1.4073, 1.3210, 1.2627, 1.0954], + device='cuda:0'), covar=tensor([0.0135, 0.0169, 0.0290, 0.0183, 0.0143, 0.0120, 0.0192, 0.0165], + device='cuda:0'), in_proj_covar=tensor([0.0031, 0.0028, 0.0028, 0.0030, 0.0028, 0.0027, 0.0030, 0.0038], + device='cuda:0'), out_proj_covar=tensor([3.5205e-05, 3.1910e-05, 3.2563e-05, 3.3650e-05, 3.1616e-05, 3.0466e-05, + 3.3843e-05, 4.2554e-05], device='cuda:0') +2023-03-21 04:22:55,756 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.463e+02 1.784e+02 2.140e+02 2.549e+02 4.343e+02, threshold=4.279e+02, percent-clipped=1.0 +2023-03-21 04:22:57,309 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 04:23:02,453 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=74944.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:23:06,646 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=74951.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:23:12,736 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7733, 3.1741, 3.8185, 3.8006, 3.7715, 3.9254, 3.8670, 3.7139], + device='cuda:0'), covar=tensor([0.0032, 0.0104, 0.0029, 0.0030, 0.0033, 0.0024, 0.0039, 0.0044], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0062, 0.0053, 0.0051, 0.0051, 0.0054, 0.0047, 0.0069], + device='cuda:0'), out_proj_covar=tensor([8.2071e-05, 1.3735e-04, 1.0582e-04, 9.6268e-05, 9.5366e-05, 1.0143e-04, + 9.7422e-05, 1.3768e-04], device='cuda:0') +2023-03-21 04:23:13,249 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=74964.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:23:16,223 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=74970.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:23:18,515 INFO [train.py:901] (0/2) Epoch 27, batch 1550, loss[loss=0.1578, simple_loss=0.2379, pruned_loss=0.03882, over 7276.00 frames. ], tot_loss[loss=0.1398, simple_loss=0.2202, pruned_loss=0.02972, over 1441613.57 frames. ], batch size: 57, lr: 6.12e-03, grad_scale: 8.0 +2023-03-21 04:23:21,958 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 04:23:44,532 INFO [train.py:901] (0/2) Epoch 27, batch 1600, loss[loss=0.1759, simple_loss=0.2571, pruned_loss=0.04737, over 6705.00 frames. ], tot_loss[loss=0.1399, simple_loss=0.2203, pruned_loss=0.02979, over 1440282.69 frames. ], batch size: 106, lr: 6.11e-03, grad_scale: 8.0 +2023-03-21 04:23:44,689 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=75025.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:23:47,210 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5497, 3.1328, 2.3647, 3.9318, 2.9258, 3.3434, 1.7625, 2.4738], + device='cuda:0'), covar=tensor([0.0466, 0.0662, 0.2313, 0.0379, 0.0442, 0.0619, 0.3044, 0.1673], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0251, 0.0289, 0.0266, 0.0267, 0.0266, 0.0246, 0.0269], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:23:48,021 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.244e+02 1.920e+02 2.273e+02 2.923e+02 5.005e+02, threshold=4.545e+02, percent-clipped=3.0 +2023-03-21 04:23:53,564 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 04:23:54,088 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 04:23:55,192 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9911, 3.3196, 3.9473, 3.8832, 3.9017, 4.0739, 4.1043, 3.9201], + device='cuda:0'), covar=tensor([0.0034, 0.0106, 0.0030, 0.0035, 0.0033, 0.0025, 0.0031, 0.0048], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0063, 0.0053, 0.0051, 0.0051, 0.0055, 0.0047, 0.0070], + device='cuda:0'), out_proj_covar=tensor([8.2495e-05, 1.3945e-04, 1.0646e-04, 9.7358e-05, 9.6346e-05, 1.0243e-04, + 9.7879e-05, 1.3871e-04], device='cuda:0') +2023-03-21 04:23:57,095 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 04:24:06,196 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 04:24:07,882 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9613, 2.2819, 1.7541, 2.6444, 2.5767, 2.7710, 2.3397, 2.4069], + device='cuda:0'), covar=tensor([0.1808, 0.0893, 0.3470, 0.0835, 0.0261, 0.0269, 0.0396, 0.0399], + device='cuda:0'), in_proj_covar=tensor([0.0268, 0.0240, 0.0265, 0.0271, 0.0187, 0.0185, 0.0213, 0.0225], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 04:24:10,231 INFO [train.py:901] (0/2) Epoch 27, batch 1650, loss[loss=0.107, simple_loss=0.1693, pruned_loss=0.02235, over 6078.00 frames. ], tot_loss[loss=0.1407, simple_loss=0.2206, pruned_loss=0.0304, over 1437457.47 frames. ], batch size: 26, lr: 6.11e-03, grad_scale: 8.0 +2023-03-21 04:24:11,270 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 04:24:19,342 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 04:24:32,949 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-03-21 04:24:36,049 INFO [train.py:901] (0/2) Epoch 27, batch 1700, loss[loss=0.1349, simple_loss=0.2116, pruned_loss=0.02912, over 7270.00 frames. ], tot_loss[loss=0.1412, simple_loss=0.221, pruned_loss=0.03074, over 1438458.08 frames. ], batch size: 89, lr: 6.11e-03, grad_scale: 8.0 +2023-03-21 04:24:37,557 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 04:24:40,082 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.205e+02 1.858e+02 2.103e+02 2.498e+02 6.232e+02, threshold=4.206e+02, percent-clipped=1.0 +2023-03-21 04:24:42,227 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 04:24:52,206 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 04:24:56,970 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.87 vs. limit=2.0 +2023-03-21 04:25:01,678 INFO [train.py:901] (0/2) Epoch 27, batch 1750, loss[loss=0.1397, simple_loss=0.2222, pruned_loss=0.02863, over 7302.00 frames. ], tot_loss[loss=0.1415, simple_loss=0.2214, pruned_loss=0.0308, over 1438722.43 frames. ], batch size: 49, lr: 6.11e-03, grad_scale: 8.0 +2023-03-21 04:25:16,219 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 04:25:17,242 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 04:25:26,012 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.88 vs. limit=2.0 +2023-03-21 04:25:28,195 INFO [train.py:901] (0/2) Epoch 27, batch 1800, loss[loss=0.1455, simple_loss=0.2266, pruned_loss=0.03223, over 7284.00 frames. ], tot_loss[loss=0.1408, simple_loss=0.2209, pruned_loss=0.03039, over 1440877.03 frames. ], batch size: 52, lr: 6.11e-03, grad_scale: 8.0 +2023-03-21 04:25:29,845 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=75228.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:25:31,712 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.312e+02 1.830e+02 2.153e+02 2.628e+02 4.402e+02, threshold=4.306e+02, percent-clipped=1.0 +2023-03-21 04:25:39,336 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 04:25:41,413 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=75251.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:25:48,528 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=75265.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:25:51,048 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=75270.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:25:53,640 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 04:25:54,103 INFO [train.py:901] (0/2) Epoch 27, batch 1850, loss[loss=0.131, simple_loss=0.2169, pruned_loss=0.02256, over 7341.00 frames. ], tot_loss[loss=0.1408, simple_loss=0.2209, pruned_loss=0.03029, over 1442783.17 frames. ], batch size: 44, lr: 6.10e-03, grad_scale: 8.0 +2023-03-21 04:26:01,391 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=75289.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:26:03,282 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 04:26:06,304 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=75299.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:26:17,403 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=75320.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:26:19,788 INFO [train.py:901] (0/2) Epoch 27, batch 1900, loss[loss=0.1112, simple_loss=0.1852, pruned_loss=0.01858, over 7180.00 frames. ], tot_loss[loss=0.1403, simple_loss=0.2203, pruned_loss=0.03018, over 1442529.24 frames. ], batch size: 39, lr: 6.10e-03, grad_scale: 8.0 +2023-03-21 04:26:20,330 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 04:26:23,030 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=75331.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 04:26:23,334 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.215e+02 1.826e+02 2.085e+02 2.530e+02 5.155e+02, threshold=4.170e+02, percent-clipped=3.0 +2023-03-21 04:26:33,890 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4108, 3.1698, 3.1684, 3.3547, 2.8955, 2.8484, 3.5014, 2.5384], + device='cuda:0'), covar=tensor([0.0445, 0.0513, 0.0570, 0.0578, 0.0767, 0.0975, 0.0578, 0.1760], + device='cuda:0'), in_proj_covar=tensor([0.0337, 0.0346, 0.0273, 0.0363, 0.0300, 0.0301, 0.0345, 0.0276], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:26:43,968 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 04:26:44,599 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2167, 1.3956, 1.1477, 1.3168, 1.4048, 1.2842, 1.2534, 1.1289], + device='cuda:0'), covar=tensor([0.0156, 0.0162, 0.0227, 0.0157, 0.0166, 0.0132, 0.0140, 0.0147], + device='cuda:0'), in_proj_covar=tensor([0.0032, 0.0029, 0.0029, 0.0030, 0.0028, 0.0028, 0.0031, 0.0038], + device='cuda:0'), out_proj_covar=tensor([3.6226e-05, 3.2300e-05, 3.2685e-05, 3.4064e-05, 3.2240e-05, 3.1064e-05, + 3.4569e-05, 4.3208e-05], device='cuda:0') +2023-03-21 04:26:45,473 INFO [train.py:901] (0/2) Epoch 27, batch 1950, loss[loss=0.1345, simple_loss=0.2202, pruned_loss=0.02436, over 7264.00 frames. ], tot_loss[loss=0.1405, simple_loss=0.2207, pruned_loss=0.03016, over 1442482.48 frames. ], batch size: 64, lr: 6.10e-03, grad_scale: 8.0 +2023-03-21 04:26:47,142 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2221, 2.9226, 3.2237, 3.1099, 2.7856, 2.7970, 3.2812, 2.4537], + device='cuda:0'), covar=tensor([0.0416, 0.0474, 0.0526, 0.0533, 0.0525, 0.0812, 0.0641, 0.1653], + device='cuda:0'), in_proj_covar=tensor([0.0336, 0.0343, 0.0271, 0.0362, 0.0299, 0.0300, 0.0343, 0.0275], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:26:55,194 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 04:26:55,792 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0462, 3.9495, 3.3351, 3.7126, 3.1862, 2.4816, 1.9411, 4.2813], + device='cuda:0'), covar=tensor([0.0060, 0.0122, 0.0126, 0.0070, 0.0121, 0.0460, 0.0564, 0.0046], + device='cuda:0'), in_proj_covar=tensor([0.0103, 0.0083, 0.0105, 0.0088, 0.0117, 0.0126, 0.0124, 0.0097], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 04:27:00,015 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 04:27:01,031 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 04:27:11,520 INFO [train.py:901] (0/2) Epoch 27, batch 2000, loss[loss=0.1282, simple_loss=0.2118, pruned_loss=0.02227, over 7239.00 frames. ], tot_loss[loss=0.1408, simple_loss=0.2212, pruned_loss=0.03019, over 1445576.25 frames. ], batch size: 55, lr: 6.10e-03, grad_scale: 8.0 +2023-03-21 04:27:14,943 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.375e+02 1.802e+02 2.059e+02 2.426e+02 4.961e+02, threshold=4.118e+02, percent-clipped=2.0 +2023-03-21 04:27:16,466 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 04:27:20,886 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.36 vs. limit=5.0 +2023-03-21 04:27:21,292 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2066, 3.0594, 3.2044, 3.1777, 2.7960, 2.7158, 3.3127, 2.4711], + device='cuda:0'), covar=tensor([0.0409, 0.0448, 0.0447, 0.0505, 0.0497, 0.0732, 0.0520, 0.1527], + device='cuda:0'), in_proj_covar=tensor([0.0334, 0.0342, 0.0270, 0.0360, 0.0298, 0.0297, 0.0343, 0.0273], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:27:28,166 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 04:27:35,241 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 04:27:37,898 INFO [train.py:901] (0/2) Epoch 27, batch 2050, loss[loss=0.1435, simple_loss=0.2261, pruned_loss=0.03046, over 7284.00 frames. ], tot_loss[loss=0.1407, simple_loss=0.2209, pruned_loss=0.03026, over 1443470.19 frames. ], batch size: 86, lr: 6.10e-03, grad_scale: 8.0 +2023-03-21 04:28:03,073 INFO [train.py:901] (0/2) Epoch 27, batch 2100, loss[loss=0.1391, simple_loss=0.2215, pruned_loss=0.02832, over 7255.00 frames. ], tot_loss[loss=0.1417, simple_loss=0.2219, pruned_loss=0.03079, over 1442560.72 frames. ], batch size: 89, lr: 6.09e-03, grad_scale: 8.0 +2023-03-21 04:28:07,140 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.255e+02 1.775e+02 2.143e+02 2.491e+02 4.702e+02, threshold=4.286e+02, percent-clipped=1.0 +2023-03-21 04:28:08,668 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 04:28:11,611 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 04:28:24,225 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=75565.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:28:29,121 INFO [train.py:901] (0/2) Epoch 27, batch 2150, loss[loss=0.1407, simple_loss=0.2243, pruned_loss=0.02858, over 7278.00 frames. ], tot_loss[loss=0.1407, simple_loss=0.2212, pruned_loss=0.03008, over 1445170.91 frames. ], batch size: 70, lr: 6.09e-03, grad_scale: 8.0 +2023-03-21 04:28:33,727 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=75584.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:28:49,215 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=75613.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:28:52,757 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=75620.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:28:55,171 INFO [train.py:901] (0/2) Epoch 27, batch 2200, loss[loss=0.158, simple_loss=0.239, pruned_loss=0.03851, over 7305.00 frames. ], tot_loss[loss=0.1411, simple_loss=0.2212, pruned_loss=0.03049, over 1443739.78 frames. ], batch size: 59, lr: 6.09e-03, grad_scale: 8.0 +2023-03-21 04:28:55,543 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.92 vs. limit=5.0 +2023-03-21 04:28:55,744 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=75626.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 04:28:58,679 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.144e+02 1.743e+02 2.140e+02 2.621e+02 4.761e+02, threshold=4.279e+02, percent-clipped=4.0 +2023-03-21 04:28:58,704 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 04:29:17,308 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=75668.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:29:20,772 INFO [train.py:901] (0/2) Epoch 27, batch 2250, loss[loss=0.114, simple_loss=0.1911, pruned_loss=0.01844, over 7190.00 frames. ], tot_loss[loss=0.141, simple_loss=0.2211, pruned_loss=0.03047, over 1444729.91 frames. ], batch size: 39, lr: 6.09e-03, grad_scale: 8.0 +2023-03-21 04:29:31,899 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 04:29:31,911 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 04:29:44,955 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 04:29:46,399 INFO [train.py:901] (0/2) Epoch 27, batch 2300, loss[loss=0.1338, simple_loss=0.2239, pruned_loss=0.02184, over 7364.00 frames. ], tot_loss[loss=0.1405, simple_loss=0.2206, pruned_loss=0.03023, over 1443788.46 frames. ], batch size: 73, lr: 6.09e-03, grad_scale: 8.0 +2023-03-21 04:29:50,602 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.180e+02 1.982e+02 2.222e+02 2.648e+02 6.200e+02, threshold=4.445e+02, percent-clipped=3.0 +2023-03-21 04:30:12,335 INFO [train.py:901] (0/2) Epoch 27, batch 2350, loss[loss=0.1172, simple_loss=0.1952, pruned_loss=0.01962, over 7349.00 frames. ], tot_loss[loss=0.1401, simple_loss=0.2205, pruned_loss=0.02985, over 1443784.71 frames. ], batch size: 44, lr: 6.08e-03, grad_scale: 8.0 +2023-03-21 04:30:32,257 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 04:30:32,913 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9695, 2.6588, 1.7990, 2.9613, 2.7239, 2.9127, 2.2412, 2.5037], + device='cuda:0'), covar=tensor([0.2229, 0.0829, 0.3793, 0.0667, 0.0223, 0.0181, 0.0345, 0.0309], + device='cuda:0'), in_proj_covar=tensor([0.0260, 0.0232, 0.0259, 0.0265, 0.0184, 0.0182, 0.0208, 0.0219], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:30:38,955 INFO [train.py:901] (0/2) Epoch 27, batch 2400, loss[loss=0.126, simple_loss=0.212, pruned_loss=0.01996, over 7340.00 frames. ], tot_loss[loss=0.1396, simple_loss=0.2199, pruned_loss=0.02967, over 1443939.92 frames. ], batch size: 44, lr: 6.08e-03, grad_scale: 8.0 +2023-03-21 04:30:39,971 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 04:30:42,375 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.255e+02 1.828e+02 2.153e+02 2.703e+02 5.254e+02, threshold=4.305e+02, percent-clipped=1.0 +2023-03-21 04:30:51,134 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 04:30:53,636 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 04:31:04,306 INFO [train.py:901] (0/2) Epoch 27, batch 2450, loss[loss=0.1489, simple_loss=0.2307, pruned_loss=0.03352, over 7259.00 frames. ], tot_loss[loss=0.14, simple_loss=0.2201, pruned_loss=0.02993, over 1444080.32 frames. ], batch size: 89, lr: 6.08e-03, grad_scale: 8.0 +2023-03-21 04:31:09,798 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=75884.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:31:21,429 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 04:31:30,682 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7171, 2.3349, 1.6523, 2.6144, 2.5772, 2.5218, 1.9630, 2.3319], + device='cuda:0'), covar=tensor([0.1819, 0.0829, 0.3613, 0.0646, 0.0191, 0.0220, 0.0368, 0.0353], + device='cuda:0'), in_proj_covar=tensor([0.0255, 0.0228, 0.0254, 0.0259, 0.0181, 0.0179, 0.0204, 0.0215], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:31:30,996 INFO [train.py:901] (0/2) Epoch 27, batch 2500, loss[loss=0.1509, simple_loss=0.2385, pruned_loss=0.03169, over 7211.00 frames. ], tot_loss[loss=0.1397, simple_loss=0.2196, pruned_loss=0.02988, over 1443593.99 frames. ], batch size: 93, lr: 6.08e-03, grad_scale: 8.0 +2023-03-21 04:31:31,574 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=75926.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 04:31:34,497 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.362e+02 1.829e+02 2.210e+02 2.665e+02 4.876e+02, threshold=4.419e+02, percent-clipped=2.0 +2023-03-21 04:31:34,578 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=75932.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:31:41,264 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4288, 2.3989, 2.1927, 3.7044, 1.8607, 3.3301, 1.4085, 2.9683], + device='cuda:0'), covar=tensor([0.0136, 0.1237, 0.1658, 0.0216, 0.3394, 0.0245, 0.1152, 0.0355], + device='cuda:0'), in_proj_covar=tensor([0.0198, 0.0254, 0.0277, 0.0198, 0.0259, 0.0209, 0.0243, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:31:46,188 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 04:31:47,263 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5993, 3.7087, 3.4771, 3.7646, 3.5845, 3.6194, 3.8255, 3.9317], + device='cuda:0'), covar=tensor([0.0309, 0.0246, 0.0312, 0.0281, 0.0436, 0.0347, 0.0403, 0.0303], + device='cuda:0'), in_proj_covar=tensor([0.0114, 0.0118, 0.0107, 0.0115, 0.0105, 0.0094, 0.0095, 0.0091], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:31:56,259 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=75974.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:31:56,725 INFO [train.py:901] (0/2) Epoch 27, batch 2550, loss[loss=0.1486, simple_loss=0.2226, pruned_loss=0.03723, over 7264.00 frames. ], tot_loss[loss=0.14, simple_loss=0.22, pruned_loss=0.03002, over 1442731.64 frames. ], batch size: 55, lr: 6.08e-03, grad_scale: 8.0 +2023-03-21 04:32:06,060 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.29 vs. limit=2.0 +2023-03-21 04:32:08,646 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-03-21 04:32:10,167 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-76000.pt +2023-03-21 04:32:25,135 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7757, 3.8338, 2.8507, 3.3609, 2.8857, 2.0054, 1.6866, 3.8253], + device='cuda:0'), covar=tensor([0.0058, 0.0051, 0.0176, 0.0071, 0.0151, 0.0556, 0.0677, 0.0058], + device='cuda:0'), in_proj_covar=tensor([0.0104, 0.0084, 0.0106, 0.0089, 0.0119, 0.0127, 0.0126, 0.0099], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 04:32:26,511 INFO [train.py:901] (0/2) Epoch 27, batch 2600, loss[loss=0.1513, simple_loss=0.2314, pruned_loss=0.03566, over 7306.00 frames. ], tot_loss[loss=0.1405, simple_loss=0.2205, pruned_loss=0.03026, over 1443369.04 frames. ], batch size: 83, lr: 6.07e-03, grad_scale: 8.0 +2023-03-21 04:32:29,975 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.271e+02 1.838e+02 2.095e+02 2.590e+02 5.003e+02, threshold=4.189e+02, percent-clipped=2.0 +2023-03-21 04:32:50,538 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5220, 2.7897, 2.4536, 2.7505, 2.8387, 2.4960, 2.7996, 2.6252], + device='cuda:0'), covar=tensor([0.0742, 0.0658, 0.0938, 0.1040, 0.1020, 0.0747, 0.0378, 0.0979], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0053, 0.0061, 0.0054, 0.0051, 0.0055, 0.0051, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:32:51,390 INFO [train.py:901] (0/2) Epoch 27, batch 2650, loss[loss=0.1087, simple_loss=0.1886, pruned_loss=0.01444, over 7128.00 frames. ], tot_loss[loss=0.1396, simple_loss=0.2197, pruned_loss=0.02976, over 1444342.29 frames. ], batch size: 39, lr: 6.07e-03, grad_scale: 8.0 +2023-03-21 04:32:58,083 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.33 vs. limit=2.0 +2023-03-21 04:33:06,803 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9094, 3.3586, 3.8041, 3.8915, 3.8499, 3.8198, 3.7685, 3.7572], + device='cuda:0'), covar=tensor([0.0028, 0.0102, 0.0032, 0.0030, 0.0035, 0.0029, 0.0050, 0.0047], + device='cuda:0'), in_proj_covar=tensor([0.0045, 0.0064, 0.0054, 0.0053, 0.0052, 0.0056, 0.0048, 0.0071], + device='cuda:0'), out_proj_covar=tensor([8.2636e-05, 1.4109e-04, 1.0919e-04, 9.9796e-05, 9.6584e-05, 1.0590e-04, + 1.0081e-04, 1.3974e-04], device='cuda:0') +2023-03-21 04:33:16,554 INFO [train.py:901] (0/2) Epoch 27, batch 2700, loss[loss=0.1341, simple_loss=0.2161, pruned_loss=0.02604, over 7369.00 frames. ], tot_loss[loss=0.1395, simple_loss=0.2196, pruned_loss=0.02968, over 1442573.30 frames. ], batch size: 65, lr: 6.07e-03, grad_scale: 8.0 +2023-03-21 04:33:19,922 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.171e+02 1.823e+02 2.064e+02 2.639e+02 4.176e+02, threshold=4.127e+02, percent-clipped=0.0 +2023-03-21 04:33:41,163 INFO [train.py:901] (0/2) Epoch 27, batch 2750, loss[loss=0.146, simple_loss=0.2235, pruned_loss=0.03428, over 7265.00 frames. ], tot_loss[loss=0.1395, simple_loss=0.2192, pruned_loss=0.02989, over 1439896.73 frames. ], batch size: 55, lr: 6.07e-03, grad_scale: 8.0 +2023-03-21 04:34:06,262 INFO [train.py:901] (0/2) Epoch 27, batch 2800, loss[loss=0.1293, simple_loss=0.2138, pruned_loss=0.02246, over 7316.00 frames. ], tot_loss[loss=0.1398, simple_loss=0.2198, pruned_loss=0.02995, over 1440139.25 frames. ], batch size: 80, lr: 6.07e-03, grad_scale: 8.0 +2023-03-21 04:34:07,888 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9922, 2.8384, 3.1823, 3.0576, 2.8440, 2.7467, 3.0963, 2.4040], + device='cuda:0'), covar=tensor([0.0379, 0.0481, 0.0490, 0.0505, 0.0551, 0.0788, 0.0547, 0.1540], + device='cuda:0'), in_proj_covar=tensor([0.0334, 0.0346, 0.0275, 0.0363, 0.0301, 0.0300, 0.0349, 0.0273], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:34:08,029 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.63 vs. limit=5.0 +2023-03-21 04:34:09,662 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.382e+02 1.810e+02 2.119e+02 2.567e+02 4.960e+02, threshold=4.237e+02, percent-clipped=2.0 +2023-03-21 04:34:18,828 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-27.pt +2023-03-21 04:34:36,762 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 04:34:36,814 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 04:34:36,840 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 04:34:36,856 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 04:34:37,035 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 04:34:37,052 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 04:34:37,062 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 04:34:37,163 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 04:34:37,585 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 04:34:37,601 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 04:34:37,617 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 04:34:37,781 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 04:34:37,786 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 04:34:37,794 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 04:34:37,934 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 04:34:37,937 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 04:34:37,996 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 04:34:40,234 INFO [train.py:901] (0/2) Epoch 28, batch 0, loss[loss=0.1465, simple_loss=0.2262, pruned_loss=0.03338, over 7268.00 frames. ], tot_loss[loss=0.1465, simple_loss=0.2262, pruned_loss=0.03338, over 7268.00 frames. ], batch size: 52, lr: 5.96e-03, grad_scale: 8.0 +2023-03-21 04:34:40,235 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 04:34:47,482 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3383, 1.5591, 1.3922, 1.4206, 1.5054, 1.4952, 1.4308, 1.2295], + device='cuda:0'), covar=tensor([0.0145, 0.0131, 0.0121, 0.0110, 0.0114, 0.0075, 0.0090, 0.0120], + device='cuda:0'), in_proj_covar=tensor([0.0032, 0.0029, 0.0029, 0.0030, 0.0028, 0.0028, 0.0030, 0.0038], + device='cuda:0'), out_proj_covar=tensor([3.6160e-05, 3.3013e-05, 3.2865e-05, 3.3256e-05, 3.2287e-05, 3.1621e-05, + 3.4396e-05, 4.3169e-05], device='cuda:0') +2023-03-21 04:35:06,112 INFO [train.py:935] (0/2) Epoch 28, validation: loss=0.1647, simple_loss=0.254, pruned_loss=0.03776, over 1622729.00 frames. +2023-03-21 04:35:06,112 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 04:35:10,224 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8066, 3.8835, 3.6713, 3.6137, 3.0153, 3.5043, 3.7000, 3.5076], + device='cuda:0'), covar=tensor([0.0237, 0.0157, 0.0147, 0.0239, 0.0766, 0.0176, 0.0267, 0.0199], + device='cuda:0'), in_proj_covar=tensor([0.0088, 0.0094, 0.0094, 0.0083, 0.0163, 0.0103, 0.0097, 0.0104], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:35:13,270 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 04:35:16,445 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=76269.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 04:35:23,467 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 04:35:29,083 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=76294.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:35:30,422 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 04:35:31,422 INFO [train.py:901] (0/2) Epoch 28, batch 50, loss[loss=0.1416, simple_loss=0.2272, pruned_loss=0.02798, over 7290.00 frames. ], tot_loss[loss=0.1408, simple_loss=0.2214, pruned_loss=0.03012, over 327185.77 frames. ], batch size: 68, lr: 5.96e-03, grad_scale: 8.0 +2023-03-21 04:35:32,454 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 04:35:34,897 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 04:35:44,308 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=76323.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 04:35:48,422 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=76330.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 04:35:49,243 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.198e+02 1.914e+02 2.137e+02 2.399e+02 6.526e+02, threshold=4.275e+02, percent-clipped=3.0 +2023-03-21 04:35:52,772 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 04:35:53,307 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 04:35:57,770 INFO [train.py:901] (0/2) Epoch 28, batch 100, loss[loss=0.155, simple_loss=0.2377, pruned_loss=0.03619, over 7333.00 frames. ], tot_loss[loss=0.1395, simple_loss=0.2205, pruned_loss=0.02925, over 575546.18 frames. ], batch size: 54, lr: 5.95e-03, grad_scale: 8.0 +2023-03-21 04:36:00,969 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=76355.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:36:07,492 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3679, 1.5343, 1.2334, 1.3564, 1.6144, 1.4071, 1.4123, 1.1309], + device='cuda:0'), covar=tensor([0.0130, 0.0152, 0.0291, 0.0148, 0.0095, 0.0124, 0.0119, 0.0138], + device='cuda:0'), in_proj_covar=tensor([0.0031, 0.0029, 0.0029, 0.0029, 0.0028, 0.0028, 0.0030, 0.0038], + device='cuda:0'), out_proj_covar=tensor([3.5889e-05, 3.2764e-05, 3.2465e-05, 3.3051e-05, 3.2246e-05, 3.1315e-05, + 3.4217e-05, 4.2905e-05], device='cuda:0') +2023-03-21 04:36:15,477 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=76384.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 04:36:23,383 INFO [train.py:901] (0/2) Epoch 28, batch 150, loss[loss=0.1392, simple_loss=0.2232, pruned_loss=0.02764, over 7360.00 frames. ], tot_loss[loss=0.1393, simple_loss=0.2207, pruned_loss=0.02896, over 767821.99 frames. ], batch size: 54, lr: 5.95e-03, grad_scale: 8.0 +2023-03-21 04:36:25,129 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.90 vs. limit=2.0 +2023-03-21 04:36:32,834 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.74 vs. limit=2.0 +2023-03-21 04:36:34,640 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9598, 3.8785, 3.4457, 3.5146, 3.0729, 2.1288, 1.9024, 4.0235], + device='cuda:0'), covar=tensor([0.0053, 0.0068, 0.0110, 0.0070, 0.0138, 0.0497, 0.0613, 0.0053], + device='cuda:0'), in_proj_covar=tensor([0.0100, 0.0082, 0.0102, 0.0087, 0.0115, 0.0122, 0.0121, 0.0095], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 04:36:40,924 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.135e+02 1.678e+02 1.996e+02 2.465e+02 6.288e+02, threshold=3.991e+02, percent-clipped=1.0 +2023-03-21 04:36:49,558 INFO [train.py:901] (0/2) Epoch 28, batch 200, loss[loss=0.1419, simple_loss=0.2245, pruned_loss=0.02967, over 7277.00 frames. ], tot_loss[loss=0.1394, simple_loss=0.2208, pruned_loss=0.02905, over 919787.16 frames. ], batch size: 57, lr: 5.95e-03, grad_scale: 8.0 +2023-03-21 04:36:53,604 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 04:36:57,530 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 04:37:03,693 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 04:37:08,485 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6809, 2.6441, 2.2792, 3.7365, 1.9330, 3.5337, 1.4651, 3.1764], + device='cuda:0'), covar=tensor([0.0152, 0.1163, 0.1728, 0.0185, 0.3372, 0.0237, 0.1150, 0.0382], + device='cuda:0'), in_proj_covar=tensor([0.0196, 0.0251, 0.0270, 0.0194, 0.0255, 0.0208, 0.0240, 0.0227], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:37:12,880 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3626, 4.9037, 4.9635, 4.8704, 4.8200, 4.3918, 4.9913, 4.8104], + device='cuda:0'), covar=tensor([0.0467, 0.0406, 0.0358, 0.0460, 0.0298, 0.0402, 0.0302, 0.0410], + device='cuda:0'), in_proj_covar=tensor([0.0206, 0.0236, 0.0182, 0.0183, 0.0144, 0.0215, 0.0190, 0.0140], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 04:37:15,269 INFO [train.py:901] (0/2) Epoch 28, batch 250, loss[loss=0.129, simple_loss=0.2081, pruned_loss=0.02492, over 7307.00 frames. ], tot_loss[loss=0.1394, simple_loss=0.2202, pruned_loss=0.02928, over 1036389.69 frames. ], batch size: 80, lr: 5.95e-03, grad_scale: 8.0 +2023-03-21 04:37:17,940 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 04:37:32,544 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.451e+02 1.788e+02 2.073e+02 2.472e+02 4.824e+02, threshold=4.147e+02, percent-clipped=3.0 +2023-03-21 04:37:37,648 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 04:37:41,142 INFO [train.py:901] (0/2) Epoch 28, batch 300, loss[loss=0.1543, simple_loss=0.2336, pruned_loss=0.03746, over 7332.00 frames. ], tot_loss[loss=0.1392, simple_loss=0.2196, pruned_loss=0.0294, over 1124812.35 frames. ], batch size: 61, lr: 5.95e-03, grad_scale: 8.0 +2023-03-21 04:37:46,257 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 04:37:53,096 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4184, 2.4571, 2.2719, 3.7177, 1.7459, 3.4107, 1.3667, 3.0945], + device='cuda:0'), covar=tensor([0.0152, 0.1245, 0.1848, 0.0170, 0.3959, 0.0224, 0.1300, 0.0349], + device='cuda:0'), in_proj_covar=tensor([0.0197, 0.0255, 0.0272, 0.0195, 0.0259, 0.0209, 0.0242, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:38:07,550 INFO [train.py:901] (0/2) Epoch 28, batch 350, loss[loss=0.123, simple_loss=0.1878, pruned_loss=0.0291, over 6250.00 frames. ], tot_loss[loss=0.1394, simple_loss=0.2196, pruned_loss=0.02955, over 1196314.64 frames. ], batch size: 27, lr: 5.94e-03, grad_scale: 8.0 +2023-03-21 04:38:13,127 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8523, 2.4498, 3.1631, 2.9258, 2.9691, 2.8979, 2.5128, 3.0163], + device='cuda:0'), covar=tensor([0.1659, 0.0911, 0.1127, 0.1256, 0.1241, 0.0981, 0.2447, 0.1490], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0062, 0.0047, 0.0047, 0.0045, 0.0044, 0.0064, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:38:21,270 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=76625.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 04:38:23,198 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 04:38:24,673 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.125e+02 1.636e+02 1.963e+02 2.462e+02 4.673e+02, threshold=3.925e+02, percent-clipped=2.0 +2023-03-21 04:38:33,378 INFO [train.py:901] (0/2) Epoch 28, batch 400, loss[loss=0.1341, simple_loss=0.2175, pruned_loss=0.02533, over 7291.00 frames. ], tot_loss[loss=0.1392, simple_loss=0.2194, pruned_loss=0.02948, over 1249872.20 frames. ], batch size: 70, lr: 5.94e-03, grad_scale: 8.0 +2023-03-21 04:38:33,928 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=76650.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:38:49,072 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=76679.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 04:38:59,631 INFO [train.py:901] (0/2) Epoch 28, batch 450, loss[loss=0.1109, simple_loss=0.1782, pruned_loss=0.02184, over 6518.00 frames. ], tot_loss[loss=0.1395, simple_loss=0.2198, pruned_loss=0.02966, over 1293908.12 frames. ], batch size: 28, lr: 5.94e-03, grad_scale: 8.0 +2023-03-21 04:39:03,614 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 04:39:04,108 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 04:39:16,075 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.221e+02 1.885e+02 2.115e+02 2.485e+02 5.641e+02, threshold=4.229e+02, percent-clipped=2.0 +2023-03-21 04:39:25,348 INFO [train.py:901] (0/2) Epoch 28, batch 500, loss[loss=0.1273, simple_loss=0.2059, pruned_loss=0.02434, over 7287.00 frames. ], tot_loss[loss=0.1394, simple_loss=0.2197, pruned_loss=0.02956, over 1327595.74 frames. ], batch size: 47, lr: 5.94e-03, grad_scale: 8.0 +2023-03-21 04:39:36,614 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 04:39:38,119 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 04:39:39,076 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 04:39:41,577 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 04:39:46,055 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 04:39:51,016 INFO [train.py:901] (0/2) Epoch 28, batch 550, loss[loss=0.1567, simple_loss=0.2385, pruned_loss=0.03751, over 7334.00 frames. ], tot_loss[loss=0.1397, simple_loss=0.2202, pruned_loss=0.02961, over 1354687.27 frames. ], batch size: 61, lr: 5.94e-03, grad_scale: 8.0 +2023-03-21 04:39:58,440 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 04:40:07,188 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 04:40:08,679 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.405e+02 1.821e+02 2.250e+02 2.684e+02 4.961e+02, threshold=4.501e+02, percent-clipped=3.0 +2023-03-21 04:40:09,715 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 04:40:11,386 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2802, 1.5409, 1.2256, 1.5282, 1.6970, 1.4316, 1.4078, 1.1444], + device='cuda:0'), covar=tensor([0.0145, 0.0159, 0.0197, 0.0142, 0.0100, 0.0135, 0.0109, 0.0156], + device='cuda:0'), in_proj_covar=tensor([0.0032, 0.0029, 0.0029, 0.0030, 0.0029, 0.0028, 0.0030, 0.0039], + device='cuda:0'), out_proj_covar=tensor([3.6519e-05, 3.3293e-05, 3.2523e-05, 3.3423e-05, 3.2508e-05, 3.1395e-05, + 3.4303e-05, 4.3745e-05], device='cuda:0') +2023-03-21 04:40:12,367 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7796, 3.8335, 2.8897, 3.4288, 2.9456, 2.0490, 1.7658, 3.8230], + device='cuda:0'), covar=tensor([0.0090, 0.0072, 0.0253, 0.0081, 0.0221, 0.0679, 0.0705, 0.0086], + device='cuda:0'), in_proj_covar=tensor([0.0103, 0.0084, 0.0105, 0.0089, 0.0119, 0.0126, 0.0125, 0.0097], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 04:40:12,387 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=76839.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:40:16,741 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 04:40:17,291 INFO [train.py:901] (0/2) Epoch 28, batch 600, loss[loss=0.106, simple_loss=0.169, pruned_loss=0.02144, over 5896.00 frames. ], tot_loss[loss=0.1389, simple_loss=0.2192, pruned_loss=0.02931, over 1369963.29 frames. ], batch size: 25, lr: 5.93e-03, grad_scale: 8.0 +2023-03-21 04:40:22,447 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8975, 3.1551, 3.6366, 3.7551, 3.7616, 3.7922, 3.7941, 3.6602], + device='cuda:0'), covar=tensor([0.0027, 0.0112, 0.0042, 0.0033, 0.0036, 0.0034, 0.0044, 0.0051], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0062, 0.0053, 0.0051, 0.0050, 0.0055, 0.0047, 0.0069], + device='cuda:0'), out_proj_covar=tensor([8.0475e-05, 1.3600e-04, 1.0693e-04, 9.6122e-05, 9.1528e-05, 1.0290e-04, + 9.7995e-05, 1.3576e-04], device='cuda:0') +2023-03-21 04:40:33,027 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 04:40:37,607 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9256, 2.8959, 3.2968, 3.1686, 3.2192, 3.1733, 2.7605, 3.2670], + device='cuda:0'), covar=tensor([0.2282, 0.0740, 0.1803, 0.1241, 0.1040, 0.0957, 0.2375, 0.1481], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0063, 0.0048, 0.0047, 0.0046, 0.0044, 0.0064, 0.0048], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:40:42,038 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 04:40:42,989 INFO [train.py:901] (0/2) Epoch 28, batch 650, loss[loss=0.1592, simple_loss=0.2386, pruned_loss=0.03991, over 7115.00 frames. ], tot_loss[loss=0.1391, simple_loss=0.2195, pruned_loss=0.02936, over 1386097.34 frames. ], batch size: 98, lr: 5.93e-03, grad_scale: 8.0 +2023-03-21 04:40:43,686 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=76900.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:40:56,907 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=76925.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 04:41:00,281 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.372e+02 1.758e+02 2.082e+02 2.489e+02 4.937e+02, threshold=4.165e+02, percent-clipped=1.0 +2023-03-21 04:41:00,303 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 04:41:09,282 INFO [train.py:901] (0/2) Epoch 28, batch 700, loss[loss=0.1679, simple_loss=0.2461, pruned_loss=0.0449, over 7237.00 frames. ], tot_loss[loss=0.1393, simple_loss=0.2194, pruned_loss=0.02965, over 1398816.99 frames. ], batch size: 89, lr: 5.93e-03, grad_scale: 8.0 +2023-03-21 04:41:09,285 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 04:41:09,898 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=76950.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:41:21,234 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=76973.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 04:41:23,347 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2346, 2.9102, 2.1044, 3.0788, 3.3986, 3.0835, 2.6346, 2.6378], + device='cuda:0'), covar=tensor([0.1739, 0.0728, 0.3133, 0.0471, 0.0209, 0.0208, 0.0231, 0.0342], + device='cuda:0'), in_proj_covar=tensor([0.0259, 0.0232, 0.0260, 0.0265, 0.0186, 0.0182, 0.0207, 0.0223], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 04:41:24,293 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=76979.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 04:41:26,772 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2463, 3.8776, 4.0596, 4.0614, 4.1048, 4.1124, 4.2755, 4.0752], + device='cuda:0'), covar=tensor([0.0027, 0.0075, 0.0033, 0.0031, 0.0032, 0.0032, 0.0028, 0.0046], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0062, 0.0053, 0.0051, 0.0050, 0.0055, 0.0047, 0.0069], + device='cuda:0'), out_proj_covar=tensor([8.1356e-05, 1.3656e-04, 1.0690e-04, 9.6406e-05, 9.1367e-05, 1.0391e-04, + 9.8123e-05, 1.3614e-04], device='cuda:0') +2023-03-21 04:41:32,746 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 04:41:33,243 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 04:41:33,768 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=76998.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:41:34,208 INFO [train.py:901] (0/2) Epoch 28, batch 750, loss[loss=0.1301, simple_loss=0.2189, pruned_loss=0.02068, over 7280.00 frames. ], tot_loss[loss=0.1393, simple_loss=0.2194, pruned_loss=0.02956, over 1408342.89 frames. ], batch size: 77, lr: 5.93e-03, grad_scale: 16.0 +2023-03-21 04:41:39,415 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.99 vs. limit=5.0 +2023-03-21 04:41:47,139 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 04:41:49,147 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=77027.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 04:41:51,578 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.364e+02 1.849e+02 2.159e+02 2.501e+02 3.591e+02, threshold=4.317e+02, percent-clipped=0.0 +2023-03-21 04:41:52,117 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 04:41:52,930 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-21 04:41:58,771 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 04:42:00,308 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 04:42:00,790 INFO [train.py:901] (0/2) Epoch 28, batch 800, loss[loss=0.1492, simple_loss=0.2263, pruned_loss=0.03607, over 7265.00 frames. ], tot_loss[loss=0.1395, simple_loss=0.2196, pruned_loss=0.0297, over 1417135.23 frames. ], batch size: 64, lr: 5.93e-03, grad_scale: 16.0 +2023-03-21 04:42:11,388 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 04:42:16,628 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.74 vs. limit=2.0 +2023-03-21 04:42:24,651 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=77095.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:42:26,485 INFO [train.py:901] (0/2) Epoch 28, batch 850, loss[loss=0.147, simple_loss=0.2288, pruned_loss=0.03262, over 7330.00 frames. ], tot_loss[loss=0.1393, simple_loss=0.2191, pruned_loss=0.0297, over 1420515.22 frames. ], batch size: 44, lr: 5.92e-03, grad_scale: 4.0 +2023-03-21 04:42:30,928 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 04:42:30,937 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 04:42:36,571 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 04:42:40,672 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 04:42:41,808 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7042, 3.9321, 3.6968, 3.8665, 3.4778, 3.8906, 4.0787, 4.1654], + device='cuda:0'), covar=tensor([0.0230, 0.0172, 0.0218, 0.0229, 0.0365, 0.0270, 0.0286, 0.0216], + device='cuda:0'), in_proj_covar=tensor([0.0115, 0.0120, 0.0108, 0.0118, 0.0106, 0.0096, 0.0096, 0.0093], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:42:44,711 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.361e+02 1.829e+02 2.089e+02 2.425e+02 3.800e+02, threshold=4.177e+02, percent-clipped=0.0 +2023-03-21 04:42:49,619 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.75 vs. limit=2.0 +2023-03-21 04:42:52,353 INFO [train.py:901] (0/2) Epoch 28, batch 900, loss[loss=0.1513, simple_loss=0.2307, pruned_loss=0.03592, over 7342.00 frames. ], tot_loss[loss=0.1397, simple_loss=0.2197, pruned_loss=0.02978, over 1425518.79 frames. ], batch size: 54, lr: 5.92e-03, grad_scale: 4.0 +2023-03-21 04:42:56,079 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=77156.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 04:43:02,022 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2458, 3.9360, 3.9644, 3.9896, 3.8322, 3.8875, 4.0951, 3.7124], + device='cuda:0'), covar=tensor([0.0137, 0.0138, 0.0122, 0.0142, 0.0414, 0.0109, 0.0157, 0.0149], + device='cuda:0'), in_proj_covar=tensor([0.0088, 0.0094, 0.0093, 0.0082, 0.0161, 0.0102, 0.0096, 0.0102], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:43:09,238 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8513, 2.9106, 2.3908, 4.0047, 1.8051, 3.7974, 1.6080, 3.2456], + device='cuda:0'), covar=tensor([0.0129, 0.0908, 0.1645, 0.0123, 0.3770, 0.0153, 0.1091, 0.0299], + device='cuda:0'), in_proj_covar=tensor([0.0196, 0.0256, 0.0268, 0.0196, 0.0258, 0.0207, 0.0242, 0.0227], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:43:15,978 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=77195.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:43:17,930 INFO [train.py:901] (0/2) Epoch 28, batch 950, loss[loss=0.1325, simple_loss=0.2114, pruned_loss=0.0268, over 7264.00 frames. ], tot_loss[loss=0.1401, simple_loss=0.2201, pruned_loss=0.03006, over 1430400.91 frames. ], batch size: 52, lr: 5.92e-03, grad_scale: 4.0 +2023-03-21 04:43:18,771 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 04:43:19,214 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.68 vs. limit=2.0 +2023-03-21 04:43:25,096 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=77211.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:43:36,543 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.269e+02 1.862e+02 2.179e+02 2.499e+02 5.050e+02, threshold=4.359e+02, percent-clipped=3.0 +2023-03-21 04:43:40,599 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.62 vs. limit=2.0 +2023-03-21 04:43:42,208 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 04:43:44,219 INFO [train.py:901] (0/2) Epoch 28, batch 1000, loss[loss=0.1395, simple_loss=0.2214, pruned_loss=0.02877, over 7322.00 frames. ], tot_loss[loss=0.1401, simple_loss=0.22, pruned_loss=0.03015, over 1433449.95 frames. ], batch size: 59, lr: 5.92e-03, grad_scale: 4.0 +2023-03-21 04:43:56,686 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=77272.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:43:57,158 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=77273.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:43:59,139 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=77277.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:44:02,521 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 04:44:10,573 INFO [train.py:901] (0/2) Epoch 28, batch 1050, loss[loss=0.1179, simple_loss=0.1868, pruned_loss=0.02447, over 6991.00 frames. ], tot_loss[loss=0.1401, simple_loss=0.2201, pruned_loss=0.03004, over 1436257.28 frames. ], batch size: 35, lr: 5.92e-03, grad_scale: 4.0 +2023-03-21 04:44:24,252 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 04:44:26,859 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9950, 2.6697, 2.4998, 2.1108, 2.4760, 2.2954, 1.9297, 1.8692], + device='cuda:0'), covar=tensor([0.0464, 0.0311, 0.0280, 0.0243, 0.0458, 0.0496, 0.0338, 0.0302], + device='cuda:0'), in_proj_covar=tensor([0.0033, 0.0031, 0.0032, 0.0030, 0.0031, 0.0030, 0.0035, 0.0034], + device='cuda:0'), out_proj_covar=tensor([8.4110e-05, 8.1392e-05, 8.0921e-05, 7.6931e-05, 8.0232e-05, 7.8344e-05, + 8.6097e-05, 8.6238e-05], device='cuda:0') +2023-03-21 04:44:28,248 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.136e+02 1.790e+02 2.098e+02 2.468e+02 4.594e+02, threshold=4.197e+02, percent-clipped=2.0 +2023-03-21 04:44:28,302 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 04:44:28,428 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=77334.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:44:30,408 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=77338.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:44:35,847 INFO [train.py:901] (0/2) Epoch 28, batch 1100, loss[loss=0.1406, simple_loss=0.2236, pruned_loss=0.02881, over 7361.00 frames. ], tot_loss[loss=0.1397, simple_loss=0.2195, pruned_loss=0.02992, over 1436049.67 frames. ], batch size: 63, lr: 5.91e-03, grad_scale: 4.0 +2023-03-21 04:44:57,060 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 04:44:57,524 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 04:45:01,968 INFO [train.py:901] (0/2) Epoch 28, batch 1150, loss[loss=0.1324, simple_loss=0.2073, pruned_loss=0.02874, over 7206.00 frames. ], tot_loss[loss=0.1395, simple_loss=0.2193, pruned_loss=0.02985, over 1434611.61 frames. ], batch size: 39, lr: 5.91e-03, grad_scale: 4.0 +2023-03-21 04:45:10,361 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 04:45:10,868 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 04:45:19,866 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.393e+02 1.883e+02 2.182e+02 2.626e+02 3.951e+02, threshold=4.364e+02, percent-clipped=0.0 +2023-03-21 04:45:27,935 INFO [train.py:901] (0/2) Epoch 28, batch 1200, loss[loss=0.1353, simple_loss=0.2168, pruned_loss=0.02686, over 7306.00 frames. ], tot_loss[loss=0.1396, simple_loss=0.2191, pruned_loss=0.03007, over 1433867.53 frames. ], batch size: 83, lr: 5.91e-03, grad_scale: 8.0 +2023-03-21 04:45:28,995 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=77451.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 04:45:38,131 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5054, 2.2920, 2.0591, 3.5752, 1.6207, 3.3583, 1.3320, 3.0961], + device='cuda:0'), covar=tensor([0.0177, 0.1338, 0.1816, 0.0180, 0.4425, 0.0187, 0.1318, 0.0366], + device='cuda:0'), in_proj_covar=tensor([0.0197, 0.0258, 0.0271, 0.0195, 0.0259, 0.0207, 0.0244, 0.0231], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:45:43,457 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 04:45:48,249 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.58 vs. limit=5.0 +2023-03-21 04:45:48,275 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.03 vs. limit=5.0 +2023-03-21 04:45:51,516 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=77495.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:45:53,441 INFO [train.py:901] (0/2) Epoch 28, batch 1250, loss[loss=0.1452, simple_loss=0.2173, pruned_loss=0.03661, over 7229.00 frames. ], tot_loss[loss=0.1399, simple_loss=0.2196, pruned_loss=0.0301, over 1435320.73 frames. ], batch size: 45, lr: 5.91e-03, grad_scale: 8.0 +2023-03-21 04:45:59,108 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9332, 2.3514, 2.3955, 1.9238, 2.2868, 2.2365, 1.7785, 1.7029], + device='cuda:0'), covar=tensor([0.0539, 0.0310, 0.0200, 0.0195, 0.0327, 0.0323, 0.0390, 0.0333], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0032, 0.0033, 0.0031, 0.0032, 0.0031, 0.0036, 0.0035], + device='cuda:0'), out_proj_covar=tensor([8.8006e-05, 8.3975e-05, 8.3492e-05, 7.9415e-05, 8.2973e-05, 8.1103e-05, + 8.8826e-05, 8.8540e-05], device='cuda:0') +2023-03-21 04:46:03,201 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5483, 2.7502, 3.3194, 3.4406, 3.5016, 3.5585, 3.2770, 3.4080], + device='cuda:0'), covar=tensor([0.0024, 0.0130, 0.0039, 0.0033, 0.0031, 0.0028, 0.0075, 0.0047], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0063, 0.0053, 0.0052, 0.0050, 0.0055, 0.0047, 0.0070], + device='cuda:0'), out_proj_covar=tensor([8.1600e-05, 1.3817e-04, 1.0590e-04, 9.6996e-05, 9.2100e-05, 1.0373e-04, + 9.8249e-05, 1.3607e-04], device='cuda:0') +2023-03-21 04:46:03,400 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.81 vs. limit=2.0 +2023-03-21 04:46:05,543 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.65 vs. limit=2.0 +2023-03-21 04:46:06,107 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 04:46:10,752 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 04:46:11,742 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.150e+02 1.879e+02 2.180e+02 2.544e+02 4.448e+02, threshold=4.361e+02, percent-clipped=1.0 +2023-03-21 04:46:11,798 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 04:46:16,326 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=77543.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:46:19,408 INFO [train.py:901] (0/2) Epoch 28, batch 1300, loss[loss=0.1271, simple_loss=0.2123, pruned_loss=0.02093, over 7329.00 frames. ], tot_loss[loss=0.1397, simple_loss=0.2195, pruned_loss=0.02991, over 1436305.45 frames. ], batch size: 75, lr: 5.91e-03, grad_scale: 8.0 +2023-03-21 04:46:29,042 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=77567.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:46:36,017 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 04:46:38,001 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 04:46:39,142 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1879, 2.6635, 1.8552, 3.1471, 3.0631, 2.8826, 2.7994, 2.6439], + device='cuda:0'), covar=tensor([0.2040, 0.0908, 0.3516, 0.0543, 0.0216, 0.0259, 0.0377, 0.0357], + device='cuda:0'), in_proj_covar=tensor([0.0259, 0.0231, 0.0257, 0.0263, 0.0185, 0.0183, 0.0207, 0.0221], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 04:46:41,504 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 04:46:45,003 INFO [train.py:901] (0/2) Epoch 28, batch 1350, loss[loss=0.1322, simple_loss=0.2153, pruned_loss=0.02458, over 7283.00 frames. ], tot_loss[loss=0.1401, simple_loss=0.2199, pruned_loss=0.03016, over 1440078.02 frames. ], batch size: 77, lr: 5.91e-03, grad_scale: 8.0 +2023-03-21 04:46:52,414 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 04:46:54,544 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=77616.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:47:01,082 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=77629.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:47:03,156 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=77633.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:47:03,582 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.276e+02 1.842e+02 2.121e+02 2.615e+02 6.191e+02, threshold=4.241e+02, percent-clipped=3.0 +2023-03-21 04:47:11,226 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.04 vs. limit=2.0 +2023-03-21 04:47:11,939 INFO [train.py:901] (0/2) Epoch 28, batch 1400, loss[loss=0.1317, simple_loss=0.2133, pruned_loss=0.02503, over 7248.00 frames. ], tot_loss[loss=0.1395, simple_loss=0.2193, pruned_loss=0.02991, over 1436190.43 frames. ], batch size: 45, lr: 5.90e-03, grad_scale: 8.0 +2023-03-21 04:47:26,044 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 04:47:26,144 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=77677.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:47:26,158 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=77677.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:47:37,272 INFO [train.py:901] (0/2) Epoch 28, batch 1450, loss[loss=0.1615, simple_loss=0.2343, pruned_loss=0.04435, over 7335.00 frames. ], tot_loss[loss=0.1396, simple_loss=0.2193, pruned_loss=0.02994, over 1436822.51 frames. ], batch size: 54, lr: 5.90e-03, grad_scale: 8.0 +2023-03-21 04:47:50,497 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 04:47:55,968 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.062e+02 1.839e+02 2.145e+02 2.638e+02 3.761e+02, threshold=4.290e+02, percent-clipped=0.0 +2023-03-21 04:47:58,151 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=77738.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:48:03,440 INFO [train.py:901] (0/2) Epoch 28, batch 1500, loss[loss=0.1503, simple_loss=0.2272, pruned_loss=0.03671, over 7277.00 frames. ], tot_loss[loss=0.1391, simple_loss=0.2188, pruned_loss=0.02972, over 1438811.46 frames. ], batch size: 68, lr: 5.90e-03, grad_scale: 8.0 +2023-03-21 04:48:04,569 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=77751.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 04:48:07,563 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 04:48:10,689 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8143, 2.4421, 3.2028, 3.0226, 3.0389, 2.8587, 2.5256, 2.9463], + device='cuda:0'), covar=tensor([0.1526, 0.0746, 0.0813, 0.0855, 0.0663, 0.1214, 0.2013, 0.1669], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0063, 0.0048, 0.0047, 0.0046, 0.0045, 0.0065, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:48:17,300 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4328, 1.5585, 1.3350, 1.5704, 1.5821, 1.4948, 1.4691, 1.1462], + device='cuda:0'), covar=tensor([0.0091, 0.0177, 0.0292, 0.0089, 0.0106, 0.0134, 0.0123, 0.0150], + device='cuda:0'), in_proj_covar=tensor([0.0032, 0.0030, 0.0029, 0.0030, 0.0029, 0.0029, 0.0030, 0.0039], + device='cuda:0'), out_proj_covar=tensor([3.6693e-05, 3.3977e-05, 3.3283e-05, 3.3189e-05, 3.2452e-05, 3.2163e-05, + 3.4142e-05, 4.4071e-05], device='cuda:0') +2023-03-21 04:48:18,925 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.06 vs. limit=5.0 +2023-03-21 04:48:29,480 INFO [train.py:901] (0/2) Epoch 28, batch 1550, loss[loss=0.1487, simple_loss=0.2269, pruned_loss=0.03524, over 7356.00 frames. ], tot_loss[loss=0.1386, simple_loss=0.2185, pruned_loss=0.02937, over 1438966.41 frames. ], batch size: 73, lr: 5.90e-03, grad_scale: 8.0 +2023-03-21 04:48:29,545 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=77799.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:48:32,344 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 04:48:48,024 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.264e+02 1.872e+02 2.235e+02 2.567e+02 5.065e+02, threshold=4.469e+02, percent-clipped=1.0 +2023-03-21 04:48:55,578 INFO [train.py:901] (0/2) Epoch 28, batch 1600, loss[loss=0.1347, simple_loss=0.222, pruned_loss=0.02365, over 7255.00 frames. ], tot_loss[loss=0.1396, simple_loss=0.2195, pruned_loss=0.02984, over 1441235.27 frames. ], batch size: 47, lr: 5.90e-03, grad_scale: 8.0 +2023-03-21 04:48:59,991 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.64 vs. limit=2.0 +2023-03-21 04:49:03,067 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 04:49:04,050 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 04:49:04,630 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0913, 4.5359, 4.6018, 4.5641, 4.5744, 4.1243, 4.6085, 4.5043], + device='cuda:0'), covar=tensor([0.0515, 0.0434, 0.0428, 0.0449, 0.0304, 0.0446, 0.0377, 0.0457], + device='cuda:0'), in_proj_covar=tensor([0.0209, 0.0240, 0.0184, 0.0185, 0.0146, 0.0218, 0.0193, 0.0143], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 04:49:04,669 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=77867.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:49:06,572 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 04:49:16,774 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 04:49:21,315 INFO [train.py:901] (0/2) Epoch 28, batch 1650, loss[loss=0.1137, simple_loss=0.1925, pruned_loss=0.01748, over 7179.00 frames. ], tot_loss[loss=0.1402, simple_loss=0.2198, pruned_loss=0.03026, over 1440872.24 frames. ], batch size: 41, lr: 5.89e-03, grad_scale: 8.0 +2023-03-21 04:49:21,350 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 04:49:21,491 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4357, 2.1963, 2.0533, 3.5702, 1.6731, 3.2831, 1.2577, 2.9310], + device='cuda:0'), covar=tensor([0.0129, 0.1208, 0.1639, 0.0127, 0.3593, 0.0174, 0.1225, 0.0284], + device='cuda:0'), in_proj_covar=tensor([0.0196, 0.0257, 0.0271, 0.0195, 0.0257, 0.0208, 0.0244, 0.0230], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:49:22,132 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.92 vs. limit=2.0 +2023-03-21 04:49:29,482 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 04:49:29,974 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=77915.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:49:37,124 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=77929.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:49:39,075 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=77933.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:49:39,435 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.287e+02 1.899e+02 2.148e+02 2.752e+02 4.294e+02, threshold=4.295e+02, percent-clipped=0.0 +2023-03-21 04:49:46,485 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 04:49:46,980 INFO [train.py:901] (0/2) Epoch 28, batch 1700, loss[loss=0.1377, simple_loss=0.2184, pruned_loss=0.02847, over 7287.00 frames. ], tot_loss[loss=0.1405, simple_loss=0.2203, pruned_loss=0.03038, over 1442649.84 frames. ], batch size: 57, lr: 5.89e-03, grad_scale: 8.0 +2023-03-21 04:49:50,052 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 04:49:59,397 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=77972.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:50:01,355 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 04:50:01,905 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=77977.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:50:03,947 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=77981.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:50:13,656 INFO [train.py:901] (0/2) Epoch 28, batch 1750, loss[loss=0.1401, simple_loss=0.2206, pruned_loss=0.02978, over 7332.00 frames. ], tot_loss[loss=0.14, simple_loss=0.2198, pruned_loss=0.0301, over 1441115.03 frames. ], batch size: 54, lr: 5.89e-03, grad_scale: 8.0 +2023-03-21 04:50:26,549 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 04:50:28,081 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 04:50:31,245 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=78033.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:50:31,655 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.311e+02 1.834e+02 2.126e+02 2.360e+02 4.852e+02, threshold=4.252e+02, percent-clipped=1.0 +2023-03-21 04:50:35,516 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-21 04:50:39,129 INFO [train.py:901] (0/2) Epoch 28, batch 1800, loss[loss=0.137, simple_loss=0.2197, pruned_loss=0.02711, over 7261.00 frames. ], tot_loss[loss=0.1397, simple_loss=0.2196, pruned_loss=0.02988, over 1442777.04 frames. ], batch size: 47, lr: 5.89e-03, grad_scale: 8.0 +2023-03-21 04:50:39,532 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.56 vs. limit=2.0 +2023-03-21 04:50:44,888 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8748, 4.0680, 3.7794, 4.0318, 3.6626, 3.9848, 4.2265, 4.3369], + device='cuda:0'), covar=tensor([0.0214, 0.0171, 0.0257, 0.0184, 0.0348, 0.0235, 0.0280, 0.0180], + device='cuda:0'), in_proj_covar=tensor([0.0117, 0.0120, 0.0109, 0.0117, 0.0107, 0.0097, 0.0096, 0.0093], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:50:49,912 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.2618, 4.7664, 4.6766, 5.2094, 5.1430, 5.1335, 4.5825, 4.7308], + device='cuda:0'), covar=tensor([0.0729, 0.2335, 0.2049, 0.0881, 0.0858, 0.1278, 0.0805, 0.1051], + device='cuda:0'), in_proj_covar=tensor([0.0236, 0.0361, 0.0279, 0.0285, 0.0210, 0.0346, 0.0206, 0.0256], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:50:50,363 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 04:51:04,989 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 04:51:05,477 INFO [train.py:901] (0/2) Epoch 28, batch 1850, loss[loss=0.1543, simple_loss=0.2353, pruned_loss=0.03667, over 7357.00 frames. ], tot_loss[loss=0.139, simple_loss=0.2188, pruned_loss=0.02954, over 1442384.65 frames. ], batch size: 51, lr: 5.89e-03, grad_scale: 8.0 +2023-03-21 04:51:09,210 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3988, 1.5240, 1.4421, 1.6147, 1.7152, 1.6423, 1.4652, 1.1217], + device='cuda:0'), covar=tensor([0.0120, 0.0127, 0.0145, 0.0099, 0.0099, 0.0078, 0.0115, 0.0141], + device='cuda:0'), in_proj_covar=tensor([0.0032, 0.0030, 0.0029, 0.0030, 0.0029, 0.0028, 0.0030, 0.0039], + device='cuda:0'), out_proj_covar=tensor([3.6907e-05, 3.3723e-05, 3.3119e-05, 3.3269e-05, 3.2281e-05, 3.1839e-05, + 3.4074e-05, 4.4101e-05], device='cuda:0') +2023-03-21 04:51:14,542 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 04:51:23,156 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.319e+02 1.957e+02 2.174e+02 2.564e+02 5.105e+02, threshold=4.348e+02, percent-clipped=1.0 +2023-03-21 04:51:31,380 INFO [train.py:901] (0/2) Epoch 28, batch 1900, loss[loss=0.1399, simple_loss=0.2186, pruned_loss=0.03057, over 7282.00 frames. ], tot_loss[loss=0.139, simple_loss=0.2188, pruned_loss=0.02957, over 1441573.19 frames. ], batch size: 47, lr: 5.88e-03, grad_scale: 8.0 +2023-03-21 04:51:31,873 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 04:51:34,941 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9019, 4.4240, 4.3002, 4.8867, 4.7583, 4.8168, 4.3234, 4.3988], + device='cuda:0'), covar=tensor([0.0772, 0.2463, 0.2135, 0.0985, 0.0882, 0.1198, 0.0756, 0.1141], + device='cuda:0'), in_proj_covar=tensor([0.0236, 0.0360, 0.0279, 0.0286, 0.0211, 0.0346, 0.0205, 0.0256], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:51:56,679 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 04:51:57,180 INFO [train.py:901] (0/2) Epoch 28, batch 1950, loss[loss=0.1687, simple_loss=0.2508, pruned_loss=0.04334, over 6677.00 frames. ], tot_loss[loss=0.1395, simple_loss=0.2197, pruned_loss=0.02969, over 1443653.36 frames. ], batch size: 107, lr: 5.88e-03, grad_scale: 8.0 +2023-03-21 04:52:06,935 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 04:52:07,062 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9407, 2.7692, 3.4216, 2.9957, 3.2098, 3.0720, 2.7253, 3.0216], + device='cuda:0'), covar=tensor([0.1874, 0.0904, 0.0908, 0.1902, 0.1024, 0.0964, 0.1940, 0.2082], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0064, 0.0049, 0.0048, 0.0047, 0.0046, 0.0066, 0.0050], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 04:52:12,115 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 04:52:12,640 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 04:52:14,668 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=78232.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:52:15,524 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.183e+02 1.838e+02 2.132e+02 2.471e+02 5.256e+02, threshold=4.265e+02, percent-clipped=3.0 +2023-03-21 04:52:23,076 INFO [train.py:901] (0/2) Epoch 28, batch 2000, loss[loss=0.1373, simple_loss=0.218, pruned_loss=0.02825, over 7319.00 frames. ], tot_loss[loss=0.1393, simple_loss=0.2193, pruned_loss=0.02962, over 1443351.21 frames. ], batch size: 83, lr: 5.88e-03, grad_scale: 8.0 +2023-03-21 04:52:30,206 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 04:52:35,233 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=78272.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:52:40,578 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 04:52:45,747 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=78293.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 04:52:48,099 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 04:52:48,584 INFO [train.py:901] (0/2) Epoch 28, batch 2050, loss[loss=0.1495, simple_loss=0.2346, pruned_loss=0.03216, over 7116.00 frames. ], tot_loss[loss=0.1392, simple_loss=0.2192, pruned_loss=0.02966, over 1445326.03 frames. ], batch size: 98, lr: 5.88e-03, grad_scale: 8.0 +2023-03-21 04:52:52,704 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0615, 3.0035, 3.3529, 3.1923, 3.3815, 3.1046, 2.8571, 3.3159], + device='cuda:0'), covar=tensor([0.1530, 0.0597, 0.1448, 0.1138, 0.0887, 0.0987, 0.1905, 0.1400], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0063, 0.0049, 0.0048, 0.0048, 0.0046, 0.0066, 0.0050], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 04:52:58,972 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2378, 4.2561, 3.7154, 3.8012, 3.3500, 2.6233, 2.2416, 4.3483], + device='cuda:0'), covar=tensor([0.0051, 0.0059, 0.0094, 0.0055, 0.0121, 0.0419, 0.0512, 0.0039], + device='cuda:0'), in_proj_covar=tensor([0.0104, 0.0085, 0.0105, 0.0089, 0.0121, 0.0127, 0.0125, 0.0098], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 04:52:59,899 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=78320.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:53:06,522 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=78333.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:53:06,912 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.335e+02 1.680e+02 2.089e+02 2.473e+02 3.796e+02, threshold=4.179e+02, percent-clipped=0.0 +2023-03-21 04:53:15,220 INFO [train.py:901] (0/2) Epoch 28, batch 2100, loss[loss=0.1468, simple_loss=0.2265, pruned_loss=0.03349, over 7286.00 frames. ], tot_loss[loss=0.1397, simple_loss=0.2199, pruned_loss=0.0297, over 1444780.93 frames. ], batch size: 70, lr: 5.88e-03, grad_scale: 8.0 +2023-03-21 04:53:15,820 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7651, 5.2795, 5.3590, 5.3147, 5.0813, 4.7854, 5.3818, 5.1726], + device='cuda:0'), covar=tensor([0.0442, 0.0406, 0.0353, 0.0433, 0.0325, 0.0345, 0.0332, 0.0471], + device='cuda:0'), in_proj_covar=tensor([0.0209, 0.0241, 0.0185, 0.0186, 0.0147, 0.0220, 0.0194, 0.0144], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 04:53:21,712 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 04:53:24,152 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 04:53:31,131 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=78381.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:53:40,735 INFO [train.py:901] (0/2) Epoch 28, batch 2150, loss[loss=0.1596, simple_loss=0.2377, pruned_loss=0.04079, over 7317.00 frames. ], tot_loss[loss=0.1393, simple_loss=0.2193, pruned_loss=0.02961, over 1444645.28 frames. ], batch size: 59, lr: 5.88e-03, grad_scale: 8.0 +2023-03-21 04:53:45,862 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.21 vs. limit=2.0 +2023-03-21 04:53:54,999 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.61 vs. limit=2.0 +2023-03-21 04:53:58,815 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2110, 4.6953, 4.8140, 4.7381, 4.6671, 4.2620, 4.8058, 4.6385], + device='cuda:0'), covar=tensor([0.0495, 0.0441, 0.0354, 0.0495, 0.0316, 0.0465, 0.0369, 0.0419], + device='cuda:0'), in_proj_covar=tensor([0.0207, 0.0240, 0.0183, 0.0185, 0.0146, 0.0218, 0.0193, 0.0142], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 04:53:59,250 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.122e+02 1.886e+02 2.193e+02 2.710e+02 5.513e+02, threshold=4.386e+02, percent-clipped=4.0 +2023-03-21 04:54:06,881 INFO [train.py:901] (0/2) Epoch 28, batch 2200, loss[loss=0.1224, simple_loss=0.2038, pruned_loss=0.02057, over 7326.00 frames. ], tot_loss[loss=0.139, simple_loss=0.2192, pruned_loss=0.02946, over 1443805.59 frames. ], batch size: 44, lr: 5.87e-03, grad_scale: 8.0 +2023-03-21 04:54:10,927 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 04:54:12,491 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=78460.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:54:23,484 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6647, 4.1751, 4.1017, 4.6560, 4.5323, 4.6330, 4.0587, 4.1841], + device='cuda:0'), covar=tensor([0.0921, 0.2782, 0.2348, 0.1025, 0.0919, 0.1252, 0.0841, 0.1259], + device='cuda:0'), in_proj_covar=tensor([0.0236, 0.0358, 0.0275, 0.0283, 0.0207, 0.0344, 0.0204, 0.0253], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:54:32,494 INFO [train.py:901] (0/2) Epoch 28, batch 2250, loss[loss=0.1293, simple_loss=0.2139, pruned_loss=0.02232, over 7238.00 frames. ], tot_loss[loss=0.139, simple_loss=0.2192, pruned_loss=0.02934, over 1443258.62 frames. ], batch size: 55, lr: 5.87e-03, grad_scale: 8.0 +2023-03-21 04:54:41,139 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8850, 4.0706, 3.8780, 4.0383, 3.5907, 4.0025, 4.3090, 4.3736], + device='cuda:0'), covar=tensor([0.0226, 0.0159, 0.0193, 0.0153, 0.0405, 0.0348, 0.0225, 0.0145], + device='cuda:0'), in_proj_covar=tensor([0.0117, 0.0120, 0.0109, 0.0115, 0.0108, 0.0097, 0.0095, 0.0093], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:54:44,209 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=78521.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:54:45,103 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 04:54:45,116 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 04:54:50,363 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.188e+02 1.836e+02 2.134e+02 2.527e+02 4.591e+02, threshold=4.269e+02, percent-clipped=1.0 +2023-03-21 04:54:56,379 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 04:54:57,854 INFO [train.py:901] (0/2) Epoch 28, batch 2300, loss[loss=0.1367, simple_loss=0.2175, pruned_loss=0.02798, over 7311.00 frames. ], tot_loss[loss=0.1398, simple_loss=0.2203, pruned_loss=0.02966, over 1443148.70 frames. ], batch size: 75, lr: 5.87e-03, grad_scale: 8.0 +2023-03-21 04:55:01,379 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=78556.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:55:06,581 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3375, 3.2486, 2.3678, 3.8181, 2.9387, 3.0803, 1.5505, 2.3543], + device='cuda:0'), covar=tensor([0.0377, 0.0701, 0.2217, 0.0587, 0.0513, 0.0520, 0.3444, 0.1700], + device='cuda:0'), in_proj_covar=tensor([0.0276, 0.0255, 0.0290, 0.0273, 0.0272, 0.0269, 0.0248, 0.0270], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:55:17,914 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=78588.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 04:55:23,958 INFO [train.py:901] (0/2) Epoch 28, batch 2350, loss[loss=0.1272, simple_loss=0.2121, pruned_loss=0.02115, over 7325.00 frames. ], tot_loss[loss=0.1398, simple_loss=0.2203, pruned_loss=0.02967, over 1445852.64 frames. ], batch size: 83, lr: 5.87e-03, grad_scale: 8.0 +2023-03-21 04:55:33,515 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=78617.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:55:36,574 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0800, 2.5881, 1.7441, 3.3492, 3.0643, 3.0897, 2.8798, 2.6544], + device='cuda:0'), covar=tensor([0.2098, 0.0895, 0.3824, 0.0801, 0.0260, 0.0202, 0.0386, 0.0402], + device='cuda:0'), in_proj_covar=tensor([0.0255, 0.0229, 0.0254, 0.0260, 0.0185, 0.0182, 0.0204, 0.0218], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:55:41,568 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 04:55:42,034 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.400e+02 1.841e+02 2.209e+02 2.542e+02 3.662e+02, threshold=4.418e+02, percent-clipped=0.0 +2023-03-21 04:55:43,244 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0313, 3.0757, 2.1687, 3.4020, 2.5383, 2.9431, 1.3804, 2.0720], + device='cuda:0'), covar=tensor([0.0395, 0.0949, 0.2386, 0.0667, 0.0590, 0.0718, 0.3313, 0.1823], + device='cuda:0'), in_proj_covar=tensor([0.0276, 0.0255, 0.0289, 0.0273, 0.0274, 0.0269, 0.0248, 0.0271], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:55:47,267 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1772, 3.0004, 2.8858, 3.0728, 2.7137, 2.5996, 3.2717, 2.1620], + device='cuda:0'), covar=tensor([0.0688, 0.0659, 0.0547, 0.0764, 0.0880, 0.1200, 0.0702, 0.2288], + device='cuda:0'), in_proj_covar=tensor([0.0337, 0.0340, 0.0271, 0.0358, 0.0301, 0.0297, 0.0343, 0.0270], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 04:55:47,575 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 04:55:50,115 INFO [train.py:901] (0/2) Epoch 28, batch 2400, loss[loss=0.1582, simple_loss=0.2351, pruned_loss=0.04062, over 7379.00 frames. ], tot_loss[loss=0.14, simple_loss=0.2206, pruned_loss=0.02971, over 1445631.63 frames. ], batch size: 56, lr: 5.87e-03, grad_scale: 8.0 +2023-03-21 04:55:59,265 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 04:56:01,845 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 04:56:15,917 INFO [train.py:901] (0/2) Epoch 28, batch 2450, loss[loss=0.1437, simple_loss=0.2278, pruned_loss=0.02979, over 6775.00 frames. ], tot_loss[loss=0.1396, simple_loss=0.22, pruned_loss=0.02958, over 1443687.15 frames. ], batch size: 106, lr: 5.86e-03, grad_scale: 8.0 +2023-03-21 04:56:21,786 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.00 vs. limit=2.0 +2023-03-21 04:56:28,205 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9520, 1.9679, 2.2738, 1.8262, 2.2299, 2.2457, 1.8838, 1.6426], + device='cuda:0'), covar=tensor([0.0468, 0.0445, 0.0234, 0.0221, 0.0453, 0.0308, 0.0340, 0.0311], + device='cuda:0'), in_proj_covar=tensor([0.0034, 0.0032, 0.0032, 0.0031, 0.0030, 0.0030, 0.0034, 0.0034], + device='cuda:0'), out_proj_covar=tensor([8.4533e-05, 8.3191e-05, 8.0817e-05, 7.8340e-05, 8.0061e-05, 7.8418e-05, + 8.4790e-05, 8.6418e-05], device='cuda:0') +2023-03-21 04:56:29,089 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 04:56:33,610 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.123e+02 1.829e+02 2.117e+02 2.491e+02 4.848e+02, threshold=4.235e+02, percent-clipped=1.0 +2023-03-21 04:56:35,434 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.71 vs. limit=5.0 +2023-03-21 04:56:41,773 INFO [train.py:901] (0/2) Epoch 28, batch 2500, loss[loss=0.1253, simple_loss=0.2126, pruned_loss=0.01905, over 7320.00 frames. ], tot_loss[loss=0.1397, simple_loss=0.2201, pruned_loss=0.02969, over 1444440.46 frames. ], batch size: 49, lr: 5.86e-03, grad_scale: 8.0 +2023-03-21 04:56:41,901 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9678, 4.0236, 3.1650, 3.4537, 2.9247, 2.2396, 1.8678, 4.0151], + device='cuda:0'), covar=tensor([0.0037, 0.0042, 0.0118, 0.0068, 0.0147, 0.0481, 0.0569, 0.0041], + device='cuda:0'), in_proj_covar=tensor([0.0101, 0.0084, 0.0103, 0.0087, 0.0117, 0.0124, 0.0123, 0.0096], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 04:56:51,268 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-03-21 04:56:51,959 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9051, 3.6700, 3.6624, 3.7084, 3.6154, 3.4067, 3.8000, 3.4505], + device='cuda:0'), covar=tensor([0.0145, 0.0164, 0.0112, 0.0137, 0.0359, 0.0126, 0.0157, 0.0147], + device='cuda:0'), in_proj_covar=tensor([0.0089, 0.0096, 0.0093, 0.0082, 0.0166, 0.0102, 0.0099, 0.0102], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 04:56:55,484 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 04:57:07,553 INFO [train.py:901] (0/2) Epoch 28, batch 2550, loss[loss=0.1508, simple_loss=0.226, pruned_loss=0.0378, over 7307.00 frames. ], tot_loss[loss=0.1399, simple_loss=0.2201, pruned_loss=0.0298, over 1445856.78 frames. ], batch size: 83, lr: 5.86e-03, grad_scale: 8.0 +2023-03-21 04:57:16,610 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=78816.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:57:26,231 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.209e+02 1.802e+02 2.097e+02 2.497e+02 4.089e+02, threshold=4.193e+02, percent-clipped=0.0 +2023-03-21 04:57:33,691 INFO [train.py:901] (0/2) Epoch 28, batch 2600, loss[loss=0.1422, simple_loss=0.2146, pruned_loss=0.03489, over 7205.00 frames. ], tot_loss[loss=0.1395, simple_loss=0.2196, pruned_loss=0.02965, over 1444291.54 frames. ], batch size: 50, lr: 5.86e-03, grad_scale: 8.0 +2023-03-21 04:57:37,358 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.30 vs. limit=2.0 +2023-03-21 04:57:52,179 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.59 vs. limit=2.0 +2023-03-21 04:57:52,882 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=78888.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:57:58,184 INFO [train.py:901] (0/2) Epoch 28, batch 2650, loss[loss=0.1372, simple_loss=0.214, pruned_loss=0.03023, over 7312.00 frames. ], tot_loss[loss=0.139, simple_loss=0.2188, pruned_loss=0.0296, over 1444738.47 frames. ], batch size: 59, lr: 5.86e-03, grad_scale: 8.0 +2023-03-21 04:58:04,645 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=78912.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:58:15,876 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.313e+02 1.795e+02 2.054e+02 2.347e+02 4.847e+02, threshold=4.109e+02, percent-clipped=2.0 +2023-03-21 04:58:16,947 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=78936.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:58:23,309 INFO [train.py:901] (0/2) Epoch 28, batch 2700, loss[loss=0.1419, simple_loss=0.2315, pruned_loss=0.02618, over 7344.00 frames. ], tot_loss[loss=0.1384, simple_loss=0.2181, pruned_loss=0.02939, over 1441787.27 frames. ], batch size: 61, lr: 5.85e-03, grad_scale: 8.0 +2023-03-21 04:58:31,353 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=78965.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:58:47,900 INFO [train.py:901] (0/2) Epoch 28, batch 2750, loss[loss=0.1462, simple_loss=0.2269, pruned_loss=0.03275, over 7310.00 frames. ], tot_loss[loss=0.1393, simple_loss=0.2189, pruned_loss=0.02982, over 1440819.95 frames. ], batch size: 59, lr: 5.85e-03, grad_scale: 8.0 +2023-03-21 04:58:48,059 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3813, 1.5328, 1.3509, 1.6340, 1.5802, 1.4318, 1.5377, 1.0722], + device='cuda:0'), covar=tensor([0.0124, 0.0196, 0.0235, 0.0116, 0.0072, 0.0171, 0.0145, 0.0178], + device='cuda:0'), in_proj_covar=tensor([0.0032, 0.0030, 0.0029, 0.0030, 0.0029, 0.0028, 0.0029, 0.0039], + device='cuda:0'), out_proj_covar=tensor([3.6503e-05, 3.3416e-05, 3.2477e-05, 3.3214e-05, 3.2194e-05, 3.1481e-05, + 3.3337e-05, 4.3389e-05], device='cuda:0') +2023-03-21 04:59:01,608 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=79026.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 04:59:05,338 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.304e+02 1.791e+02 2.019e+02 2.450e+02 4.572e+02, threshold=4.037e+02, percent-clipped=2.0 +2023-03-21 04:59:12,704 INFO [train.py:901] (0/2) Epoch 28, batch 2800, loss[loss=0.1556, simple_loss=0.2388, pruned_loss=0.0362, over 6490.00 frames. ], tot_loss[loss=0.1399, simple_loss=0.2199, pruned_loss=0.0299, over 1439801.98 frames. ], batch size: 106, lr: 5.85e-03, grad_scale: 8.0 +2023-03-21 04:59:25,609 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-28.pt +2023-03-21 04:59:37,209 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 04:59:37,264 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 04:59:37,293 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 04:59:37,311 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 04:59:37,486 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 04:59:37,501 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 04:59:37,513 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 04:59:37,612 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 04:59:37,693 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 04:59:37,708 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 04:59:37,726 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 04:59:38,258 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 04:59:38,263 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 04:59:38,272 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 04:59:38,423 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 04:59:38,426 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 04:59:38,485 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 04:59:40,822 INFO [train.py:901] (0/2) Epoch 29, batch 0, loss[loss=0.1351, simple_loss=0.2239, pruned_loss=0.02313, over 7266.00 frames. ], tot_loss[loss=0.1351, simple_loss=0.2239, pruned_loss=0.02313, over 7266.00 frames. ], batch size: 89, lr: 5.75e-03, grad_scale: 8.0 +2023-03-21 04:59:40,824 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 05:00:03,405 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3719, 3.9736, 3.9343, 4.1046, 4.0926, 3.8716, 4.2705, 3.9267], + device='cuda:0'), covar=tensor([0.0144, 0.0160, 0.0148, 0.0151, 0.0409, 0.0155, 0.0149, 0.0150], + device='cuda:0'), in_proj_covar=tensor([0.0090, 0.0098, 0.0095, 0.0084, 0.0168, 0.0103, 0.0101, 0.0104], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:00:07,206 INFO [train.py:935] (0/2) Epoch 29, validation: loss=0.1643, simple_loss=0.2541, pruned_loss=0.03725, over 1622729.00 frames. +2023-03-21 05:00:07,207 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 05:00:09,542 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.28 vs. limit=5.0 +2023-03-21 05:00:13,241 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 05:00:24,316 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 05:00:29,057 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=79116.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:00:30,985 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 05:00:32,446 INFO [train.py:901] (0/2) Epoch 29, batch 50, loss[loss=0.1446, simple_loss=0.2248, pruned_loss=0.03222, over 7306.00 frames. ], tot_loss[loss=0.1384, simple_loss=0.2198, pruned_loss=0.02851, over 326772.34 frames. ], batch size: 80, lr: 5.75e-03, grad_scale: 16.0 +2023-03-21 05:00:33,507 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 05:00:37,148 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 05:00:38,554 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.286e+02 1.920e+02 2.167e+02 2.557e+02 4.902e+02, threshold=4.333e+02, percent-clipped=3.0 +2023-03-21 05:00:53,492 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=79164.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:00:53,929 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 05:00:55,106 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 05:00:58,601 INFO [train.py:901] (0/2) Epoch 29, batch 100, loss[loss=0.1397, simple_loss=0.2202, pruned_loss=0.02964, over 7258.00 frames. ], tot_loss[loss=0.1371, simple_loss=0.2177, pruned_loss=0.02828, over 573818.96 frames. ], batch size: 57, lr: 5.75e-03, grad_scale: 16.0 +2023-03-21 05:01:07,036 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0147, 3.4526, 3.9698, 4.0293, 4.0704, 4.0673, 4.1477, 3.9623], + device='cuda:0'), covar=tensor([0.0035, 0.0100, 0.0030, 0.0033, 0.0028, 0.0029, 0.0031, 0.0049], + device='cuda:0'), in_proj_covar=tensor([0.0046, 0.0064, 0.0054, 0.0053, 0.0052, 0.0057, 0.0048, 0.0072], + device='cuda:0'), out_proj_covar=tensor([8.4216e-05, 1.4002e-04, 1.0667e-04, 9.8508e-05, 9.5417e-05, 1.0533e-04, + 1.0115e-04, 1.4023e-04], device='cuda:0') +2023-03-21 05:01:18,182 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.84 vs. limit=2.0 +2023-03-21 05:01:18,456 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=79212.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:01:20,526 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3484, 2.3494, 2.4052, 2.0197, 2.4860, 2.2632, 2.0186, 1.9052], + device='cuda:0'), covar=tensor([0.0231, 0.0374, 0.0298, 0.0289, 0.0655, 0.0510, 0.0427, 0.0319], + device='cuda:0'), in_proj_covar=tensor([0.0034, 0.0032, 0.0033, 0.0031, 0.0031, 0.0031, 0.0035, 0.0034], + device='cuda:0'), out_proj_covar=tensor([8.5369e-05, 8.4071e-05, 8.2632e-05, 7.9025e-05, 8.0911e-05, 8.0019e-05, + 8.5652e-05, 8.7419e-05], device='cuda:0') +2023-03-21 05:01:24,450 INFO [train.py:901] (0/2) Epoch 29, batch 150, loss[loss=0.1508, simple_loss=0.2219, pruned_loss=0.03984, over 7259.00 frames. ], tot_loss[loss=0.137, simple_loss=0.2173, pruned_loss=0.02837, over 762225.01 frames. ], batch size: 64, lr: 5.74e-03, grad_scale: 16.0 +2023-03-21 05:01:29,959 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.271e+02 1.931e+02 2.300e+02 2.732e+02 4.086e+02, threshold=4.601e+02, percent-clipped=1.0 +2023-03-21 05:01:43,586 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=79260.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:01:49,936 INFO [train.py:901] (0/2) Epoch 29, batch 200, loss[loss=0.1498, simple_loss=0.2304, pruned_loss=0.03462, over 7349.00 frames. ], tot_loss[loss=0.1377, simple_loss=0.2175, pruned_loss=0.02899, over 913302.05 frames. ], batch size: 63, lr: 5.74e-03, grad_scale: 16.0 +2023-03-21 05:01:55,417 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 05:01:55,983 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=79285.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 05:02:00,296 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 05:02:05,347 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 05:02:14,502 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=79321.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:02:15,453 INFO [train.py:901] (0/2) Epoch 29, batch 250, loss[loss=0.1217, simple_loss=0.199, pruned_loss=0.02219, over 7276.00 frames. ], tot_loss[loss=0.138, simple_loss=0.2177, pruned_loss=0.02919, over 1030394.86 frames. ], batch size: 64, lr: 5.74e-03, grad_scale: 16.0 +2023-03-21 05:02:18,893 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 05:02:20,546 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=79333.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:02:20,897 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.137e+02 1.752e+02 2.021e+02 2.403e+02 4.904e+02, threshold=4.043e+02, percent-clipped=2.0 +2023-03-21 05:02:27,659 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=79346.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 05:02:36,467 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=79364.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:02:39,335 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 05:02:40,817 INFO [train.py:901] (0/2) Epoch 29, batch 300, loss[loss=0.1126, simple_loss=0.1715, pruned_loss=0.02683, over 6088.00 frames. ], tot_loss[loss=0.1395, simple_loss=0.2195, pruned_loss=0.02977, over 1122092.28 frames. ], batch size: 26, lr: 5.74e-03, grad_scale: 16.0 +2023-03-21 05:02:42,022 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=79375.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 05:02:48,368 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 05:02:51,075 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.91 vs. limit=5.0 +2023-03-21 05:02:52,079 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=79394.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:03:06,647 INFO [train.py:901] (0/2) Epoch 29, batch 350, loss[loss=0.1423, simple_loss=0.2208, pruned_loss=0.03191, over 7285.00 frames. ], tot_loss[loss=0.1381, simple_loss=0.2182, pruned_loss=0.02894, over 1193129.96 frames. ], batch size: 68, lr: 5.74e-03, grad_scale: 16.0 +2023-03-21 05:03:07,821 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=79425.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:03:08,998 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8672, 2.6032, 1.9067, 2.9822, 2.1624, 2.5312, 1.2413, 1.8993], + device='cuda:0'), covar=tensor([0.0545, 0.1165, 0.2437, 0.0845, 0.0476, 0.0470, 0.3246, 0.1595], + device='cuda:0'), in_proj_covar=tensor([0.0277, 0.0253, 0.0289, 0.0270, 0.0269, 0.0267, 0.0249, 0.0269], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 05:03:12,767 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.257e+02 2.121e+02 2.387e+02 2.963e+02 4.532e+02, threshold=4.773e+02, percent-clipped=2.0 +2023-03-21 05:03:13,894 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=79436.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 05:03:21,285 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 05:03:32,726 INFO [train.py:901] (0/2) Epoch 29, batch 400, loss[loss=0.1219, simple_loss=0.1981, pruned_loss=0.02288, over 7174.00 frames. ], tot_loss[loss=0.139, simple_loss=0.2191, pruned_loss=0.02946, over 1247936.28 frames. ], batch size: 41, lr: 5.74e-03, grad_scale: 16.0 +2023-03-21 05:03:48,442 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1383, 3.6972, 3.4648, 3.9396, 3.5041, 3.4008, 4.1044, 3.0049], + device='cuda:0'), covar=tensor([0.0419, 0.0479, 0.0351, 0.0449, 0.0741, 0.0789, 0.0617, 0.1255], + device='cuda:0'), in_proj_covar=tensor([0.0331, 0.0337, 0.0269, 0.0353, 0.0295, 0.0294, 0.0343, 0.0266], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 05:03:55,664 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.62 vs. limit=2.0 +2023-03-21 05:03:58,245 INFO [train.py:901] (0/2) Epoch 29, batch 450, loss[loss=0.1375, simple_loss=0.2262, pruned_loss=0.02437, over 7277.00 frames. ], tot_loss[loss=0.1384, simple_loss=0.2185, pruned_loss=0.02913, over 1290170.22 frames. ], batch size: 66, lr: 5.73e-03, grad_scale: 16.0 +2023-03-21 05:04:03,214 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 05:04:03,228 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 05:04:04,212 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.097e+02 1.851e+02 2.031e+02 2.665e+02 3.786e+02, threshold=4.062e+02, percent-clipped=0.0 +2023-03-21 05:04:07,352 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0714, 3.1346, 3.2468, 2.8455, 3.2958, 3.2566, 2.8213, 3.2555], + device='cuda:0'), covar=tensor([0.1303, 0.0674, 0.1452, 0.2393, 0.1384, 0.0841, 0.2078, 0.1812], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0063, 0.0048, 0.0047, 0.0047, 0.0045, 0.0065, 0.0048], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 05:04:23,690 INFO [train.py:901] (0/2) Epoch 29, batch 500, loss[loss=0.1356, simple_loss=0.216, pruned_loss=0.0276, over 7272.00 frames. ], tot_loss[loss=0.1376, simple_loss=0.2177, pruned_loss=0.02873, over 1326087.57 frames. ], batch size: 52, lr: 5.73e-03, grad_scale: 8.0 +2023-03-21 05:04:29,829 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=79585.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:04:36,198 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 05:04:38,035 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 05:04:38,528 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 05:04:41,608 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 05:04:45,527 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 05:04:48,599 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=79621.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:04:49,484 INFO [train.py:901] (0/2) Epoch 29, batch 550, loss[loss=0.1399, simple_loss=0.2194, pruned_loss=0.03019, over 7268.00 frames. ], tot_loss[loss=0.1372, simple_loss=0.2175, pruned_loss=0.02848, over 1351602.31 frames. ], batch size: 70, lr: 5.73e-03, grad_scale: 8.0 +2023-03-21 05:04:55,504 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.176e+02 1.766e+02 2.094e+02 2.391e+02 9.522e+02, threshold=4.188e+02, percent-clipped=4.0 +2023-03-21 05:04:55,529 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 05:04:58,552 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=79641.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 05:05:01,154 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=79646.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:05:04,538 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 05:05:07,039 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=79657.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:05:07,445 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 05:05:10,804 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.81 vs. limit=5.0 +2023-03-21 05:05:13,013 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=79669.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:05:14,482 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 05:05:14,966 INFO [train.py:901] (0/2) Epoch 29, batch 600, loss[loss=0.1436, simple_loss=0.2237, pruned_loss=0.0317, over 7280.00 frames. ], tot_loss[loss=0.1376, simple_loss=0.2179, pruned_loss=0.02859, over 1372446.27 frames. ], batch size: 66, lr: 5.73e-03, grad_scale: 8.0 +2023-03-21 05:05:18,503 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8469, 4.0833, 3.9418, 4.0572, 3.7185, 4.1168, 4.3653, 4.4183], + device='cuda:0'), covar=tensor([0.0231, 0.0159, 0.0168, 0.0154, 0.0315, 0.0196, 0.0215, 0.0160], + device='cuda:0'), in_proj_covar=tensor([0.0112, 0.0116, 0.0106, 0.0113, 0.0103, 0.0093, 0.0091, 0.0089], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:05:18,577 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5796, 2.6110, 2.2214, 3.8785, 1.7388, 3.5448, 1.4301, 3.0035], + device='cuda:0'), covar=tensor([0.0133, 0.1175, 0.1919, 0.0167, 0.4066, 0.0211, 0.1207, 0.0396], + device='cuda:0'), in_proj_covar=tensor([0.0198, 0.0259, 0.0273, 0.0200, 0.0259, 0.0210, 0.0245, 0.0228], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 05:05:23,039 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=79689.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:05:31,997 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 05:05:35,752 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.48 vs. limit=2.0 +2023-03-21 05:05:38,055 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=79718.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:05:39,003 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=79720.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:05:39,060 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2765, 4.2704, 3.5585, 3.7655, 3.4025, 2.4133, 1.9778, 4.2807], + device='cuda:0'), covar=tensor([0.0042, 0.0052, 0.0114, 0.0064, 0.0114, 0.0471, 0.0583, 0.0050], + device='cuda:0'), in_proj_covar=tensor([0.0102, 0.0084, 0.0104, 0.0087, 0.0118, 0.0126, 0.0124, 0.0096], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 05:05:40,448 INFO [train.py:901] (0/2) Epoch 29, batch 650, loss[loss=0.1486, simple_loss=0.2244, pruned_loss=0.03642, over 7343.00 frames. ], tot_loss[loss=0.1374, simple_loss=0.2174, pruned_loss=0.02868, over 1385597.30 frames. ], batch size: 73, lr: 5.73e-03, grad_scale: 8.0 +2023-03-21 05:05:40,945 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 05:05:42,984 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8853, 3.1361, 3.7544, 3.8635, 4.0759, 3.9739, 4.0172, 3.8350], + device='cuda:0'), covar=tensor([0.0037, 0.0130, 0.0036, 0.0039, 0.0031, 0.0031, 0.0036, 0.0051], + device='cuda:0'), in_proj_covar=tensor([0.0046, 0.0064, 0.0054, 0.0052, 0.0052, 0.0056, 0.0048, 0.0071], + device='cuda:0'), out_proj_covar=tensor([8.3260e-05, 1.3889e-04, 1.0706e-04, 9.7556e-05, 9.4041e-05, 1.0420e-04, + 9.9856e-05, 1.3822e-04], device='cuda:0') +2023-03-21 05:05:44,401 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=79731.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 05:05:46,326 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.341e+02 1.881e+02 2.206e+02 2.571e+02 4.157e+02, threshold=4.413e+02, percent-clipped=0.0 +2023-03-21 05:05:58,437 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 05:06:00,567 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=79762.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:06:05,898 INFO [train.py:901] (0/2) Epoch 29, batch 700, loss[loss=0.1568, simple_loss=0.2396, pruned_loss=0.03699, over 7240.00 frames. ], tot_loss[loss=0.1383, simple_loss=0.2184, pruned_loss=0.02909, over 1399133.88 frames. ], batch size: 55, lr: 5.72e-03, grad_scale: 8.0 +2023-03-21 05:06:06,421 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 05:06:31,473 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 05:06:31,981 INFO [train.py:901] (0/2) Epoch 29, batch 750, loss[loss=0.1082, simple_loss=0.1767, pruned_loss=0.01985, over 7056.00 frames. ], tot_loss[loss=0.1386, simple_loss=0.2185, pruned_loss=0.02932, over 1408896.75 frames. ], batch size: 35, lr: 5.72e-03, grad_scale: 8.0 +2023-03-21 05:06:31,985 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 05:06:32,102 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6921, 3.0081, 2.6853, 2.8086, 2.8830, 2.4966, 2.8319, 2.5478], + device='cuda:0'), covar=tensor([0.1018, 0.0690, 0.0883, 0.1260, 0.0861, 0.0642, 0.0742, 0.2046], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0054, 0.0061, 0.0053, 0.0051, 0.0054, 0.0052, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:06:32,120 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=79823.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:06:38,570 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.148e+02 1.855e+02 2.149e+02 2.429e+02 6.154e+02, threshold=4.298e+02, percent-clipped=2.0 +2023-03-21 05:06:46,675 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 05:06:52,083 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 05:06:57,596 INFO [train.py:901] (0/2) Epoch 29, batch 800, loss[loss=0.1121, simple_loss=0.1953, pruned_loss=0.0145, over 7345.00 frames. ], tot_loss[loss=0.1376, simple_loss=0.2175, pruned_loss=0.02885, over 1416153.76 frames. ], batch size: 44, lr: 5.72e-03, grad_scale: 8.0 +2023-03-21 05:06:58,719 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 05:06:59,767 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 05:07:10,171 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 05:07:12,763 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7445, 3.1580, 3.6628, 3.8331, 3.8821, 3.8280, 3.7160, 3.7124], + device='cuda:0'), covar=tensor([0.0035, 0.0112, 0.0033, 0.0029, 0.0032, 0.0032, 0.0042, 0.0050], + device='cuda:0'), in_proj_covar=tensor([0.0045, 0.0063, 0.0053, 0.0051, 0.0051, 0.0056, 0.0047, 0.0070], + device='cuda:0'), out_proj_covar=tensor([8.2198e-05, 1.3686e-04, 1.0427e-04, 9.5362e-05, 9.3176e-05, 1.0304e-04, + 9.7919e-05, 1.3647e-04], device='cuda:0') +2023-03-21 05:07:23,767 INFO [train.py:901] (0/2) Epoch 29, batch 850, loss[loss=0.1109, simple_loss=0.1934, pruned_loss=0.01423, over 7171.00 frames. ], tot_loss[loss=0.1376, simple_loss=0.2177, pruned_loss=0.02878, over 1420587.32 frames. ], batch size: 39, lr: 5.72e-03, grad_scale: 8.0 +2023-03-21 05:07:29,305 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 05:07:29,315 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 05:07:29,785 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.291e+02 1.737e+02 2.030e+02 2.344e+02 3.800e+02, threshold=4.059e+02, percent-clipped=0.0 +2023-03-21 05:07:32,890 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=79941.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:07:32,918 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=79941.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 05:07:34,863 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 05:07:38,441 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 05:07:49,413 INFO [train.py:901] (0/2) Epoch 29, batch 900, loss[loss=0.1303, simple_loss=0.2102, pruned_loss=0.02515, over 7322.00 frames. ], tot_loss[loss=0.1373, simple_loss=0.2174, pruned_loss=0.02856, over 1425325.84 frames. ], batch size: 49, lr: 5.72e-03, grad_scale: 8.0 +2023-03-21 05:07:57,540 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=79989.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 05:07:57,585 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=79989.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:08:01,607 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9738, 3.4440, 3.9238, 3.9735, 4.0690, 4.0145, 4.1155, 3.9055], + device='cuda:0'), covar=tensor([0.0030, 0.0100, 0.0034, 0.0035, 0.0032, 0.0031, 0.0032, 0.0047], + device='cuda:0'), in_proj_covar=tensor([0.0045, 0.0064, 0.0054, 0.0053, 0.0052, 0.0056, 0.0048, 0.0071], + device='cuda:0'), out_proj_covar=tensor([8.2939e-05, 1.3896e-04, 1.0610e-04, 9.7506e-05, 9.4489e-05, 1.0472e-04, + 1.0023e-04, 1.3881e-04], device='cuda:0') +2023-03-21 05:08:03,335 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-80000.pt +2023-03-21 05:08:13,985 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=80013.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:08:17,440 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=80020.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:08:18,837 INFO [train.py:901] (0/2) Epoch 29, batch 950, loss[loss=0.1461, simple_loss=0.2235, pruned_loss=0.03435, over 7270.00 frames. ], tot_loss[loss=0.137, simple_loss=0.2172, pruned_loss=0.02844, over 1430254.66 frames. ], batch size: 47, lr: 5.72e-03, grad_scale: 8.0 +2023-03-21 05:08:20,834 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 05:08:22,953 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=80031.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 05:08:24,816 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.204e+02 1.694e+02 2.042e+02 2.334e+02 4.192e+02, threshold=4.084e+02, percent-clipped=1.0 +2023-03-21 05:08:25,856 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=80037.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:08:41,914 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=80068.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:08:43,911 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 05:08:44,392 INFO [train.py:901] (0/2) Epoch 29, batch 1000, loss[loss=0.1404, simple_loss=0.2261, pruned_loss=0.02739, over 7336.00 frames. ], tot_loss[loss=0.1372, simple_loss=0.2173, pruned_loss=0.02854, over 1431013.24 frames. ], batch size: 61, lr: 5.71e-03, grad_scale: 8.0 +2023-03-21 05:08:47,418 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=80079.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 05:08:58,193 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5020, 3.5524, 2.3628, 3.9118, 2.9411, 3.4708, 1.6872, 2.2526], + device='cuda:0'), covar=tensor([0.0464, 0.1157, 0.2466, 0.0573, 0.0543, 0.0658, 0.3137, 0.1809], + device='cuda:0'), in_proj_covar=tensor([0.0277, 0.0255, 0.0285, 0.0268, 0.0268, 0.0267, 0.0244, 0.0267], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 05:09:05,362 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 05:09:07,460 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=80118.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:09:09,826 INFO [train.py:901] (0/2) Epoch 29, batch 1050, loss[loss=0.1357, simple_loss=0.2235, pruned_loss=0.02398, over 7257.00 frames. ], tot_loss[loss=0.138, simple_loss=0.2176, pruned_loss=0.02916, over 1434480.33 frames. ], batch size: 64, lr: 5.71e-03, grad_scale: 8.0 +2023-03-21 05:09:16,381 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.326e+02 1.871e+02 2.139e+02 2.585e+02 3.936e+02, threshold=4.278e+02, percent-clipped=0.0 +2023-03-21 05:09:27,005 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 05:09:30,993 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 05:09:35,420 INFO [train.py:901] (0/2) Epoch 29, batch 1100, loss[loss=0.141, simple_loss=0.2267, pruned_loss=0.02766, over 7296.00 frames. ], tot_loss[loss=0.1388, simple_loss=0.2185, pruned_loss=0.02956, over 1435062.40 frames. ], batch size: 57, lr: 5.71e-03, grad_scale: 8.0 +2023-03-21 05:09:59,453 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 05:09:59,994 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 05:10:01,979 INFO [train.py:901] (0/2) Epoch 29, batch 1150, loss[loss=0.1366, simple_loss=0.2224, pruned_loss=0.02547, over 7277.00 frames. ], tot_loss[loss=0.1381, simple_loss=0.218, pruned_loss=0.02906, over 1435749.44 frames. ], batch size: 77, lr: 5.71e-03, grad_scale: 8.0 +2023-03-21 05:10:08,131 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.333e+02 1.785e+02 2.047e+02 2.431e+02 3.454e+02, threshold=4.094e+02, percent-clipped=0.0 +2023-03-21 05:10:10,716 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 05:10:11,215 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 05:10:11,317 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=80241.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:10:27,691 INFO [train.py:901] (0/2) Epoch 29, batch 1200, loss[loss=0.1585, simple_loss=0.2383, pruned_loss=0.0394, over 7361.00 frames. ], tot_loss[loss=0.1386, simple_loss=0.2185, pruned_loss=0.02932, over 1435581.07 frames. ], batch size: 73, lr: 5.71e-03, grad_scale: 8.0 +2023-03-21 05:10:32,834 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=80283.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:10:35,733 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=80289.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:10:43,331 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 05:10:48,362 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=80313.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:10:53,184 INFO [train.py:901] (0/2) Epoch 29, batch 1250, loss[loss=0.1373, simple_loss=0.2175, pruned_loss=0.02849, over 7310.00 frames. ], tot_loss[loss=0.1395, simple_loss=0.2194, pruned_loss=0.02977, over 1438181.99 frames. ], batch size: 49, lr: 5.70e-03, grad_scale: 8.0 +2023-03-21 05:10:59,174 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.228e+02 1.776e+02 2.060e+02 2.370e+02 4.071e+02, threshold=4.121e+02, percent-clipped=0.0 +2023-03-21 05:11:03,892 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=80344.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:11:06,375 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 05:11:08,512 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8198, 2.4959, 2.8736, 2.5201, 2.8638, 2.7626, 2.3232, 2.9350], + device='cuda:0'), covar=tensor([0.1770, 0.0887, 0.1728, 0.2291, 0.0965, 0.1286, 0.2377, 0.1516], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0063, 0.0048, 0.0047, 0.0047, 0.0045, 0.0065, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 05:11:10,405 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 05:11:11,927 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 05:11:12,928 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=80361.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:11:14,489 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4307, 4.1410, 4.4923, 4.6322, 4.5986, 4.5800, 4.6021, 4.5797], + device='cuda:0'), covar=tensor([0.0019, 0.0070, 0.0023, 0.0019, 0.0021, 0.0021, 0.0021, 0.0031], + device='cuda:0'), in_proj_covar=tensor([0.0045, 0.0063, 0.0054, 0.0052, 0.0051, 0.0056, 0.0048, 0.0071], + device='cuda:0'), out_proj_covar=tensor([8.1585e-05, 1.3728e-04, 1.0621e-04, 9.7328e-05, 9.3612e-05, 1.0386e-04, + 1.0006e-04, 1.3757e-04], device='cuda:0') +2023-03-21 05:11:15,546 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1509, 4.0570, 3.5343, 3.6461, 3.0804, 2.3164, 1.9559, 4.1074], + device='cuda:0'), covar=tensor([0.0041, 0.0083, 0.0092, 0.0060, 0.0130, 0.0463, 0.0552, 0.0044], + device='cuda:0'), in_proj_covar=tensor([0.0103, 0.0085, 0.0105, 0.0089, 0.0120, 0.0127, 0.0125, 0.0097], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 05:11:19,042 INFO [train.py:901] (0/2) Epoch 29, batch 1300, loss[loss=0.134, simple_loss=0.2203, pruned_loss=0.02386, over 7237.00 frames. ], tot_loss[loss=0.1392, simple_loss=0.2194, pruned_loss=0.02954, over 1438226.84 frames. ], batch size: 93, lr: 5.70e-03, grad_scale: 8.0 +2023-03-21 05:11:19,666 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9208, 3.2524, 3.8549, 3.9876, 3.9948, 4.0088, 3.9626, 3.9471], + device='cuda:0'), covar=tensor([0.0027, 0.0109, 0.0035, 0.0027, 0.0029, 0.0029, 0.0044, 0.0045], + device='cuda:0'), in_proj_covar=tensor([0.0045, 0.0063, 0.0054, 0.0052, 0.0051, 0.0056, 0.0048, 0.0070], + device='cuda:0'), out_proj_covar=tensor([8.1531e-05, 1.3710e-04, 1.0609e-04, 9.7347e-05, 9.3640e-05, 1.0364e-04, + 9.9961e-05, 1.3737e-04], device='cuda:0') +2023-03-21 05:11:29,897 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=80393.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:11:36,153 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 05:11:38,143 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 05:11:41,604 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 05:11:42,741 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=80418.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:11:45,131 INFO [train.py:901] (0/2) Epoch 29, batch 1350, loss[loss=0.1397, simple_loss=0.2191, pruned_loss=0.03019, over 7353.00 frames. ], tot_loss[loss=0.1385, simple_loss=0.219, pruned_loss=0.02907, over 1438974.70 frames. ], batch size: 63, lr: 5.70e-03, grad_scale: 8.0 +2023-03-21 05:11:51,634 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.192e+02 1.765e+02 2.078e+02 2.410e+02 4.380e+02, threshold=4.156e+02, percent-clipped=1.0 +2023-03-21 05:11:52,168 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 05:11:55,430 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.60 vs. limit=5.0 +2023-03-21 05:12:01,206 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=80454.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:12:05,652 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=80463.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:12:06,994 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=80466.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:12:10,974 INFO [train.py:901] (0/2) Epoch 29, batch 1400, loss[loss=0.1593, simple_loss=0.2418, pruned_loss=0.03843, over 6797.00 frames. ], tot_loss[loss=0.1387, simple_loss=0.2189, pruned_loss=0.02924, over 1438820.67 frames. ], batch size: 107, lr: 5.70e-03, grad_scale: 8.0 +2023-03-21 05:12:21,796 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=80494.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:12:24,698 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 05:12:33,757 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4906, 1.6806, 1.3632, 1.6474, 1.7161, 1.4852, 1.5896, 1.2156], + device='cuda:0'), covar=tensor([0.0131, 0.0134, 0.0262, 0.0092, 0.0084, 0.0087, 0.0138, 0.0130], + device='cuda:0'), in_proj_covar=tensor([0.0033, 0.0030, 0.0030, 0.0031, 0.0030, 0.0029, 0.0031, 0.0040], + device='cuda:0'), out_proj_covar=tensor([3.7532e-05, 3.4309e-05, 3.3625e-05, 3.4301e-05, 3.3636e-05, 3.2905e-05, + 3.4848e-05, 4.4673e-05], device='cuda:0') +2023-03-21 05:12:36,601 INFO [train.py:901] (0/2) Epoch 29, batch 1450, loss[loss=0.1643, simple_loss=0.2412, pruned_loss=0.04374, over 7136.00 frames. ], tot_loss[loss=0.1385, simple_loss=0.2186, pruned_loss=0.02915, over 1439651.95 frames. ], batch size: 98, lr: 5.70e-03, grad_scale: 8.0 +2023-03-21 05:12:37,260 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=80524.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:12:42,705 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.303e+02 1.883e+02 2.179e+02 2.600e+02 4.285e+02, threshold=4.357e+02, percent-clipped=2.0 +2023-03-21 05:12:48,552 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.86 vs. limit=2.0 +2023-03-21 05:12:48,784 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 05:12:53,502 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=80555.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:12:54,685 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.32 vs. limit=5.0 +2023-03-21 05:12:55,007 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2102, 3.5198, 2.7703, 3.2238, 3.1485, 2.7814, 3.2163, 2.7371], + device='cuda:0'), covar=tensor([0.0722, 0.0624, 0.1824, 0.1184, 0.1607, 0.1121, 0.0781, 0.1929], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0053, 0.0061, 0.0053, 0.0051, 0.0055, 0.0052, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:13:02,257 INFO [train.py:901] (0/2) Epoch 29, batch 1500, loss[loss=0.1498, simple_loss=0.2259, pruned_loss=0.03686, over 7314.00 frames. ], tot_loss[loss=0.1379, simple_loss=0.2181, pruned_loss=0.02888, over 1439203.43 frames. ], batch size: 49, lr: 5.70e-03, grad_scale: 8.0 +2023-03-21 05:13:04,815 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 05:13:18,267 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7906, 2.3555, 3.0023, 2.7094, 2.7801, 2.7427, 2.3694, 2.7873], + device='cuda:0'), covar=tensor([0.1277, 0.0710, 0.0994, 0.1552, 0.1148, 0.0936, 0.2055, 0.1681], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0063, 0.0048, 0.0047, 0.0047, 0.0045, 0.0064, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 05:13:26,885 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5022, 2.1676, 2.7219, 2.6372, 2.6621, 2.8614, 2.5220, 2.4917], + device='cuda:0'), covar=tensor([0.0539, 0.0837, 0.0357, 0.0171, 0.0664, 0.0337, 0.0237, 0.0248], + device='cuda:0'), in_proj_covar=tensor([0.0034, 0.0033, 0.0033, 0.0031, 0.0031, 0.0030, 0.0035, 0.0035], + device='cuda:0'), out_proj_covar=tensor([8.6338e-05, 8.5083e-05, 8.3122e-05, 8.0057e-05, 8.1912e-05, 7.9715e-05, + 8.6444e-05, 8.8901e-05], device='cuda:0') +2023-03-21 05:13:26,912 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8694, 2.2444, 1.5415, 2.1697, 2.2858, 1.8444, 1.7475, 1.5648], + device='cuda:0'), covar=tensor([0.0104, 0.0090, 0.0294, 0.0121, 0.0082, 0.0128, 0.0145, 0.0176], + device='cuda:0'), in_proj_covar=tensor([0.0032, 0.0030, 0.0029, 0.0030, 0.0029, 0.0029, 0.0030, 0.0039], + device='cuda:0'), out_proj_covar=tensor([3.6935e-05, 3.3629e-05, 3.3022e-05, 3.3704e-05, 3.2879e-05, 3.2263e-05, + 3.4403e-05, 4.3982e-05], device='cuda:0') +2023-03-21 05:13:28,301 INFO [train.py:901] (0/2) Epoch 29, batch 1550, loss[loss=0.1066, simple_loss=0.1779, pruned_loss=0.01768, over 7012.00 frames. ], tot_loss[loss=0.1381, simple_loss=0.218, pruned_loss=0.02904, over 1437213.88 frames. ], batch size: 35, lr: 5.69e-03, grad_scale: 8.0 +2023-03-21 05:13:29,840 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 05:13:29,937 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1193, 3.7784, 3.7285, 3.8748, 3.8019, 3.6594, 3.9698, 3.5290], + device='cuda:0'), covar=tensor([0.0185, 0.0177, 0.0141, 0.0157, 0.0415, 0.0122, 0.0162, 0.0171], + device='cuda:0'), in_proj_covar=tensor([0.0087, 0.0096, 0.0093, 0.0082, 0.0164, 0.0101, 0.0098, 0.0101], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:13:34,384 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.279e+02 1.859e+02 2.000e+02 2.568e+02 5.498e+02, threshold=3.999e+02, percent-clipped=3.0 +2023-03-21 05:13:36,487 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=80639.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:13:41,493 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-21 05:13:51,606 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4847, 5.0344, 5.1113, 5.0586, 4.8030, 4.4864, 5.1520, 4.8949], + device='cuda:0'), covar=tensor([0.0501, 0.0430, 0.0393, 0.0467, 0.0390, 0.0421, 0.0307, 0.0532], + device='cuda:0'), in_proj_covar=tensor([0.0209, 0.0242, 0.0184, 0.0187, 0.0149, 0.0218, 0.0195, 0.0143], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 05:13:54,013 INFO [train.py:901] (0/2) Epoch 29, batch 1600, loss[loss=0.1417, simple_loss=0.2225, pruned_loss=0.03043, over 7283.00 frames. ], tot_loss[loss=0.138, simple_loss=0.2183, pruned_loss=0.02886, over 1438282.16 frames. ], batch size: 57, lr: 5.69e-03, grad_scale: 8.0 +2023-03-21 05:13:58,689 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5987, 3.7924, 3.5823, 3.6880, 3.4017, 3.6486, 4.0560, 4.0729], + device='cuda:0'), covar=tensor([0.0235, 0.0179, 0.0230, 0.0217, 0.0392, 0.0485, 0.0240, 0.0205], + device='cuda:0'), in_proj_covar=tensor([0.0115, 0.0119, 0.0109, 0.0116, 0.0107, 0.0096, 0.0092, 0.0092], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:14:01,554 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 05:14:02,059 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 05:14:05,025 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 05:14:10,538 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.71 vs. limit=5.0 +2023-03-21 05:14:10,719 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.3198, 4.8661, 4.6749, 5.3463, 5.0728, 5.2892, 4.6612, 4.8638], + device='cuda:0'), covar=tensor([0.0683, 0.2218, 0.1958, 0.0833, 0.0864, 0.1015, 0.0720, 0.1058], + device='cuda:0'), in_proj_covar=tensor([0.0239, 0.0374, 0.0282, 0.0292, 0.0217, 0.0350, 0.0212, 0.0261], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:14:15,124 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 05:14:19,245 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 05:14:19,736 INFO [train.py:901] (0/2) Epoch 29, batch 1650, loss[loss=0.116, simple_loss=0.1939, pruned_loss=0.01907, over 6970.00 frames. ], tot_loss[loss=0.1375, simple_loss=0.2176, pruned_loss=0.02872, over 1440866.05 frames. ], batch size: 35, lr: 5.69e-03, grad_scale: 8.0 +2023-03-21 05:14:26,394 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.202e+02 1.728e+02 1.955e+02 2.318e+02 5.387e+02, threshold=3.909e+02, percent-clipped=3.0 +2023-03-21 05:14:27,077 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=80736.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:14:28,039 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 05:14:33,592 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=80749.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:14:43,563 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9346, 2.0201, 2.4437, 2.0608, 2.3783, 2.2619, 2.1619, 1.8173], + device='cuda:0'), covar=tensor([0.0621, 0.0511, 0.0208, 0.0188, 0.0627, 0.0427, 0.0285, 0.0445], + device='cuda:0'), in_proj_covar=tensor([0.0034, 0.0032, 0.0033, 0.0031, 0.0031, 0.0030, 0.0034, 0.0035], + device='cuda:0'), out_proj_covar=tensor([8.5568e-05, 8.4017e-05, 8.2082e-05, 7.9186e-05, 8.0442e-05, 7.9108e-05, + 8.5015e-05, 8.8658e-05], device='cuda:0') +2023-03-21 05:14:44,922 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 05:14:45,408 INFO [train.py:901] (0/2) Epoch 29, batch 1700, loss[loss=0.1449, simple_loss=0.2258, pruned_loss=0.032, over 7312.00 frames. ], tot_loss[loss=0.1374, simple_loss=0.2175, pruned_loss=0.02862, over 1441382.19 frames. ], batch size: 59, lr: 5.69e-03, grad_scale: 8.0 +2023-03-21 05:14:45,547 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=80773.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:14:46,573 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9862, 3.9928, 3.2247, 3.4545, 2.8176, 2.3060, 1.8968, 4.0125], + device='cuda:0'), covar=tensor([0.0047, 0.0043, 0.0118, 0.0076, 0.0166, 0.0503, 0.0628, 0.0052], + device='cuda:0'), in_proj_covar=tensor([0.0105, 0.0087, 0.0106, 0.0091, 0.0122, 0.0129, 0.0128, 0.0099], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 05:14:49,445 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 05:14:58,103 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=80797.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:14:59,882 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2458, 2.7616, 3.4177, 2.9683, 3.3537, 3.0661, 2.8226, 3.1236], + device='cuda:0'), covar=tensor([0.1556, 0.0823, 0.1034, 0.2261, 0.0952, 0.0895, 0.1869, 0.2147], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0064, 0.0049, 0.0048, 0.0047, 0.0046, 0.0065, 0.0050], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 05:15:00,356 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.6244, 5.0962, 5.0408, 5.6639, 5.4304, 5.5599, 5.0569, 4.9362], + device='cuda:0'), covar=tensor([0.0724, 0.2391, 0.2027, 0.0746, 0.0978, 0.1081, 0.0690, 0.1257], + device='cuda:0'), in_proj_covar=tensor([0.0240, 0.0377, 0.0282, 0.0292, 0.0218, 0.0351, 0.0213, 0.0260], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:15:00,836 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 05:15:09,976 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=80819.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:15:11,910 INFO [train.py:901] (0/2) Epoch 29, batch 1750, loss[loss=0.1266, simple_loss=0.2131, pruned_loss=0.02002, over 7345.00 frames. ], tot_loss[loss=0.1373, simple_loss=0.2173, pruned_loss=0.02867, over 1440201.79 frames. ], batch size: 73, lr: 5.69e-03, grad_scale: 8.0 +2023-03-21 05:15:17,551 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=80834.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:15:17,860 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.159e+02 1.786e+02 2.210e+02 2.520e+02 4.800e+02, threshold=4.419e+02, percent-clipped=3.0 +2023-03-21 05:15:24,270 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 05:15:25,298 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 05:15:25,338 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=80850.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:15:27,325 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2666, 4.5369, 4.3540, 4.5516, 4.4255, 4.2635, 4.6369, 4.4248], + device='cuda:0'), covar=tensor([0.0951, 0.1163, 0.1206, 0.1139, 0.0859, 0.0812, 0.0788, 0.1119], + device='cuda:0'), in_proj_covar=tensor([0.0210, 0.0243, 0.0185, 0.0188, 0.0150, 0.0219, 0.0196, 0.0144], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 05:15:31,036 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.00 vs. limit=2.0 +2023-03-21 05:15:37,356 INFO [train.py:901] (0/2) Epoch 29, batch 1800, loss[loss=0.1294, simple_loss=0.2139, pruned_loss=0.02245, over 7353.00 frames. ], tot_loss[loss=0.1374, simple_loss=0.2177, pruned_loss=0.02853, over 1440528.26 frames. ], batch size: 73, lr: 5.69e-03, grad_scale: 8.0 +2023-03-21 05:15:47,381 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 05:16:00,780 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 05:16:02,663 INFO [train.py:901] (0/2) Epoch 29, batch 1850, loss[loss=0.1515, simple_loss=0.2279, pruned_loss=0.03752, over 7278.00 frames. ], tot_loss[loss=0.1386, simple_loss=0.2189, pruned_loss=0.02912, over 1442817.49 frames. ], batch size: 89, lr: 5.68e-03, grad_scale: 8.0 +2023-03-21 05:16:05,267 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8377, 5.3548, 5.4515, 5.3669, 5.1606, 4.7897, 5.5091, 5.2751], + device='cuda:0'), covar=tensor([0.0435, 0.0383, 0.0310, 0.0429, 0.0338, 0.0385, 0.0257, 0.0475], + device='cuda:0'), in_proj_covar=tensor([0.0210, 0.0245, 0.0185, 0.0189, 0.0150, 0.0219, 0.0196, 0.0144], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 05:16:08,665 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.303e+02 1.763e+02 2.062e+02 2.438e+02 5.216e+02, threshold=4.123e+02, percent-clipped=1.0 +2023-03-21 05:16:10,709 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 05:16:10,808 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=80939.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:16:24,466 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7941, 2.2618, 2.8861, 2.5640, 2.7313, 2.6138, 2.3948, 2.7509], + device='cuda:0'), covar=tensor([0.1410, 0.0863, 0.0807, 0.1449, 0.0950, 0.1016, 0.1861, 0.1179], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0061, 0.0047, 0.0047, 0.0046, 0.0044, 0.0063, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 05:16:27,934 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 05:16:28,402 INFO [train.py:901] (0/2) Epoch 29, batch 1900, loss[loss=0.154, simple_loss=0.2381, pruned_loss=0.03493, over 7141.00 frames. ], tot_loss[loss=0.1386, simple_loss=0.219, pruned_loss=0.02913, over 1442250.33 frames. ], batch size: 98, lr: 5.68e-03, grad_scale: 8.0 +2023-03-21 05:16:29,998 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8170, 3.9612, 3.8119, 3.9747, 3.7504, 3.9545, 4.2605, 4.3279], + device='cuda:0'), covar=tensor([0.0187, 0.0142, 0.0177, 0.0162, 0.0246, 0.0265, 0.0206, 0.0134], + device='cuda:0'), in_proj_covar=tensor([0.0113, 0.0118, 0.0108, 0.0114, 0.0105, 0.0094, 0.0091, 0.0090], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:16:35,370 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=80987.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:16:52,828 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 05:16:54,336 INFO [train.py:901] (0/2) Epoch 29, batch 1950, loss[loss=0.1422, simple_loss=0.2281, pruned_loss=0.02812, over 7280.00 frames. ], tot_loss[loss=0.1391, simple_loss=0.2194, pruned_loss=0.02937, over 1443772.09 frames. ], batch size: 77, lr: 5.68e-03, grad_scale: 8.0 +2023-03-21 05:16:59,682 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-03-21 05:17:00,347 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.014e+02 1.770e+02 2.041e+02 2.360e+02 4.145e+02, threshold=4.083e+02, percent-clipped=1.0 +2023-03-21 05:17:03,412 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 05:17:08,030 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=81049.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:17:08,457 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 05:17:08,944 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 05:17:19,952 INFO [train.py:901] (0/2) Epoch 29, batch 2000, loss[loss=0.1438, simple_loss=0.2238, pruned_loss=0.03191, over 7298.00 frames. ], tot_loss[loss=0.1388, simple_loss=0.2195, pruned_loss=0.02906, over 1445089.79 frames. ], batch size: 86, lr: 5.68e-03, grad_scale: 8.0 +2023-03-21 05:17:25,548 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 05:17:30,204 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=81092.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:17:32,719 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=81097.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:17:36,641 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 05:17:39,740 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2919, 2.3802, 2.7090, 2.2577, 2.5396, 2.4416, 2.1395, 2.2188], + device='cuda:0'), covar=tensor([0.0444, 0.0429, 0.0228, 0.0264, 0.0443, 0.0425, 0.0316, 0.0313], + device='cuda:0'), in_proj_covar=tensor([0.0034, 0.0032, 0.0033, 0.0031, 0.0031, 0.0030, 0.0034, 0.0035], + device='cuda:0'), out_proj_covar=tensor([8.5439e-05, 8.4032e-05, 8.2098e-05, 8.0262e-05, 8.0759e-05, 7.9014e-05, + 8.5291e-05, 8.9129e-05], device='cuda:0') +2023-03-21 05:17:40,287 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2114, 2.5527, 1.8809, 3.1795, 3.1038, 3.3045, 2.2670, 2.8622], + device='cuda:0'), covar=tensor([0.2117, 0.1158, 0.3870, 0.0734, 0.0262, 0.0238, 0.0364, 0.0454], + device='cuda:0'), in_proj_covar=tensor([0.0259, 0.0232, 0.0258, 0.0265, 0.0189, 0.0183, 0.0209, 0.0221], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 05:17:41,816 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4764, 1.2259, 1.6461, 2.0244, 1.6162, 2.0003, 1.4043, 2.0265], + device='cuda:0'), covar=tensor([0.2423, 0.4085, 0.1898, 0.1203, 0.1916, 0.1839, 0.1770, 0.1985], + device='cuda:0'), in_proj_covar=tensor([0.0066, 0.0068, 0.0058, 0.0052, 0.0055, 0.0055, 0.0085, 0.0055], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:17:42,275 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=81116.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:17:43,784 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=81119.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:17:44,195 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 05:17:45,681 INFO [train.py:901] (0/2) Epoch 29, batch 2050, loss[loss=0.1334, simple_loss=0.2166, pruned_loss=0.02514, over 7250.00 frames. ], tot_loss[loss=0.1377, simple_loss=0.2186, pruned_loss=0.02846, over 1445542.15 frames. ], batch size: 89, lr: 5.68e-03, grad_scale: 8.0 +2023-03-21 05:17:48,711 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=81129.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:17:48,802 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3942, 2.4449, 2.2825, 3.5867, 1.7955, 3.4643, 1.5122, 3.1070], + device='cuda:0'), covar=tensor([0.0177, 0.1217, 0.1689, 0.0156, 0.3655, 0.0241, 0.1127, 0.0389], + device='cuda:0'), in_proj_covar=tensor([0.0201, 0.0260, 0.0271, 0.0201, 0.0258, 0.0214, 0.0243, 0.0230], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 05:17:52,342 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.043e+02 1.747e+02 2.128e+02 2.530e+02 3.662e+02, threshold=4.256e+02, percent-clipped=0.0 +2023-03-21 05:17:59,403 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.08 vs. limit=2.0 +2023-03-21 05:18:00,130 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=81150.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:18:07,799 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-03-21 05:18:08,560 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=81167.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:18:12,083 INFO [train.py:901] (0/2) Epoch 29, batch 2100, loss[loss=0.1325, simple_loss=0.2213, pruned_loss=0.02182, over 7319.00 frames. ], tot_loss[loss=0.1375, simple_loss=0.2182, pruned_loss=0.02836, over 1442206.54 frames. ], batch size: 59, lr: 5.67e-03, grad_scale: 8.0 +2023-03-21 05:18:14,143 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=81177.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:18:19,089 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 05:18:22,083 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 05:18:24,634 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=81198.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:18:37,950 INFO [train.py:901] (0/2) Epoch 29, batch 2150, loss[loss=0.1313, simple_loss=0.2108, pruned_loss=0.02593, over 7290.00 frames. ], tot_loss[loss=0.1377, simple_loss=0.2183, pruned_loss=0.02859, over 1440201.56 frames. ], batch size: 66, lr: 5.67e-03, grad_scale: 8.0 +2023-03-21 05:18:43,971 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.259e+02 1.793e+02 2.082e+02 2.494e+02 7.569e+02, threshold=4.163e+02, percent-clipped=1.0 +2023-03-21 05:18:59,493 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9309, 2.0958, 1.5337, 2.0055, 2.2042, 1.5337, 1.5565, 1.5649], + device='cuda:0'), covar=tensor([0.0132, 0.0140, 0.0274, 0.0118, 0.0082, 0.0206, 0.0142, 0.0222], + device='cuda:0'), in_proj_covar=tensor([0.0033, 0.0030, 0.0030, 0.0031, 0.0030, 0.0029, 0.0031, 0.0041], + device='cuda:0'), out_proj_covar=tensor([3.7562e-05, 3.4086e-05, 3.4035e-05, 3.4624e-05, 3.3706e-05, 3.2762e-05, + 3.5249e-05, 4.5964e-05], device='cuda:0') +2023-03-21 05:19:03,837 INFO [train.py:901] (0/2) Epoch 29, batch 2200, loss[loss=0.1493, simple_loss=0.2323, pruned_loss=0.03313, over 7264.00 frames. ], tot_loss[loss=0.1382, simple_loss=0.2187, pruned_loss=0.02889, over 1438390.74 frames. ], batch size: 64, lr: 5.67e-03, grad_scale: 8.0 +2023-03-21 05:19:09,344 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 05:19:16,944 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=81299.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:19:17,920 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=81301.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:19:27,911 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.4757, 4.9764, 4.8893, 5.4031, 5.3120, 5.3940, 4.8736, 4.9958], + device='cuda:0'), covar=tensor([0.0719, 0.2308, 0.1907, 0.0911, 0.0730, 0.0986, 0.0608, 0.1046], + device='cuda:0'), in_proj_covar=tensor([0.0242, 0.0376, 0.0283, 0.0294, 0.0215, 0.0353, 0.0214, 0.0261], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:19:29,324 INFO [train.py:901] (0/2) Epoch 29, batch 2250, loss[loss=0.1313, simple_loss=0.2182, pruned_loss=0.0222, over 7344.00 frames. ], tot_loss[loss=0.1385, simple_loss=0.2189, pruned_loss=0.02904, over 1437558.02 frames. ], batch size: 73, lr: 5.67e-03, grad_scale: 8.0 +2023-03-21 05:19:31,439 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6753, 5.1238, 5.2511, 5.1531, 4.9513, 4.5891, 5.2356, 5.0539], + device='cuda:0'), covar=tensor([0.0395, 0.0359, 0.0331, 0.0408, 0.0303, 0.0358, 0.0357, 0.0431], + device='cuda:0'), in_proj_covar=tensor([0.0210, 0.0245, 0.0184, 0.0190, 0.0151, 0.0220, 0.0197, 0.0143], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 05:19:33,086 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1352, 3.1970, 2.3685, 3.8175, 2.5961, 3.4984, 1.5549, 2.2614], + device='cuda:0'), covar=tensor([0.0361, 0.0637, 0.2376, 0.0502, 0.0339, 0.0532, 0.3407, 0.1885], + device='cuda:0'), in_proj_covar=tensor([0.0280, 0.0256, 0.0285, 0.0267, 0.0270, 0.0266, 0.0245, 0.0265], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 05:19:35,473 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.215e+02 1.800e+02 2.074e+02 2.534e+02 4.493e+02, threshold=4.147e+02, percent-clipped=2.0 +2023-03-21 05:19:44,671 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 05:19:45,195 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 05:19:48,839 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=81360.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:19:49,830 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=81362.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:19:55,265 INFO [train.py:901] (0/2) Epoch 29, batch 2300, loss[loss=0.1375, simple_loss=0.2217, pruned_loss=0.02662, over 7352.00 frames. ], tot_loss[loss=0.1381, simple_loss=0.2184, pruned_loss=0.02891, over 1438426.14 frames. ], batch size: 63, lr: 5.67e-03, grad_scale: 8.0 +2023-03-21 05:19:57,253 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 05:19:58,804 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6386, 3.9143, 3.7376, 3.8889, 3.6014, 3.9364, 4.2283, 4.2390], + device='cuda:0'), covar=tensor([0.0245, 0.0179, 0.0209, 0.0184, 0.0379, 0.0272, 0.0207, 0.0156], + device='cuda:0'), in_proj_covar=tensor([0.0112, 0.0115, 0.0106, 0.0112, 0.0104, 0.0092, 0.0090, 0.0089], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:20:04,859 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=81392.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:20:06,599 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.62 vs. limit=2.0 +2023-03-21 05:20:21,610 INFO [train.py:901] (0/2) Epoch 29, batch 2350, loss[loss=0.1383, simple_loss=0.2233, pruned_loss=0.02666, over 7300.00 frames. ], tot_loss[loss=0.1382, simple_loss=0.2186, pruned_loss=0.02894, over 1441275.85 frames. ], batch size: 80, lr: 5.67e-03, grad_scale: 8.0 +2023-03-21 05:20:24,698 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=81429.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:20:27,535 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.392e+02 1.900e+02 2.109e+02 2.477e+02 4.318e+02, threshold=4.218e+02, percent-clipped=1.0 +2023-03-21 05:20:30,775 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=81440.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:20:44,923 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 05:20:46,975 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=81472.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:20:47,402 INFO [train.py:901] (0/2) Epoch 29, batch 2400, loss[loss=0.1503, simple_loss=0.233, pruned_loss=0.03383, over 7334.00 frames. ], tot_loss[loss=0.1376, simple_loss=0.218, pruned_loss=0.02858, over 1441625.34 frames. ], batch size: 75, lr: 5.66e-03, grad_scale: 8.0 +2023-03-21 05:20:49,482 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=81477.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:20:51,391 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 05:20:54,134 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.67 vs. limit=2.0 +2023-03-21 05:20:57,624 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8931, 2.4163, 1.8287, 2.8441, 2.8124, 2.8356, 1.9520, 2.6758], + device='cuda:0'), covar=tensor([0.2413, 0.0998, 0.3803, 0.0745, 0.0240, 0.0220, 0.0298, 0.0405], + device='cuda:0'), in_proj_covar=tensor([0.0260, 0.0232, 0.0260, 0.0265, 0.0190, 0.0185, 0.0211, 0.0223], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 05:21:02,498 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 05:21:05,009 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 05:21:09,373 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.60 vs. limit=2.0 +2023-03-21 05:21:13,044 INFO [train.py:901] (0/2) Epoch 29, batch 2450, loss[loss=0.1477, simple_loss=0.2277, pruned_loss=0.03382, over 7279.00 frames. ], tot_loss[loss=0.1383, simple_loss=0.2186, pruned_loss=0.02902, over 1442940.66 frames. ], batch size: 70, lr: 5.66e-03, grad_scale: 8.0 +2023-03-21 05:21:19,661 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.288e+02 1.787e+02 2.100e+02 2.538e+02 7.882e+02, threshold=4.199e+02, percent-clipped=1.0 +2023-03-21 05:21:31,132 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 05:21:36,040 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.58 vs. limit=2.0 +2023-03-21 05:21:39,151 INFO [train.py:901] (0/2) Epoch 29, batch 2500, loss[loss=0.1292, simple_loss=0.213, pruned_loss=0.02266, over 7327.00 frames. ], tot_loss[loss=0.1384, simple_loss=0.2185, pruned_loss=0.02916, over 1442035.39 frames. ], batch size: 75, lr: 5.66e-03, grad_scale: 16.0 +2023-03-21 05:21:47,914 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9986, 2.9284, 2.3767, 4.0344, 1.8821, 3.7907, 1.4892, 3.2804], + device='cuda:0'), covar=tensor([0.0202, 0.1143, 0.1773, 0.0160, 0.4441, 0.0219, 0.1485, 0.0412], + device='cuda:0'), in_proj_covar=tensor([0.0199, 0.0257, 0.0267, 0.0200, 0.0258, 0.0210, 0.0241, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 05:21:57,455 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 05:21:58,580 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0528, 2.5719, 3.2145, 2.9718, 3.1157, 2.8714, 2.5145, 2.9226], + device='cuda:0'), covar=tensor([0.1270, 0.0745, 0.1022, 0.1197, 0.0826, 0.0979, 0.2022, 0.1464], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0063, 0.0048, 0.0048, 0.0046, 0.0045, 0.0065, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 05:22:05,110 INFO [train.py:901] (0/2) Epoch 29, batch 2550, loss[loss=0.1409, simple_loss=0.229, pruned_loss=0.02642, over 7259.00 frames. ], tot_loss[loss=0.1388, simple_loss=0.2192, pruned_loss=0.02926, over 1443817.45 frames. ], batch size: 55, lr: 5.66e-03, grad_scale: 16.0 +2023-03-21 05:22:11,195 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.193e+02 1.737e+02 2.185e+02 2.560e+02 4.324e+02, threshold=4.370e+02, percent-clipped=1.0 +2023-03-21 05:22:18,513 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.68 vs. limit=5.0 +2023-03-21 05:22:21,287 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=81655.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:22:22,877 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=81657.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:22:22,951 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6378, 3.0956, 2.6664, 2.8092, 2.9561, 2.4676, 2.9406, 2.9337], + device='cuda:0'), covar=tensor([0.0898, 0.0648, 0.1014, 0.1748, 0.0786, 0.0864, 0.0641, 0.0857], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0053, 0.0061, 0.0053, 0.0052, 0.0056, 0.0053, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:22:29,436 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=81670.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:22:30,800 INFO [train.py:901] (0/2) Epoch 29, batch 2600, loss[loss=0.1367, simple_loss=0.2238, pruned_loss=0.0248, over 7279.00 frames. ], tot_loss[loss=0.1388, simple_loss=0.2191, pruned_loss=0.02928, over 1442014.14 frames. ], batch size: 86, lr: 5.66e-03, grad_scale: 16.0 +2023-03-21 05:22:55,363 INFO [train.py:901] (0/2) Epoch 29, batch 2650, loss[loss=0.1679, simple_loss=0.2508, pruned_loss=0.04255, over 6683.00 frames. ], tot_loss[loss=0.1388, simple_loss=0.2193, pruned_loss=0.02911, over 1442201.96 frames. ], batch size: 106, lr: 5.66e-03, grad_scale: 8.0 +2023-03-21 05:22:58,432 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5054, 1.5428, 1.4396, 1.5985, 1.5142, 1.5032, 1.4624, 1.1149], + device='cuda:0'), covar=tensor([0.0108, 0.0155, 0.0209, 0.0116, 0.0108, 0.0110, 0.0149, 0.0150], + device='cuda:0'), in_proj_covar=tensor([0.0032, 0.0029, 0.0029, 0.0030, 0.0029, 0.0028, 0.0030, 0.0039], + device='cuda:0'), out_proj_covar=tensor([3.6572e-05, 3.2716e-05, 3.2983e-05, 3.3246e-05, 3.2765e-05, 3.1778e-05, + 3.4164e-05, 4.4221e-05], device='cuda:0') +2023-03-21 05:22:59,431 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=81731.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 05:23:01,692 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.150e+02 1.754e+02 2.071e+02 2.481e+02 5.009e+02, threshold=4.143e+02, percent-clipped=1.0 +2023-03-21 05:23:20,009 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=81772.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:23:20,406 INFO [train.py:901] (0/2) Epoch 29, batch 2700, loss[loss=0.1365, simple_loss=0.2247, pruned_loss=0.02415, over 7255.00 frames. ], tot_loss[loss=0.1388, simple_loss=0.2192, pruned_loss=0.02919, over 1441866.33 frames. ], batch size: 55, lr: 5.65e-03, grad_scale: 8.0 +2023-03-21 05:23:36,052 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5511, 1.4117, 1.9321, 2.2919, 2.0210, 2.0711, 2.0175, 2.0748], + device='cuda:0'), covar=tensor([0.3655, 0.3819, 0.2044, 0.1501, 0.1462, 0.3402, 0.1803, 0.3999], + device='cuda:0'), in_proj_covar=tensor([0.0069, 0.0071, 0.0060, 0.0054, 0.0057, 0.0057, 0.0088, 0.0058], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:23:43,933 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=81820.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:23:45,386 INFO [train.py:901] (0/2) Epoch 29, batch 2750, loss[loss=0.1486, simple_loss=0.2346, pruned_loss=0.03132, over 7353.00 frames. ], tot_loss[loss=0.1389, simple_loss=0.2195, pruned_loss=0.02912, over 1442046.23 frames. ], batch size: 73, lr: 5.65e-03, grad_scale: 8.0 +2023-03-21 05:23:51,726 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.223e+02 1.802e+02 2.176e+02 2.517e+02 3.886e+02, threshold=4.352e+02, percent-clipped=0.0 +2023-03-21 05:24:09,870 INFO [train.py:901] (0/2) Epoch 29, batch 2800, loss[loss=0.1468, simple_loss=0.2302, pruned_loss=0.03172, over 7317.00 frames. ], tot_loss[loss=0.1387, simple_loss=0.2194, pruned_loss=0.02904, over 1442483.70 frames. ], batch size: 83, lr: 5.65e-03, grad_scale: 8.0 +2023-03-21 05:24:22,362 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-29.pt +2023-03-21 05:24:40,407 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 05:24:40,462 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 05:24:40,494 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 05:24:40,513 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 05:24:40,691 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 05:24:40,707 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 05:24:40,717 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 05:24:40,823 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 05:24:41,252 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 05:24:41,269 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 05:24:41,285 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 05:24:41,450 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 05:24:41,456 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 05:24:41,463 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 05:24:41,610 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 05:24:41,612 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 05:24:41,672 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 05:24:43,921 INFO [train.py:901] (0/2) Epoch 30, batch 0, loss[loss=0.1451, simple_loss=0.227, pruned_loss=0.03161, over 7278.00 frames. ], tot_loss[loss=0.1451, simple_loss=0.227, pruned_loss=0.03161, over 7278.00 frames. ], batch size: 57, lr: 5.56e-03, grad_scale: 8.0 +2023-03-21 05:24:43,923 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 05:24:47,530 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6758, 2.4150, 2.7817, 2.6879, 2.6988, 2.5927, 2.7548, 2.2001], + device='cuda:0'), covar=tensor([0.0418, 0.0472, 0.0560, 0.0557, 0.0605, 0.0799, 0.0524, 0.1586], + device='cuda:0'), in_proj_covar=tensor([0.0341, 0.0344, 0.0274, 0.0362, 0.0301, 0.0301, 0.0349, 0.0271], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 05:24:56,182 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8111, 2.6805, 2.1648, 3.2240, 1.8998, 2.7368, 1.3861, 2.1308], + device='cuda:0'), covar=tensor([0.0383, 0.0587, 0.2570, 0.0615, 0.0348, 0.0464, 0.3371, 0.1478], + device='cuda:0'), in_proj_covar=tensor([0.0280, 0.0255, 0.0287, 0.0268, 0.0271, 0.0266, 0.0244, 0.0265], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 05:25:09,058 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4664, 4.6960, 4.7672, 4.6929, 4.5370, 4.2683, 4.7568, 4.4828], + device='cuda:0'), covar=tensor([0.0351, 0.0360, 0.0333, 0.0475, 0.0336, 0.0287, 0.0288, 0.0470], + device='cuda:0'), in_proj_covar=tensor([0.0208, 0.0239, 0.0182, 0.0187, 0.0149, 0.0217, 0.0193, 0.0142], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 05:25:10,055 INFO [train.py:935] (0/2) Epoch 30, validation: loss=0.164, simple_loss=0.254, pruned_loss=0.03703, over 1622729.00 frames. +2023-03-21 05:25:10,055 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 05:25:10,207 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5593, 1.2459, 1.7512, 2.1032, 1.6652, 1.9006, 1.5949, 1.9460], + device='cuda:0'), covar=tensor([0.1725, 0.4577, 0.1196, 0.0832, 0.1602, 0.2119, 0.2261, 0.2315], + device='cuda:0'), in_proj_covar=tensor([0.0069, 0.0070, 0.0060, 0.0054, 0.0057, 0.0057, 0.0088, 0.0058], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:25:16,641 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 05:25:21,754 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4005, 1.5370, 1.2729, 1.6366, 1.7106, 1.5878, 1.4254, 1.1233], + device='cuda:0'), covar=tensor([0.0145, 0.0125, 0.0317, 0.0132, 0.0096, 0.0096, 0.0129, 0.0163], + device='cuda:0'), in_proj_covar=tensor([0.0032, 0.0029, 0.0029, 0.0030, 0.0029, 0.0028, 0.0030, 0.0039], + device='cuda:0'), out_proj_covar=tensor([3.6494e-05, 3.2431e-05, 3.3006e-05, 3.3378e-05, 3.2417e-05, 3.1467e-05, + 3.4030e-05, 4.4000e-05], device='cuda:0') +2023-03-21 05:25:28,172 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 05:25:29,653 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.214e+02 1.769e+02 2.116e+02 2.355e+02 5.614e+02, threshold=4.232e+02, percent-clipped=2.0 +2023-03-21 05:25:35,184 INFO [train.py:901] (0/2) Epoch 30, batch 50, loss[loss=0.1396, simple_loss=0.2156, pruned_loss=0.03179, over 7263.00 frames. ], tot_loss[loss=0.1381, simple_loss=0.2176, pruned_loss=0.02924, over 328019.97 frames. ], batch size: 47, lr: 5.55e-03, grad_scale: 8.0 +2023-03-21 05:25:35,209 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 05:25:37,231 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 05:25:39,319 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=81955.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:25:40,279 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 05:25:40,365 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=81957.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:25:53,229 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0920, 2.1598, 2.3701, 1.9067, 2.2664, 2.2679, 1.9076, 1.8252], + device='cuda:0'), covar=tensor([0.0401, 0.0311, 0.0210, 0.0222, 0.0425, 0.0329, 0.0276, 0.0268], + device='cuda:0'), in_proj_covar=tensor([0.0036, 0.0033, 0.0033, 0.0033, 0.0032, 0.0031, 0.0036, 0.0036], + device='cuda:0'), out_proj_covar=tensor([8.8839e-05, 8.6673e-05, 8.3983e-05, 8.3479e-05, 8.3322e-05, 8.1245e-05, + 8.8851e-05, 9.0965e-05], device='cuda:0') +2023-03-21 05:25:56,324 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=81987.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:25:57,225 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 05:25:57,735 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 05:26:01,878 INFO [train.py:901] (0/2) Epoch 30, batch 100, loss[loss=0.1375, simple_loss=0.22, pruned_loss=0.02748, over 7290.00 frames. ], tot_loss[loss=0.1385, simple_loss=0.2193, pruned_loss=0.02888, over 573236.46 frames. ], batch size: 66, lr: 5.55e-03, grad_scale: 8.0 +2023-03-21 05:26:05,355 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=82003.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:26:06,393 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=82005.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:26:16,968 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=82026.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 05:26:21,838 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.292e+02 1.747e+02 2.055e+02 2.472e+02 4.255e+02, threshold=4.110e+02, percent-clipped=1.0 +2023-03-21 05:26:27,976 INFO [train.py:901] (0/2) Epoch 30, batch 150, loss[loss=0.1465, simple_loss=0.2315, pruned_loss=0.0308, over 7266.00 frames. ], tot_loss[loss=0.1389, simple_loss=0.2193, pruned_loss=0.02923, over 763373.28 frames. ], batch size: 52, lr: 5.55e-03, grad_scale: 8.0 +2023-03-21 05:26:28,635 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=82048.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:26:32,196 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=82055.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:26:33,895 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.72 vs. limit=2.0 +2023-03-21 05:26:44,757 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3114, 3.9532, 3.9901, 3.9314, 3.8179, 3.8374, 4.1222, 3.7255], + device='cuda:0'), covar=tensor([0.0097, 0.0144, 0.0103, 0.0148, 0.0408, 0.0113, 0.0131, 0.0153], + device='cuda:0'), in_proj_covar=tensor([0.0088, 0.0096, 0.0094, 0.0084, 0.0165, 0.0101, 0.0101, 0.0103], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:26:53,598 INFO [train.py:901] (0/2) Epoch 30, batch 200, loss[loss=0.125, simple_loss=0.2112, pruned_loss=0.01941, over 7279.00 frames. ], tot_loss[loss=0.1379, simple_loss=0.2184, pruned_loss=0.02866, over 915635.42 frames. ], batch size: 70, lr: 5.55e-03, grad_scale: 8.0 +2023-03-21 05:26:58,774 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 05:27:03,289 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 05:27:03,429 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=82116.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:27:09,250 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 05:27:13,829 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.287e+02 1.803e+02 2.124e+02 2.586e+02 4.041e+02, threshold=4.248e+02, percent-clipped=0.0 +2023-03-21 05:27:19,427 INFO [train.py:901] (0/2) Epoch 30, batch 250, loss[loss=0.1407, simple_loss=0.2131, pruned_loss=0.03416, over 7255.00 frames. ], tot_loss[loss=0.1382, simple_loss=0.2184, pruned_loss=0.02895, over 1029181.28 frames. ], batch size: 47, lr: 5.55e-03, grad_scale: 8.0 +2023-03-21 05:27:22,281 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 05:27:42,125 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 05:27:43,268 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4786, 4.1152, 4.4643, 4.4713, 4.5673, 4.5648, 4.5718, 4.4287], + device='cuda:0'), covar=tensor([0.0023, 0.0067, 0.0025, 0.0032, 0.0021, 0.0026, 0.0021, 0.0037], + device='cuda:0'), in_proj_covar=tensor([0.0046, 0.0065, 0.0054, 0.0054, 0.0052, 0.0057, 0.0048, 0.0072], + device='cuda:0'), out_proj_covar=tensor([8.4278e-05, 1.3923e-04, 1.0566e-04, 9.9011e-05, 9.4114e-05, 1.0380e-04, + 9.9741e-05, 1.3989e-04], device='cuda:0') +2023-03-21 05:27:45,161 INFO [train.py:901] (0/2) Epoch 30, batch 300, loss[loss=0.131, simple_loss=0.2081, pruned_loss=0.02695, over 7331.00 frames. ], tot_loss[loss=0.1385, simple_loss=0.2186, pruned_loss=0.02917, over 1122328.91 frames. ], batch size: 75, lr: 5.55e-03, grad_scale: 8.0 +2023-03-21 05:27:51,586 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 05:27:58,512 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.67 vs. limit=5.0 +2023-03-21 05:28:05,670 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.315e+02 1.824e+02 2.205e+02 2.616e+02 4.930e+02, threshold=4.410e+02, percent-clipped=1.0 +2023-03-21 05:28:11,137 INFO [train.py:901] (0/2) Epoch 30, batch 350, loss[loss=0.1399, simple_loss=0.2179, pruned_loss=0.03094, over 7355.00 frames. ], tot_loss[loss=0.1384, simple_loss=0.2184, pruned_loss=0.02914, over 1191928.25 frames. ], batch size: 73, lr: 5.54e-03, grad_scale: 8.0 +2023-03-21 05:28:25,920 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 05:28:27,591 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8934, 2.3494, 1.7507, 2.6052, 2.5472, 2.8190, 2.2984, 2.2981], + device='cuda:0'), covar=tensor([0.2267, 0.1038, 0.3832, 0.1004, 0.0288, 0.0306, 0.0332, 0.0381], + device='cuda:0'), in_proj_covar=tensor([0.0261, 0.0232, 0.0259, 0.0266, 0.0190, 0.0186, 0.0210, 0.0225], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 05:28:36,970 INFO [train.py:901] (0/2) Epoch 30, batch 400, loss[loss=0.1296, simple_loss=0.2143, pruned_loss=0.02242, over 7256.00 frames. ], tot_loss[loss=0.1377, simple_loss=0.2181, pruned_loss=0.02863, over 1248699.06 frames. ], batch size: 64, lr: 5.54e-03, grad_scale: 8.0 +2023-03-21 05:28:39,643 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0693, 3.3629, 2.8802, 2.9923, 3.3538, 2.7500, 3.3266, 3.0822], + device='cuda:0'), covar=tensor([0.1084, 0.1123, 0.0783, 0.1973, 0.1374, 0.0652, 0.0461, 0.0768], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0053, 0.0061, 0.0053, 0.0051, 0.0055, 0.0052, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:28:44,966 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5047, 3.3558, 2.3332, 3.7828, 2.8389, 3.4877, 1.5157, 2.2647], + device='cuda:0'), covar=tensor([0.0441, 0.0692, 0.2417, 0.0429, 0.0411, 0.0676, 0.3337, 0.1907], + device='cuda:0'), in_proj_covar=tensor([0.0279, 0.0255, 0.0286, 0.0268, 0.0270, 0.0266, 0.0244, 0.0266], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 05:28:52,520 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=82326.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:28:57,396 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.199e+02 1.768e+02 2.131e+02 2.450e+02 4.318e+02, threshold=4.262e+02, percent-clipped=0.0 +2023-03-21 05:29:01,613 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=82343.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:29:03,229 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=82346.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:29:03,600 INFO [train.py:901] (0/2) Epoch 30, batch 450, loss[loss=0.1144, simple_loss=0.1971, pruned_loss=0.01582, over 7136.00 frames. ], tot_loss[loss=0.1372, simple_loss=0.2179, pruned_loss=0.02828, over 1291600.64 frames. ], batch size: 41, lr: 5.54e-03, grad_scale: 8.0 +2023-03-21 05:29:08,614 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 05:29:08,627 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 05:29:15,742 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6097, 4.0875, 3.9680, 4.5819, 4.3840, 4.4381, 3.8884, 4.0715], + device='cuda:0'), covar=tensor([0.0937, 0.2876, 0.2652, 0.1168, 0.1003, 0.1498, 0.0968, 0.1118], + device='cuda:0'), in_proj_covar=tensor([0.0244, 0.0377, 0.0287, 0.0295, 0.0215, 0.0354, 0.0214, 0.0261], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:29:15,820 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5436, 2.6455, 2.4089, 2.6618, 2.6914, 2.3315, 2.7061, 2.5434], + device='cuda:0'), covar=tensor([0.0662, 0.0849, 0.0997, 0.1287, 0.0795, 0.0453, 0.0525, 0.0829], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0054, 0.0062, 0.0054, 0.0052, 0.0056, 0.0053, 0.0050], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:29:17,258 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=82374.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:29:29,393 INFO [train.py:901] (0/2) Epoch 30, batch 500, loss[loss=0.1269, simple_loss=0.2014, pruned_loss=0.02626, over 7271.00 frames. ], tot_loss[loss=0.1373, simple_loss=0.2183, pruned_loss=0.02821, over 1326082.88 frames. ], batch size: 47, lr: 5.54e-03, grad_scale: 8.0 +2023-03-21 05:29:34,969 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=82407.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:29:36,949 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=82411.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:29:40,379 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 05:29:41,911 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 05:29:42,962 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 05:29:45,534 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 05:29:50,047 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.309e+02 1.731e+02 2.051e+02 2.399e+02 6.852e+02, threshold=4.103e+02, percent-clipped=1.0 +2023-03-21 05:29:50,068 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 05:29:53,767 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1791, 3.1366, 2.1158, 3.6351, 2.5202, 3.0117, 1.3924, 2.1376], + device='cuda:0'), covar=tensor([0.0434, 0.0730, 0.2483, 0.0501, 0.0401, 0.0588, 0.3485, 0.1878], + device='cuda:0'), in_proj_covar=tensor([0.0279, 0.0255, 0.0285, 0.0269, 0.0269, 0.0266, 0.0243, 0.0266], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 05:29:55,577 INFO [train.py:901] (0/2) Epoch 30, batch 550, loss[loss=0.1334, simple_loss=0.22, pruned_loss=0.02342, over 7244.00 frames. ], tot_loss[loss=0.1376, simple_loss=0.2185, pruned_loss=0.02829, over 1355089.63 frames. ], batch size: 89, lr: 5.54e-03, grad_scale: 8.0 +2023-03-21 05:30:00,948 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 05:30:09,626 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 05:30:13,071 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 05:30:16,631 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.76 vs. limit=5.0 +2023-03-21 05:30:19,380 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=82493.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 05:30:19,738 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 05:30:21,213 INFO [train.py:901] (0/2) Epoch 30, batch 600, loss[loss=0.1294, simple_loss=0.2163, pruned_loss=0.02127, over 7303.00 frames. ], tot_loss[loss=0.1375, simple_loss=0.2184, pruned_loss=0.02828, over 1373066.67 frames. ], batch size: 77, lr: 5.54e-03, grad_scale: 8.0 +2023-03-21 05:30:36,849 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 05:30:41,268 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.247e+02 1.785e+02 2.155e+02 2.633e+02 4.871e+02, threshold=4.309e+02, percent-clipped=2.0 +2023-03-21 05:30:45,824 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 05:30:46,469 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9011, 2.5229, 3.1020, 2.9835, 3.0854, 2.6864, 2.7186, 3.0940], + device='cuda:0'), covar=tensor([0.1659, 0.0889, 0.1087, 0.1305, 0.0965, 0.1396, 0.1740, 0.1098], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0063, 0.0049, 0.0048, 0.0047, 0.0047, 0.0066, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 05:30:46,856 INFO [train.py:901] (0/2) Epoch 30, batch 650, loss[loss=0.1443, simple_loss=0.2311, pruned_loss=0.02875, over 7354.00 frames. ], tot_loss[loss=0.1377, simple_loss=0.2186, pruned_loss=0.02842, over 1390272.47 frames. ], batch size: 63, lr: 5.53e-03, grad_scale: 8.0 +2023-03-21 05:30:50,554 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=82554.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 05:30:58,221 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=82569.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:31:03,384 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 05:31:12,407 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 05:31:12,903 INFO [train.py:901] (0/2) Epoch 30, batch 700, loss[loss=0.1331, simple_loss=0.2124, pruned_loss=0.02687, over 7348.00 frames. ], tot_loss[loss=0.1378, simple_loss=0.2186, pruned_loss=0.02851, over 1400316.92 frames. ], batch size: 61, lr: 5.53e-03, grad_scale: 8.0 +2023-03-21 05:31:16,497 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3126, 1.6304, 1.3976, 1.5879, 1.7476, 1.4756, 1.5489, 1.2085], + device='cuda:0'), covar=tensor([0.0145, 0.0162, 0.0229, 0.0185, 0.0129, 0.0141, 0.0148, 0.0216], + device='cuda:0'), in_proj_covar=tensor([0.0033, 0.0030, 0.0030, 0.0031, 0.0031, 0.0030, 0.0032, 0.0040], + device='cuda:0'), out_proj_covar=tensor([3.8294e-05, 3.4308e-05, 3.4454e-05, 3.5244e-05, 3.4642e-05, 3.3379e-05, + 3.5996e-05, 4.5428e-05], device='cuda:0') +2023-03-21 05:31:19,563 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6141, 1.3086, 1.6668, 2.0564, 1.7593, 2.0017, 1.4916, 1.8926], + device='cuda:0'), covar=tensor([0.2254, 0.3925, 0.2011, 0.1543, 0.1650, 0.2576, 0.2203, 0.2366], + device='cuda:0'), in_proj_covar=tensor([0.0070, 0.0072, 0.0060, 0.0055, 0.0057, 0.0058, 0.0089, 0.0058], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:31:22,433 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.69 vs. limit=2.0 +2023-03-21 05:31:29,347 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.86 vs. limit=2.0 +2023-03-21 05:31:30,687 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=82630.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:31:33,642 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.368e+02 1.896e+02 2.225e+02 2.541e+02 4.418e+02, threshold=4.450e+02, percent-clipped=1.0 +2023-03-21 05:31:36,190 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 05:31:36,660 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 05:31:37,278 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=82643.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:31:39,219 INFO [train.py:901] (0/2) Epoch 30, batch 750, loss[loss=0.1334, simple_loss=0.2217, pruned_loss=0.02251, over 7287.00 frames. ], tot_loss[loss=0.1371, simple_loss=0.218, pruned_loss=0.02813, over 1408568.94 frames. ], batch size: 70, lr: 5.53e-03, grad_scale: 8.0 +2023-03-21 05:31:51,964 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 05:31:52,624 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3052, 2.8345, 3.0994, 3.2186, 2.7904, 2.6956, 3.0768, 2.3626], + device='cuda:0'), covar=tensor([0.0568, 0.0534, 0.0641, 0.0680, 0.0597, 0.0945, 0.0690, 0.1847], + device='cuda:0'), in_proj_covar=tensor([0.0340, 0.0343, 0.0276, 0.0361, 0.0300, 0.0302, 0.0351, 0.0271], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 05:31:56,409 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 05:32:02,519 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=82691.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:32:02,981 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 05:32:04,552 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 05:32:05,528 INFO [train.py:901] (0/2) Epoch 30, batch 800, loss[loss=0.1467, simple_loss=0.2317, pruned_loss=0.03084, over 7348.00 frames. ], tot_loss[loss=0.1371, simple_loss=0.2179, pruned_loss=0.02817, over 1417966.60 frames. ], batch size: 73, lr: 5.53e-03, grad_scale: 8.0 +2023-03-21 05:32:08,154 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=82702.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:32:12,710 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=82711.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:32:15,147 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 05:32:15,751 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0595, 3.6260, 4.1160, 4.1185, 4.1078, 4.1569, 4.2318, 4.0704], + device='cuda:0'), covar=tensor([0.0034, 0.0099, 0.0034, 0.0033, 0.0031, 0.0030, 0.0031, 0.0054], + device='cuda:0'), in_proj_covar=tensor([0.0046, 0.0066, 0.0055, 0.0053, 0.0052, 0.0057, 0.0048, 0.0072], + device='cuda:0'), out_proj_covar=tensor([8.3912e-05, 1.4150e-04, 1.0641e-04, 9.8544e-05, 9.3791e-05, 1.0430e-04, + 1.0032e-04, 1.4050e-04], device='cuda:0') +2023-03-21 05:32:25,199 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.302e+02 1.846e+02 2.089e+02 2.458e+02 4.934e+02, threshold=4.178e+02, percent-clipped=1.0 +2023-03-21 05:32:31,375 INFO [train.py:901] (0/2) Epoch 30, batch 850, loss[loss=0.1372, simple_loss=0.221, pruned_loss=0.02673, over 7286.00 frames. ], tot_loss[loss=0.1374, simple_loss=0.2181, pruned_loss=0.0283, over 1422600.25 frames. ], batch size: 77, lr: 5.53e-03, grad_scale: 8.0 +2023-03-21 05:32:33,873 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 05:32:34,352 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 05:32:37,402 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=82759.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:32:39,926 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 05:32:43,559 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 05:32:57,126 INFO [train.py:901] (0/2) Epoch 30, batch 900, loss[loss=0.1446, simple_loss=0.2242, pruned_loss=0.03253, over 7293.00 frames. ], tot_loss[loss=0.1371, simple_loss=0.2177, pruned_loss=0.02827, over 1427296.61 frames. ], batch size: 68, lr: 5.53e-03, grad_scale: 8.0 +2023-03-21 05:33:07,228 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8157, 3.3290, 3.3869, 3.4564, 3.0354, 2.9710, 3.6412, 2.7345], + device='cuda:0'), covar=tensor([0.0376, 0.0497, 0.0598, 0.0618, 0.0783, 0.1053, 0.0584, 0.1821], + device='cuda:0'), in_proj_covar=tensor([0.0334, 0.0338, 0.0271, 0.0354, 0.0294, 0.0297, 0.0345, 0.0266], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 05:33:17,779 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.226e+02 1.775e+02 2.067e+02 2.438e+02 4.554e+02, threshold=4.134e+02, percent-clipped=2.0 +2023-03-21 05:33:23,152 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.30 vs. limit=5.0 +2023-03-21 05:33:23,325 INFO [train.py:901] (0/2) Epoch 30, batch 950, loss[loss=0.113, simple_loss=0.1863, pruned_loss=0.01983, over 6958.00 frames. ], tot_loss[loss=0.1374, simple_loss=0.2179, pruned_loss=0.02844, over 1431908.15 frames. ], batch size: 35, lr: 5.52e-03, grad_scale: 8.0 +2023-03-21 05:33:23,346 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 05:33:24,402 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=82849.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 05:33:32,030 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.71 vs. limit=5.0 +2023-03-21 05:33:46,290 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 05:33:49,367 INFO [train.py:901] (0/2) Epoch 30, batch 1000, loss[loss=0.1373, simple_loss=0.2203, pruned_loss=0.02709, over 7288.00 frames. ], tot_loss[loss=0.137, simple_loss=0.2177, pruned_loss=0.02819, over 1436062.01 frames. ], batch size: 66, lr: 5.52e-03, grad_scale: 8.0 +2023-03-21 05:34:04,177 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=82925.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:34:06,639 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 05:34:09,601 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.374e+02 1.830e+02 2.186e+02 2.537e+02 4.453e+02, threshold=4.372e+02, percent-clipped=1.0 +2023-03-21 05:34:15,837 INFO [train.py:901] (0/2) Epoch 30, batch 1050, loss[loss=0.1436, simple_loss=0.2244, pruned_loss=0.03145, over 7279.00 frames. ], tot_loss[loss=0.137, simple_loss=0.218, pruned_loss=0.02803, over 1438485.69 frames. ], batch size: 66, lr: 5.52e-03, grad_scale: 8.0 +2023-03-21 05:34:24,075 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=82963.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:34:28,457 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 05:34:32,479 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 05:34:41,033 INFO [train.py:901] (0/2) Epoch 30, batch 1100, loss[loss=0.1354, simple_loss=0.2238, pruned_loss=0.02354, over 7332.00 frames. ], tot_loss[loss=0.1372, simple_loss=0.218, pruned_loss=0.02817, over 1439024.73 frames. ], batch size: 75, lr: 5.52e-03, grad_scale: 8.0 +2023-03-21 05:34:44,028 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=83002.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:34:55,726 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=83024.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:35:01,607 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.202e+02 1.822e+02 2.194e+02 2.520e+02 4.082e+02, threshold=4.387e+02, percent-clipped=0.0 +2023-03-21 05:35:03,294 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 05:35:03,793 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 05:35:07,768 INFO [train.py:901] (0/2) Epoch 30, batch 1150, loss[loss=0.1417, simple_loss=0.2233, pruned_loss=0.03004, over 7320.00 frames. ], tot_loss[loss=0.1366, simple_loss=0.217, pruned_loss=0.02814, over 1438496.44 frames. ], batch size: 83, lr: 5.52e-03, grad_scale: 8.0 +2023-03-21 05:35:09,340 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=83050.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:35:16,352 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 05:35:16,861 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 05:35:17,612 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.63 vs. limit=2.0 +2023-03-21 05:35:31,793 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6656, 3.5627, 2.5643, 3.6301, 3.4371, 3.5819, 3.3139, 3.0647], + device='cuda:0'), covar=tensor([0.1891, 0.0642, 0.3301, 0.0655, 0.0296, 0.0210, 0.0365, 0.0326], + device='cuda:0'), in_proj_covar=tensor([0.0257, 0.0232, 0.0258, 0.0262, 0.0191, 0.0184, 0.0207, 0.0223], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 05:35:32,291 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5619, 2.6937, 2.4218, 2.6220, 2.6713, 2.3917, 2.7303, 2.5994], + device='cuda:0'), covar=tensor([0.1105, 0.0887, 0.1187, 0.1105, 0.1308, 0.0932, 0.0944, 0.1059], + device='cuda:0'), in_proj_covar=tensor([0.0056, 0.0056, 0.0063, 0.0055, 0.0053, 0.0057, 0.0055, 0.0050], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:35:33,781 INFO [train.py:901] (0/2) Epoch 30, batch 1200, loss[loss=0.1271, simple_loss=0.2143, pruned_loss=0.01994, over 7302.00 frames. ], tot_loss[loss=0.137, simple_loss=0.2171, pruned_loss=0.0284, over 1437135.74 frames. ], batch size: 80, lr: 5.52e-03, grad_scale: 8.0 +2023-03-21 05:35:49,652 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 05:35:54,133 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.348e+02 1.789e+02 2.118e+02 2.602e+02 4.588e+02, threshold=4.235e+02, percent-clipped=1.0 +2023-03-21 05:35:59,542 INFO [train.py:901] (0/2) Epoch 30, batch 1250, loss[loss=0.1449, simple_loss=0.2253, pruned_loss=0.03222, over 7294.00 frames. ], tot_loss[loss=0.1373, simple_loss=0.2172, pruned_loss=0.02864, over 1435967.54 frames. ], batch size: 68, lr: 5.51e-03, grad_scale: 8.0 +2023-03-21 05:36:00,638 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=83149.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 05:36:13,165 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 05:36:17,152 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 05:36:17,236 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7277, 3.4217, 3.4243, 3.3785, 3.4196, 3.2462, 3.6253, 3.2992], + device='cuda:0'), covar=tensor([0.0117, 0.0191, 0.0138, 0.0197, 0.0394, 0.0117, 0.0143, 0.0162], + device='cuda:0'), in_proj_covar=tensor([0.0087, 0.0096, 0.0094, 0.0084, 0.0166, 0.0102, 0.0099, 0.0104], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:36:19,316 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 05:36:25,055 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=83196.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:36:25,416 INFO [train.py:901] (0/2) Epoch 30, batch 1300, loss[loss=0.1118, simple_loss=0.1777, pruned_loss=0.02299, over 6297.00 frames. ], tot_loss[loss=0.1364, simple_loss=0.2165, pruned_loss=0.02813, over 1435302.76 frames. ], batch size: 27, lr: 5.51e-03, grad_scale: 8.0 +2023-03-21 05:36:25,471 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=83197.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 05:36:26,574 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9474, 2.5361, 3.2826, 3.0886, 3.0302, 3.0451, 2.5476, 3.1175], + device='cuda:0'), covar=tensor([0.1145, 0.0794, 0.1062, 0.1134, 0.0977, 0.0870, 0.1996, 0.1182], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0062, 0.0048, 0.0048, 0.0046, 0.0046, 0.0066, 0.0048], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 05:36:29,268 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-03-21 05:36:40,687 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=83225.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:36:43,089 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 05:36:45,089 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 05:36:46,076 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.311e+02 1.913e+02 2.268e+02 2.716e+02 5.038e+02, threshold=4.536e+02, percent-clipped=3.0 +2023-03-21 05:36:48,621 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 05:36:51,611 INFO [train.py:901] (0/2) Epoch 30, batch 1350, loss[loss=0.1362, simple_loss=0.213, pruned_loss=0.02965, over 7292.00 frames. ], tot_loss[loss=0.1371, simple_loss=0.217, pruned_loss=0.02859, over 1436025.68 frames. ], batch size: 57, lr: 5.51e-03, grad_scale: 8.0 +2023-03-21 05:36:56,868 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=83257.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:36:59,309 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 05:37:04,837 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=83273.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:37:10,992 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=83284.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:37:17,376 INFO [train.py:901] (0/2) Epoch 30, batch 1400, loss[loss=0.1273, simple_loss=0.213, pruned_loss=0.02086, over 7252.00 frames. ], tot_loss[loss=0.1371, simple_loss=0.217, pruned_loss=0.02861, over 1438375.93 frames. ], batch size: 64, lr: 5.51e-03, grad_scale: 8.0 +2023-03-21 05:37:28,940 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=83319.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:37:32,360 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 05:37:37,510 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.214e+02 1.867e+02 2.201e+02 2.472e+02 4.157e+02, threshold=4.401e+02, percent-clipped=0.0 +2023-03-21 05:37:42,162 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=83345.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:37:43,002 INFO [train.py:901] (0/2) Epoch 30, batch 1450, loss[loss=0.1389, simple_loss=0.2129, pruned_loss=0.03252, over 7220.00 frames. ], tot_loss[loss=0.1367, simple_loss=0.2166, pruned_loss=0.02838, over 1439275.27 frames. ], batch size: 50, lr: 5.51e-03, grad_scale: 8.0 +2023-03-21 05:37:57,158 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 05:38:09,521 INFO [train.py:901] (0/2) Epoch 30, batch 1500, loss[loss=0.1372, simple_loss=0.2213, pruned_loss=0.02657, over 7296.00 frames. ], tot_loss[loss=0.1368, simple_loss=0.2169, pruned_loss=0.02833, over 1438178.38 frames. ], batch size: 68, lr: 5.51e-03, grad_scale: 8.0 +2023-03-21 05:38:14,475 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 05:38:16,076 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8668, 2.4193, 3.2528, 2.9072, 3.0681, 2.6880, 2.4435, 3.1105], + device='cuda:0'), covar=tensor([0.1294, 0.0965, 0.0719, 0.1571, 0.0892, 0.1487, 0.2471, 0.1088], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0062, 0.0048, 0.0047, 0.0046, 0.0045, 0.0065, 0.0048], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 05:38:29,487 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.259e+02 1.731e+02 2.059e+02 2.508e+02 4.024e+02, threshold=4.119e+02, percent-clipped=0.0 +2023-03-21 05:38:34,980 INFO [train.py:901] (0/2) Epoch 30, batch 1550, loss[loss=0.1261, simple_loss=0.2057, pruned_loss=0.0233, over 7318.00 frames. ], tot_loss[loss=0.1369, simple_loss=0.2173, pruned_loss=0.02825, over 1440108.67 frames. ], batch size: 59, lr: 5.50e-03, grad_scale: 8.0 +2023-03-21 05:38:39,084 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 05:38:44,836 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.75 vs. limit=2.0 +2023-03-21 05:38:56,889 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6485, 3.0090, 2.5952, 2.9875, 2.8825, 2.5289, 2.8577, 2.6971], + device='cuda:0'), covar=tensor([0.0878, 0.0421, 0.0958, 0.0669, 0.1198, 0.0651, 0.1266, 0.1097], + device='cuda:0'), in_proj_covar=tensor([0.0056, 0.0055, 0.0063, 0.0055, 0.0054, 0.0056, 0.0054, 0.0050], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:39:01,219 INFO [train.py:901] (0/2) Epoch 30, batch 1600, loss[loss=0.137, simple_loss=0.2159, pruned_loss=0.02898, over 7332.00 frames. ], tot_loss[loss=0.1372, simple_loss=0.2176, pruned_loss=0.02844, over 1438287.49 frames. ], batch size: 59, lr: 5.50e-03, grad_scale: 8.0 +2023-03-21 05:39:10,329 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 05:39:10,794 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 05:39:13,676 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 05:39:14,811 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9838, 2.1469, 2.1698, 2.0710, 2.3873, 2.0389, 1.8827, 1.8862], + device='cuda:0'), covar=tensor([0.0554, 0.0454, 0.0485, 0.0273, 0.0237, 0.0741, 0.0427, 0.0285], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0034, 0.0034, 0.0033, 0.0032, 0.0032, 0.0036, 0.0036], + device='cuda:0'), out_proj_covar=tensor([8.8459e-05, 8.8127e-05, 8.5525e-05, 8.3709e-05, 8.3550e-05, 8.3051e-05, + 8.9875e-05, 9.1127e-05], device='cuda:0') +2023-03-21 05:39:18,906 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4494, 2.9578, 3.3303, 3.1575, 2.8909, 2.7217, 3.5437, 2.5604], + device='cuda:0'), covar=tensor([0.0436, 0.0452, 0.0568, 0.0639, 0.0675, 0.0853, 0.0603, 0.1712], + device='cuda:0'), in_proj_covar=tensor([0.0334, 0.0341, 0.0275, 0.0361, 0.0298, 0.0296, 0.0345, 0.0267], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 05:39:20,728 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.139e+02 1.801e+02 2.093e+02 2.671e+02 5.073e+02, threshold=4.186e+02, percent-clipped=2.0 +2023-03-21 05:39:23,425 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 05:39:26,576 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7754, 2.8892, 2.5948, 2.9371, 3.0175, 2.5017, 2.9984, 2.8377], + device='cuda:0'), covar=tensor([0.0763, 0.0652, 0.1236, 0.1109, 0.0813, 0.0801, 0.0842, 0.0876], + device='cuda:0'), in_proj_covar=tensor([0.0056, 0.0055, 0.0063, 0.0055, 0.0054, 0.0056, 0.0054, 0.0050], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:39:26,956 INFO [train.py:901] (0/2) Epoch 30, batch 1650, loss[loss=0.1362, simple_loss=0.2177, pruned_loss=0.02732, over 7252.00 frames. ], tot_loss[loss=0.1373, simple_loss=0.2175, pruned_loss=0.0286, over 1439265.33 frames. ], batch size: 64, lr: 5.50e-03, grad_scale: 8.0 +2023-03-21 05:39:28,294 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.84 vs. limit=5.0 +2023-03-21 05:39:28,498 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 05:39:29,609 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=83552.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:39:36,571 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 05:39:52,622 INFO [train.py:901] (0/2) Epoch 30, batch 1700, loss[loss=0.1337, simple_loss=0.2173, pruned_loss=0.02505, over 7280.00 frames. ], tot_loss[loss=0.1369, simple_loss=0.2168, pruned_loss=0.02848, over 1438325.09 frames. ], batch size: 66, lr: 5.50e-03, grad_scale: 8.0 +2023-03-21 05:39:53,024 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.10 vs. limit=5.0 +2023-03-21 05:39:54,508 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 05:39:54,639 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7158, 2.9940, 2.6779, 3.0484, 3.0131, 2.5714, 2.9424, 2.7112], + device='cuda:0'), covar=tensor([0.0941, 0.0913, 0.1166, 0.0853, 0.0991, 0.0544, 0.1583, 0.1714], + device='cuda:0'), in_proj_covar=tensor([0.0056, 0.0055, 0.0062, 0.0055, 0.0054, 0.0056, 0.0054, 0.0050], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:39:58,551 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 05:40:04,278 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=83619.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:40:04,912 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.40 vs. limit=2.0 +2023-03-21 05:40:09,678 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 05:40:09,759 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5012, 3.6521, 3.5491, 3.6553, 3.3399, 3.5582, 3.9029, 3.9411], + device='cuda:0'), covar=tensor([0.0240, 0.0176, 0.0217, 0.0173, 0.0361, 0.0338, 0.0234, 0.0182], + device='cuda:0'), in_proj_covar=tensor([0.0113, 0.0118, 0.0109, 0.0115, 0.0106, 0.0094, 0.0092, 0.0090], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:40:13,349 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.202e+02 1.754e+02 2.177e+02 2.446e+02 3.446e+02, threshold=4.353e+02, percent-clipped=0.0 +2023-03-21 05:40:15,440 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=83640.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:40:18,871 INFO [train.py:901] (0/2) Epoch 30, batch 1750, loss[loss=0.1444, simple_loss=0.2232, pruned_loss=0.03279, over 7366.00 frames. ], tot_loss[loss=0.1372, simple_loss=0.2174, pruned_loss=0.02847, over 1439810.80 frames. ], batch size: 63, lr: 5.50e-03, grad_scale: 8.0 +2023-03-21 05:40:29,505 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=83667.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:40:33,908 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 05:40:34,919 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 05:40:39,200 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0499, 2.6008, 2.9455, 2.9394, 2.6054, 2.5610, 3.0956, 2.2706], + device='cuda:0'), covar=tensor([0.0465, 0.0482, 0.0573, 0.0670, 0.0596, 0.0844, 0.0625, 0.1635], + device='cuda:0'), in_proj_covar=tensor([0.0334, 0.0340, 0.0275, 0.0362, 0.0297, 0.0295, 0.0346, 0.0267], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 05:40:44,519 INFO [train.py:901] (0/2) Epoch 30, batch 1800, loss[loss=0.1325, simple_loss=0.2128, pruned_loss=0.02609, over 7271.00 frames. ], tot_loss[loss=0.1373, simple_loss=0.2173, pruned_loss=0.0286, over 1438990.88 frames. ], batch size: 70, lr: 5.50e-03, grad_scale: 16.0 +2023-03-21 05:40:49,948 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-03-21 05:40:55,173 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1267, 4.0285, 3.3268, 3.6672, 3.0889, 2.2397, 1.9390, 4.2281], + device='cuda:0'), covar=tensor([0.0054, 0.0070, 0.0126, 0.0071, 0.0161, 0.0561, 0.0603, 0.0052], + device='cuda:0'), in_proj_covar=tensor([0.0104, 0.0086, 0.0104, 0.0089, 0.0119, 0.0128, 0.0124, 0.0098], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 05:40:56,065 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 05:41:04,457 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.308e+02 1.893e+02 2.107e+02 2.536e+02 4.167e+02, threshold=4.214e+02, percent-clipped=0.0 +2023-03-21 05:41:08,299 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8498, 2.2661, 1.6854, 2.5568, 2.5902, 2.6795, 2.2268, 2.4026], + device='cuda:0'), covar=tensor([0.2006, 0.1035, 0.3622, 0.0822, 0.0316, 0.0242, 0.0325, 0.0304], + device='cuda:0'), in_proj_covar=tensor([0.0261, 0.0234, 0.0257, 0.0262, 0.0192, 0.0186, 0.0209, 0.0224], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 05:41:09,163 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 05:41:10,635 INFO [train.py:901] (0/2) Epoch 30, batch 1850, loss[loss=0.143, simple_loss=0.2262, pruned_loss=0.0299, over 7268.00 frames. ], tot_loss[loss=0.1374, simple_loss=0.2174, pruned_loss=0.02867, over 1438301.92 frames. ], batch size: 52, lr: 5.49e-03, grad_scale: 16.0 +2023-03-21 05:41:17,124 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=83760.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:41:19,269 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.77 vs. limit=5.0 +2023-03-21 05:41:19,548 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 05:41:21,078 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.5374, 5.0067, 5.0663, 5.4525, 5.4258, 5.4984, 4.8707, 5.1087], + device='cuda:0'), covar=tensor([0.0692, 0.2418, 0.1613, 0.1081, 0.0713, 0.1130, 0.0680, 0.1022], + device='cuda:0'), in_proj_covar=tensor([0.0243, 0.0375, 0.0284, 0.0294, 0.0213, 0.0356, 0.0214, 0.0258], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:41:35,761 INFO [train.py:901] (0/2) Epoch 30, batch 1900, loss[loss=0.1262, simple_loss=0.213, pruned_loss=0.01971, over 7348.00 frames. ], tot_loss[loss=0.1366, simple_loss=0.2166, pruned_loss=0.02833, over 1439723.40 frames. ], batch size: 61, lr: 5.49e-03, grad_scale: 16.0 +2023-03-21 05:41:36,779 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 05:41:40,813 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.13 vs. limit=5.0 +2023-03-21 05:41:43,224 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0524, 3.7301, 4.1669, 4.2242, 4.1398, 4.2053, 4.2878, 4.1330], + device='cuda:0'), covar=tensor([0.0033, 0.0094, 0.0030, 0.0028, 0.0027, 0.0029, 0.0023, 0.0044], + device='cuda:0'), in_proj_covar=tensor([0.0046, 0.0065, 0.0054, 0.0053, 0.0052, 0.0057, 0.0048, 0.0071], + device='cuda:0'), out_proj_covar=tensor([8.2554e-05, 1.3964e-04, 1.0439e-04, 9.7988e-05, 9.4294e-05, 1.0523e-04, + 9.9517e-05, 1.3757e-04], device='cuda:0') +2023-03-21 05:41:46,374 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3289, 3.4922, 2.5009, 3.8873, 2.7651, 3.4277, 1.6060, 2.3820], + device='cuda:0'), covar=tensor([0.0473, 0.0877, 0.2399, 0.0532, 0.0503, 0.0834, 0.3778, 0.1855], + device='cuda:0'), in_proj_covar=tensor([0.0277, 0.0254, 0.0286, 0.0268, 0.0267, 0.0266, 0.0244, 0.0266], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 05:41:48,834 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=83821.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:41:56,979 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.134e+02 1.794e+02 2.093e+02 2.533e+02 5.661e+02, threshold=4.186e+02, percent-clipped=2.0 +2023-03-21 05:42:02,545 INFO [train.py:901] (0/2) Epoch 30, batch 1950, loss[loss=0.1467, simple_loss=0.2211, pruned_loss=0.03619, over 7304.00 frames. ], tot_loss[loss=0.1366, simple_loss=0.2164, pruned_loss=0.02834, over 1440274.30 frames. ], batch size: 49, lr: 5.49e-03, grad_scale: 16.0 +2023-03-21 05:42:03,036 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 05:42:05,178 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=83852.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:42:12,583 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 05:42:17,222 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 05:42:17,754 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 05:42:28,461 INFO [train.py:901] (0/2) Epoch 30, batch 2000, loss[loss=0.1334, simple_loss=0.2243, pruned_loss=0.0212, over 7138.00 frames. ], tot_loss[loss=0.1378, simple_loss=0.2181, pruned_loss=0.02871, over 1442030.22 frames. ], batch size: 98, lr: 5.49e-03, grad_scale: 16.0 +2023-03-21 05:42:30,060 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=83900.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:42:31,162 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3599, 2.4874, 2.3075, 3.5132, 1.9164, 3.1932, 1.4287, 3.0001], + device='cuda:0'), covar=tensor([0.0176, 0.1230, 0.1791, 0.0246, 0.3541, 0.0289, 0.1322, 0.0314], + device='cuda:0'), in_proj_covar=tensor([0.0200, 0.0258, 0.0269, 0.0204, 0.0258, 0.0212, 0.0240, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 05:42:35,454 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 05:42:46,592 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 05:42:48,525 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.334e+02 1.864e+02 2.197e+02 2.640e+02 5.149e+02, threshold=4.395e+02, percent-clipped=2.0 +2023-03-21 05:42:50,666 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=83940.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:42:53,975 INFO [train.py:901] (0/2) Epoch 30, batch 2050, loss[loss=0.1179, simple_loss=0.1966, pruned_loss=0.01957, over 7175.00 frames. ], tot_loss[loss=0.1376, simple_loss=0.2178, pruned_loss=0.0287, over 1443361.69 frames. ], batch size: 39, lr: 5.49e-03, grad_scale: 16.0 +2023-03-21 05:42:54,487 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 05:42:55,082 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=83949.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:43:10,138 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0630, 3.1152, 2.2606, 3.5508, 2.6093, 3.1110, 1.4695, 2.1101], + device='cuda:0'), covar=tensor([0.0439, 0.0701, 0.2388, 0.0634, 0.0318, 0.0706, 0.3523, 0.1782], + device='cuda:0'), in_proj_covar=tensor([0.0278, 0.0255, 0.0286, 0.0269, 0.0268, 0.0266, 0.0243, 0.0266], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 05:43:12,398 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.82 vs. limit=5.0 +2023-03-21 05:43:13,682 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=83985.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:43:15,111 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=83988.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:43:19,525 INFO [train.py:901] (0/2) Epoch 30, batch 2100, loss[loss=0.146, simple_loss=0.2279, pruned_loss=0.03208, over 7304.00 frames. ], tot_loss[loss=0.1373, simple_loss=0.2174, pruned_loss=0.02855, over 1441255.91 frames. ], batch size: 83, lr: 5.49e-03, grad_scale: 16.0 +2023-03-21 05:43:21,349 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-84000.pt +2023-03-21 05:43:25,207 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7708, 3.9278, 3.7027, 4.0026, 3.5344, 3.9205, 4.2508, 4.2569], + device='cuda:0'), covar=tensor([0.0239, 0.0166, 0.0256, 0.0154, 0.0369, 0.0309, 0.0224, 0.0179], + device='cuda:0'), in_proj_covar=tensor([0.0113, 0.0118, 0.0109, 0.0115, 0.0105, 0.0095, 0.0091, 0.0090], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:43:30,850 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=84010.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 05:43:32,801 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 05:43:35,792 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 05:43:43,748 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.225e+02 1.924e+02 2.210e+02 2.575e+02 6.608e+02, threshold=4.420e+02, percent-clipped=1.0 +2023-03-21 05:43:48,908 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=84046.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:43:49,291 INFO [train.py:901] (0/2) Epoch 30, batch 2150, loss[loss=0.1486, simple_loss=0.2271, pruned_loss=0.03502, over 7344.00 frames. ], tot_loss[loss=0.1373, simple_loss=0.2177, pruned_loss=0.02843, over 1440800.02 frames. ], batch size: 54, lr: 5.48e-03, grad_scale: 16.0 +2023-03-21 05:44:15,520 INFO [train.py:901] (0/2) Epoch 30, batch 2200, loss[loss=0.128, simple_loss=0.2122, pruned_loss=0.02188, over 7322.00 frames. ], tot_loss[loss=0.1371, simple_loss=0.2177, pruned_loss=0.0282, over 1443311.95 frames. ], batch size: 59, lr: 5.48e-03, grad_scale: 16.0 +2023-03-21 05:44:15,647 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=84097.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:44:21,039 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 05:44:25,098 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=84116.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:44:34,904 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.268e+02 1.821e+02 2.183e+02 2.536e+02 5.271e+02, threshold=4.365e+02, percent-clipped=2.0 +2023-03-21 05:44:40,502 INFO [train.py:901] (0/2) Epoch 30, batch 2250, loss[loss=0.1417, simple_loss=0.223, pruned_loss=0.03021, over 7323.00 frames. ], tot_loss[loss=0.1373, simple_loss=0.2178, pruned_loss=0.02839, over 1442758.56 frames. ], batch size: 61, lr: 5.48e-03, grad_scale: 16.0 +2023-03-21 05:44:46,722 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=84158.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:44:54,828 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 05:44:55,327 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 05:45:04,001 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5745, 2.1519, 2.7352, 2.6396, 2.6642, 2.5022, 2.1690, 2.6966], + device='cuda:0'), covar=tensor([0.1652, 0.1470, 0.1492, 0.1575, 0.1319, 0.1266, 0.2787, 0.1463], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0062, 0.0049, 0.0048, 0.0046, 0.0046, 0.0066, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 05:45:06,334 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 05:45:06,808 INFO [train.py:901] (0/2) Epoch 30, batch 2300, loss[loss=0.1398, simple_loss=0.2247, pruned_loss=0.02743, over 7370.00 frames. ], tot_loss[loss=0.1371, simple_loss=0.2174, pruned_loss=0.02838, over 1441303.50 frames. ], batch size: 73, lr: 5.48e-03, grad_scale: 16.0 +2023-03-21 05:45:07,956 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=84199.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:45:26,700 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.187e+02 1.852e+02 2.178e+02 2.778e+02 7.190e+02, threshold=4.356e+02, percent-clipped=4.0 +2023-03-21 05:45:26,858 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=84236.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:45:32,858 INFO [train.py:901] (0/2) Epoch 30, batch 2350, loss[loss=0.1432, simple_loss=0.2214, pruned_loss=0.03253, over 7318.00 frames. ], tot_loss[loss=0.1382, simple_loss=0.2185, pruned_loss=0.02896, over 1441845.06 frames. ], batch size: 59, lr: 5.48e-03, grad_scale: 16.0 +2023-03-21 05:45:40,183 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=84260.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:45:42,793 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.83 vs. limit=5.0 +2023-03-21 05:45:54,626 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 05:45:58,570 INFO [train.py:901] (0/2) Epoch 30, batch 2400, loss[loss=0.1021, simple_loss=0.1715, pruned_loss=0.01635, over 6970.00 frames. ], tot_loss[loss=0.1373, simple_loss=0.218, pruned_loss=0.02831, over 1440881.10 frames. ], batch size: 35, lr: 5.48e-03, grad_scale: 8.0 +2023-03-21 05:45:58,739 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=84297.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:46:00,681 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 05:46:02,737 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=84305.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 05:46:10,731 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 05:46:10,897 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5794, 3.3059, 3.3472, 3.4278, 3.0076, 2.8839, 3.4283, 2.7195], + device='cuda:0'), covar=tensor([0.0521, 0.0514, 0.0636, 0.0607, 0.0739, 0.0966, 0.0639, 0.1968], + device='cuda:0'), in_proj_covar=tensor([0.0332, 0.0340, 0.0273, 0.0360, 0.0295, 0.0293, 0.0345, 0.0265], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 05:46:13,758 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 05:46:18,999 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=84336.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:46:19,342 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.243e+02 1.768e+02 2.044e+02 2.372e+02 6.082e+02, threshold=4.088e+02, percent-clipped=2.0 +2023-03-21 05:46:21,385 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=84341.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:46:24,982 INFO [train.py:901] (0/2) Epoch 30, batch 2450, loss[loss=0.1282, simple_loss=0.2102, pruned_loss=0.02307, over 7288.00 frames. ], tot_loss[loss=0.1369, simple_loss=0.2174, pruned_loss=0.02819, over 1439469.25 frames. ], batch size: 57, lr: 5.47e-03, grad_scale: 8.0 +2023-03-21 05:46:40,409 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 05:46:44,798 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.09 vs. limit=5.0 +2023-03-21 05:46:49,911 INFO [train.py:901] (0/2) Epoch 30, batch 2500, loss[loss=0.1338, simple_loss=0.2147, pruned_loss=0.02646, over 7292.00 frames. ], tot_loss[loss=0.1374, simple_loss=0.2182, pruned_loss=0.02829, over 1440233.96 frames. ], batch size: 68, lr: 5.47e-03, grad_scale: 8.0 +2023-03-21 05:46:50,072 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=84397.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:46:59,891 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=84416.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:47:01,546 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0141, 3.2706, 2.9573, 3.2819, 3.2332, 2.6591, 3.1410, 3.0537], + device='cuda:0'), covar=tensor([0.0734, 0.1383, 0.1097, 0.0792, 0.1205, 0.0702, 0.0561, 0.1226], + device='cuda:0'), in_proj_covar=tensor([0.0056, 0.0056, 0.0063, 0.0055, 0.0053, 0.0057, 0.0054, 0.0051], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:47:05,758 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.90 vs. limit=5.0 +2023-03-21 05:47:06,554 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 05:47:11,410 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.270e+02 1.780e+02 2.087e+02 2.604e+02 4.810e+02, threshold=4.173e+02, percent-clipped=3.0 +2023-03-21 05:47:16,440 INFO [train.py:901] (0/2) Epoch 30, batch 2550, loss[loss=0.1386, simple_loss=0.21, pruned_loss=0.0336, over 7275.00 frames. ], tot_loss[loss=0.1371, simple_loss=0.2178, pruned_loss=0.02823, over 1441524.32 frames. ], batch size: 47, lr: 5.47e-03, grad_scale: 8.0 +2023-03-21 05:47:19,554 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=84453.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:47:25,091 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=84464.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:47:41,567 INFO [train.py:901] (0/2) Epoch 30, batch 2600, loss[loss=0.155, simple_loss=0.231, pruned_loss=0.03945, over 7287.00 frames. ], tot_loss[loss=0.1373, simple_loss=0.2177, pruned_loss=0.02843, over 1439705.98 frames. ], batch size: 66, lr: 5.47e-03, grad_scale: 8.0 +2023-03-21 05:47:45,621 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4921, 1.2985, 1.6050, 1.9031, 1.6432, 1.8086, 1.3446, 1.8276], + device='cuda:0'), covar=tensor([0.1918, 0.2598, 0.1136, 0.1391, 0.1534, 0.1747, 0.1820, 0.2184], + device='cuda:0'), in_proj_covar=tensor([0.0071, 0.0075, 0.0062, 0.0057, 0.0058, 0.0060, 0.0094, 0.0059], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 05:47:49,786 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.87 vs. limit=2.0 +2023-03-21 05:47:55,284 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=84525.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:48:01,089 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.046e+02 1.691e+02 2.069e+02 2.465e+02 5.601e+02, threshold=4.139e+02, percent-clipped=3.0 +2023-03-21 05:48:06,542 INFO [train.py:901] (0/2) Epoch 30, batch 2650, loss[loss=0.1353, simple_loss=0.2194, pruned_loss=0.02563, over 7332.00 frames. ], tot_loss[loss=0.1374, simple_loss=0.2179, pruned_loss=0.02842, over 1441080.80 frames. ], batch size: 54, lr: 5.47e-03, grad_scale: 8.0 +2023-03-21 05:48:11,016 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=84555.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:48:26,209 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=84586.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:48:29,119 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=84592.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:48:31,476 INFO [train.py:901] (0/2) Epoch 30, batch 2700, loss[loss=0.1463, simple_loss=0.2195, pruned_loss=0.03655, over 7253.00 frames. ], tot_loss[loss=0.1372, simple_loss=0.2179, pruned_loss=0.02828, over 1443399.80 frames. ], batch size: 47, lr: 5.47e-03, grad_scale: 8.0 +2023-03-21 05:48:35,975 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=84605.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:48:49,559 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=84633.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:48:51,390 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.327e+02 1.736e+02 2.051e+02 2.342e+02 3.644e+02, threshold=4.103e+02, percent-clipped=0.0 +2023-03-21 05:48:51,581 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0075, 3.0150, 2.0781, 3.3344, 2.4861, 3.2383, 1.4569, 1.9946], + device='cuda:0'), covar=tensor([0.0516, 0.1010, 0.2945, 0.0721, 0.0675, 0.0667, 0.3860, 0.1972], + device='cuda:0'), in_proj_covar=tensor([0.0281, 0.0256, 0.0290, 0.0271, 0.0270, 0.0268, 0.0245, 0.0269], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 05:48:53,529 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=84641.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:48:56,453 INFO [train.py:901] (0/2) Epoch 30, batch 2750, loss[loss=0.1355, simple_loss=0.22, pruned_loss=0.02552, over 7252.00 frames. ], tot_loss[loss=0.1371, simple_loss=0.2178, pruned_loss=0.02823, over 1444744.39 frames. ], batch size: 89, lr: 5.47e-03, grad_scale: 8.0 +2023-03-21 05:48:59,522 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=84653.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:49:17,178 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=84689.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:49:18,600 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=84692.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:49:19,657 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=84694.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:49:21,000 INFO [train.py:901] (0/2) Epoch 30, batch 2800, loss[loss=0.1458, simple_loss=0.2257, pruned_loss=0.03292, over 7266.00 frames. ], tot_loss[loss=0.1371, simple_loss=0.2175, pruned_loss=0.02839, over 1442529.94 frames. ], batch size: 70, lr: 5.46e-03, grad_scale: 8.0 +2023-03-21 05:49:33,650 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-30.pt +2023-03-21 05:49:52,079 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 05:49:52,155 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 05:49:52,192 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 05:49:52,219 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 05:49:52,456 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 05:49:52,479 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 05:49:52,495 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 05:49:52,997 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 05:49:53,095 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 05:49:53,115 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 05:49:53,134 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 05:49:53,353 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 05:49:53,360 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 05:49:53,371 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 05:49:53,580 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 05:49:53,584 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 05:49:53,653 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 05:49:55,877 INFO [train.py:901] (0/2) Epoch 31, batch 0, loss[loss=0.1379, simple_loss=0.2155, pruned_loss=0.03015, over 7279.00 frames. ], tot_loss[loss=0.1379, simple_loss=0.2155, pruned_loss=0.03015, over 7279.00 frames. ], batch size: 57, lr: 5.37e-03, grad_scale: 8.0 +2023-03-21 05:49:55,879 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 05:50:18,104 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5198, 3.3043, 3.3177, 3.4715, 3.0102, 2.8577, 3.4743, 2.6231], + device='cuda:0'), covar=tensor([0.0314, 0.0451, 0.0564, 0.0524, 0.0584, 0.0779, 0.0548, 0.1836], + device='cuda:0'), in_proj_covar=tensor([0.0330, 0.0336, 0.0271, 0.0357, 0.0294, 0.0291, 0.0343, 0.0264], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 05:50:21,869 INFO [train.py:935] (0/2) Epoch 31, validation: loss=0.1654, simple_loss=0.2551, pruned_loss=0.03787, over 1622729.00 frames. +2023-03-21 05:50:21,869 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 05:50:28,510 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 05:50:29,973 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.305e+02 1.866e+02 2.119e+02 2.515e+02 6.246e+02, threshold=4.238e+02, percent-clipped=1.0 +2023-03-21 05:50:38,125 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=84753.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:50:39,050 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 05:50:46,719 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 05:50:48,343 INFO [train.py:901] (0/2) Epoch 31, batch 50, loss[loss=0.1437, simple_loss=0.2277, pruned_loss=0.02982, over 7324.00 frames. ], tot_loss[loss=0.1393, simple_loss=0.2182, pruned_loss=0.03015, over 322591.51 frames. ], batch size: 75, lr: 5.37e-03, grad_scale: 8.0 +2023-03-21 05:50:49,801 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 05:50:50,396 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9563, 3.6427, 3.6746, 3.7395, 3.6518, 3.5480, 3.8606, 3.5010], + device='cuda:0'), covar=tensor([0.0153, 0.0209, 0.0123, 0.0149, 0.0450, 0.0117, 0.0159, 0.0166], + device='cuda:0'), in_proj_covar=tensor([0.0090, 0.0099, 0.0096, 0.0086, 0.0169, 0.0105, 0.0103, 0.0107], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:50:52,818 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 05:51:03,230 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=84800.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:51:03,697 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=84801.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:51:06,777 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9958, 2.2119, 1.7581, 2.8594, 2.7163, 2.8446, 2.4259, 2.5410], + device='cuda:0'), covar=tensor([0.1794, 0.1041, 0.3238, 0.0810, 0.0351, 0.0279, 0.0423, 0.0532], + device='cuda:0'), in_proj_covar=tensor([0.0258, 0.0232, 0.0257, 0.0263, 0.0191, 0.0187, 0.0210, 0.0225], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 05:51:10,601 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 05:51:11,115 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 05:51:13,504 INFO [train.py:901] (0/2) Epoch 31, batch 100, loss[loss=0.1267, simple_loss=0.203, pruned_loss=0.02517, over 7329.00 frames. ], tot_loss[loss=0.1383, simple_loss=0.2184, pruned_loss=0.02912, over 570794.87 frames. ], batch size: 61, lr: 5.37e-03, grad_scale: 8.0 +2023-03-21 05:51:21,417 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.272e+02 1.807e+02 2.107e+02 2.444e+02 8.407e+02, threshold=4.214e+02, percent-clipped=1.0 +2023-03-21 05:51:22,104 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5786, 1.2602, 1.6064, 2.0231, 1.6752, 1.8201, 1.4192, 1.8458], + device='cuda:0'), covar=tensor([0.2229, 0.4155, 0.0852, 0.0939, 0.1848, 0.1935, 0.1998, 0.2086], + device='cuda:0'), in_proj_covar=tensor([0.0071, 0.0074, 0.0062, 0.0057, 0.0057, 0.0059, 0.0092, 0.0059], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 05:51:30,585 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=84855.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:51:34,840 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=84861.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:51:39,598 INFO [train.py:901] (0/2) Epoch 31, batch 150, loss[loss=0.15, simple_loss=0.2228, pruned_loss=0.03858, over 7253.00 frames. ], tot_loss[loss=0.1392, simple_loss=0.2196, pruned_loss=0.02935, over 764871.46 frames. ], batch size: 64, lr: 5.37e-03, grad_scale: 8.0 +2023-03-21 05:51:44,722 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=84881.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:51:50,298 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=84892.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:51:55,837 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=84903.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:52:04,878 INFO [train.py:901] (0/2) Epoch 31, batch 200, loss[loss=0.1291, simple_loss=0.2118, pruned_loss=0.02316, over 7283.00 frames. ], tot_loss[loss=0.138, simple_loss=0.2189, pruned_loss=0.02854, over 915196.89 frames. ], batch size: 64, lr: 5.37e-03, grad_scale: 8.0 +2023-03-21 05:52:09,347 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 05:52:12,945 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.363e+02 1.822e+02 2.128e+02 2.384e+02 5.145e+02, threshold=4.256e+02, percent-clipped=2.0 +2023-03-21 05:52:14,576 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 05:52:15,114 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=84940.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:52:21,226 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 05:52:21,332 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=84951.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:52:31,089 INFO [train.py:901] (0/2) Epoch 31, batch 250, loss[loss=0.1095, simple_loss=0.1935, pruned_loss=0.0127, over 7224.00 frames. ], tot_loss[loss=0.1371, simple_loss=0.2177, pruned_loss=0.02821, over 1032839.72 frames. ], batch size: 39, lr: 5.37e-03, grad_scale: 8.0 +2023-03-21 05:52:34,142 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 05:52:40,295 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=84989.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:52:41,822 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=84992.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:52:52,153 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=85012.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:52:53,014 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 05:52:56,388 INFO [train.py:901] (0/2) Epoch 31, batch 300, loss[loss=0.1205, simple_loss=0.1934, pruned_loss=0.02378, over 7002.00 frames. ], tot_loss[loss=0.1366, simple_loss=0.2176, pruned_loss=0.02782, over 1125892.01 frames. ], batch size: 35, lr: 5.37e-03, grad_scale: 8.0 +2023-03-21 05:53:03,982 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 05:53:05,474 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.264e+02 1.823e+02 2.082e+02 2.425e+02 4.646e+02, threshold=4.164e+02, percent-clipped=1.0 +2023-03-21 05:53:06,966 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=85040.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:53:22,339 INFO [train.py:901] (0/2) Epoch 31, batch 350, loss[loss=0.1656, simple_loss=0.2395, pruned_loss=0.04588, over 7321.00 frames. ], tot_loss[loss=0.1369, simple_loss=0.2173, pruned_loss=0.02821, over 1194952.50 frames. ], batch size: 59, lr: 5.36e-03, grad_scale: 8.0 +2023-03-21 05:53:33,975 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1544, 4.5691, 4.6175, 4.5960, 4.5948, 4.1456, 4.6489, 4.5184], + device='cuda:0'), covar=tensor([0.0439, 0.0425, 0.0413, 0.0484, 0.0283, 0.0381, 0.0333, 0.0415], + device='cuda:0'), in_proj_covar=tensor([0.0208, 0.0243, 0.0185, 0.0187, 0.0148, 0.0219, 0.0194, 0.0143], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 05:53:36,937 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 05:53:38,058 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6804, 1.2740, 1.8692, 2.1961, 1.9129, 2.0125, 1.8025, 1.9684], + device='cuda:0'), covar=tensor([0.2765, 0.3515, 0.2031, 0.1312, 0.2717, 0.2009, 0.1914, 0.1936], + device='cuda:0'), in_proj_covar=tensor([0.0069, 0.0072, 0.0061, 0.0056, 0.0055, 0.0058, 0.0091, 0.0057], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:53:40,232 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-03-21 05:53:40,561 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=85107.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:53:45,797 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=85115.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:53:48,635 INFO [train.py:901] (0/2) Epoch 31, batch 400, loss[loss=0.1492, simple_loss=0.2292, pruned_loss=0.03464, over 7265.00 frames. ], tot_loss[loss=0.1366, simple_loss=0.2172, pruned_loss=0.028, over 1250815.67 frames. ], batch size: 64, lr: 5.36e-03, grad_scale: 8.0 +2023-03-21 05:53:56,683 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.161e+02 1.732e+02 2.014e+02 2.395e+02 5.773e+02, threshold=4.029e+02, percent-clipped=2.0 +2023-03-21 05:54:00,147 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-03-21 05:54:06,355 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=85156.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:54:12,536 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=85168.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 05:54:13,875 INFO [train.py:901] (0/2) Epoch 31, batch 450, loss[loss=0.1535, simple_loss=0.2275, pruned_loss=0.03976, over 7292.00 frames. ], tot_loss[loss=0.1363, simple_loss=0.2169, pruned_loss=0.02779, over 1293279.28 frames. ], batch size: 68, lr: 5.36e-03, grad_scale: 8.0 +2023-03-21 05:54:16,548 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=85176.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:54:19,052 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=85181.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:54:19,456 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 05:54:19,997 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 05:54:23,579 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0402, 4.2303, 3.9809, 4.2077, 3.8491, 4.0806, 4.4526, 4.4980], + device='cuda:0'), covar=tensor([0.0205, 0.0123, 0.0214, 0.0154, 0.0379, 0.0247, 0.0218, 0.0164], + device='cuda:0'), in_proj_covar=tensor([0.0116, 0.0121, 0.0113, 0.0119, 0.0108, 0.0098, 0.0095, 0.0095], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:54:24,680 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3924, 3.1002, 2.0627, 3.2025, 3.3828, 3.4909, 3.2183, 2.9343], + device='cuda:0'), covar=tensor([0.1897, 0.0803, 0.3639, 0.0567, 0.0294, 0.0373, 0.0391, 0.0392], + device='cuda:0'), in_proj_covar=tensor([0.0254, 0.0230, 0.0252, 0.0257, 0.0188, 0.0185, 0.0207, 0.0222], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 05:54:35,748 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3977, 1.4084, 1.4855, 1.5091, 1.5185, 1.4380, 1.4359, 1.1880], + device='cuda:0'), covar=tensor([0.0197, 0.0164, 0.0191, 0.0167, 0.0130, 0.0165, 0.0174, 0.0156], + device='cuda:0'), in_proj_covar=tensor([0.0033, 0.0030, 0.0030, 0.0031, 0.0030, 0.0028, 0.0031, 0.0040], + device='cuda:0'), out_proj_covar=tensor([3.7627e-05, 3.3832e-05, 3.3824e-05, 3.4724e-05, 3.3216e-05, 3.1810e-05, + 3.5307e-05, 4.5025e-05], device='cuda:0') +2023-03-21 05:54:40,657 INFO [train.py:901] (0/2) Epoch 31, batch 500, loss[loss=0.1315, simple_loss=0.2121, pruned_loss=0.02547, over 7305.00 frames. ], tot_loss[loss=0.1364, simple_loss=0.2168, pruned_loss=0.02803, over 1325289.86 frames. ], batch size: 80, lr: 5.36e-03, grad_scale: 8.0 +2023-03-21 05:54:44,803 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=85229.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:54:48,810 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.266e+02 1.776e+02 2.130e+02 2.483e+02 6.645e+02, threshold=4.260e+02, percent-clipped=4.0 +2023-03-21 05:54:52,886 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 05:54:54,836 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 05:54:55,352 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 05:54:57,386 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 05:55:02,372 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 05:55:05,845 INFO [train.py:901] (0/2) Epoch 31, batch 550, loss[loss=0.1396, simple_loss=0.2279, pruned_loss=0.02564, over 7325.00 frames. ], tot_loss[loss=0.1363, simple_loss=0.2166, pruned_loss=0.02803, over 1351532.35 frames. ], batch size: 61, lr: 5.36e-03, grad_scale: 8.0 +2023-03-21 05:55:11,819 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.57 vs. limit=2.0 +2023-03-21 05:55:12,241 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.88 vs. limit=2.0 +2023-03-21 05:55:13,456 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 05:55:16,274 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=85289.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:55:22,773 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 05:55:25,299 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=85307.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:55:26,229 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 05:55:32,083 INFO [train.py:901] (0/2) Epoch 31, batch 600, loss[loss=0.1322, simple_loss=0.2095, pruned_loss=0.0274, over 7332.00 frames. ], tot_loss[loss=0.1367, simple_loss=0.2167, pruned_loss=0.02829, over 1371287.73 frames. ], batch size: 75, lr: 5.36e-03, grad_scale: 8.0 +2023-03-21 05:55:32,660 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 05:55:40,079 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.256e+02 1.834e+02 2.062e+02 2.649e+02 3.908e+02, threshold=4.123e+02, percent-clipped=0.0 +2023-03-21 05:55:40,158 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=85337.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:55:45,825 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.29 vs. limit=5.0 +2023-03-21 05:55:47,538 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 05:55:57,090 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 05:55:57,582 INFO [train.py:901] (0/2) Epoch 31, batch 650, loss[loss=0.1603, simple_loss=0.2431, pruned_loss=0.03875, over 7337.00 frames. ], tot_loss[loss=0.1371, simple_loss=0.2174, pruned_loss=0.0284, over 1389512.23 frames. ], batch size: 54, lr: 5.35e-03, grad_scale: 8.0 +2023-03-21 05:56:14,057 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1618, 3.7881, 3.7489, 3.8849, 3.7689, 3.6381, 4.0111, 3.5464], + device='cuda:0'), covar=tensor([0.0156, 0.0177, 0.0151, 0.0157, 0.0472, 0.0140, 0.0151, 0.0165], + device='cuda:0'), in_proj_covar=tensor([0.0090, 0.0098, 0.0096, 0.0085, 0.0169, 0.0105, 0.0103, 0.0107], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:56:14,967 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 05:56:23,516 INFO [train.py:901] (0/2) Epoch 31, batch 700, loss[loss=0.1413, simple_loss=0.2253, pruned_loss=0.0287, over 7292.00 frames. ], tot_loss[loss=0.1366, simple_loss=0.2171, pruned_loss=0.02802, over 1402228.71 frames. ], batch size: 86, lr: 5.35e-03, grad_scale: 8.0 +2023-03-21 05:56:23,526 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 05:56:27,257 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=85428.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:56:31,664 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.177e+02 1.664e+02 1.929e+02 2.389e+02 3.949e+02, threshold=3.858e+02, percent-clipped=0.0 +2023-03-21 05:56:41,279 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=85456.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:56:45,293 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=85463.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 05:56:47,818 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 05:56:48,352 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 05:56:49,809 INFO [train.py:901] (0/2) Epoch 31, batch 750, loss[loss=0.1331, simple_loss=0.223, pruned_loss=0.02164, over 7332.00 frames. ], tot_loss[loss=0.1367, simple_loss=0.2174, pruned_loss=0.028, over 1411752.98 frames. ], batch size: 54, lr: 5.35e-03, grad_scale: 8.0 +2023-03-21 05:56:50,019 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=85471.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:56:56,107 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=85483.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:56:59,192 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=85489.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:57:03,233 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 05:57:06,815 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=85504.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:57:07,720 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 05:57:13,595 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 05:57:14,603 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 05:57:14,965 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-03-21 05:57:15,100 INFO [train.py:901] (0/2) Epoch 31, batch 800, loss[loss=0.149, simple_loss=0.2302, pruned_loss=0.03389, over 7291.00 frames. ], tot_loss[loss=0.1365, simple_loss=0.217, pruned_loss=0.02797, over 1420081.77 frames. ], batch size: 68, lr: 5.35e-03, grad_scale: 8.0 +2023-03-21 05:57:19,936 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=85530.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:57:23,323 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.239e+02 1.779e+02 2.084e+02 2.516e+02 3.956e+02, threshold=4.169e+02, percent-clipped=1.0 +2023-03-21 05:57:24,437 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0272, 3.9585, 3.5233, 3.5675, 3.0306, 2.3161, 1.9399, 4.0299], + device='cuda:0'), covar=tensor([0.0048, 0.0055, 0.0096, 0.0065, 0.0146, 0.0459, 0.0527, 0.0044], + device='cuda:0'), in_proj_covar=tensor([0.0103, 0.0085, 0.0102, 0.0088, 0.0118, 0.0125, 0.0122, 0.0096], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 05:57:24,821 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 05:57:27,011 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=85544.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:57:38,287 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6746, 3.5199, 3.4177, 3.6182, 3.0974, 2.8727, 3.8109, 2.6415], + device='cuda:0'), covar=tensor([0.0444, 0.0468, 0.0527, 0.0550, 0.0676, 0.0875, 0.0598, 0.1768], + device='cuda:0'), in_proj_covar=tensor([0.0333, 0.0339, 0.0272, 0.0356, 0.0295, 0.0290, 0.0344, 0.0264], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 05:57:38,775 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4603, 3.2938, 2.3384, 3.7073, 2.7932, 3.2864, 1.4990, 2.3372], + device='cuda:0'), covar=tensor([0.0441, 0.0748, 0.2353, 0.0530, 0.0439, 0.0672, 0.3433, 0.1932], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0250, 0.0282, 0.0266, 0.0265, 0.0264, 0.0240, 0.0260], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 05:57:41,519 INFO [train.py:901] (0/2) Epoch 31, batch 850, loss[loss=0.1298, simple_loss=0.223, pruned_loss=0.01834, over 7271.00 frames. ], tot_loss[loss=0.1368, simple_loss=0.2173, pruned_loss=0.02817, over 1423519.06 frames. ], batch size: 77, lr: 5.35e-03, grad_scale: 8.0 +2023-03-21 05:57:45,054 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 05:57:45,063 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 05:57:48,487 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.58 vs. limit=5.0 +2023-03-21 05:57:48,498 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-03-21 05:57:50,139 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 05:57:51,821 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=85591.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:57:53,764 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 05:57:56,859 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.23 vs. limit=2.0 +2023-03-21 05:57:59,847 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.22 vs. limit=5.0 +2023-03-21 05:58:00,190 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=85607.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:58:01,944 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-03-21 05:58:07,166 INFO [train.py:901] (0/2) Epoch 31, batch 900, loss[loss=0.1448, simple_loss=0.2253, pruned_loss=0.03214, over 7256.00 frames. ], tot_loss[loss=0.1369, simple_loss=0.2173, pruned_loss=0.02819, over 1428062.01 frames. ], batch size: 89, lr: 5.35e-03, grad_scale: 8.0 +2023-03-21 05:58:09,302 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=85625.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:58:15,249 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.228e+02 1.810e+02 2.129e+02 2.402e+02 5.713e+02, threshold=4.258e+02, percent-clipped=2.0 +2023-03-21 05:58:25,755 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=85655.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:58:33,617 INFO [train.py:901] (0/2) Epoch 31, batch 950, loss[loss=0.1501, simple_loss=0.2355, pruned_loss=0.0323, over 7254.00 frames. ], tot_loss[loss=0.1363, simple_loss=0.2166, pruned_loss=0.02804, over 1429482.70 frames. ], batch size: 64, lr: 5.34e-03, grad_scale: 8.0 +2023-03-21 05:58:34,639 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 05:58:41,337 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=85686.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 05:58:58,444 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 05:58:58,923 INFO [train.py:901] (0/2) Epoch 31, batch 1000, loss[loss=0.1447, simple_loss=0.2255, pruned_loss=0.03196, over 7282.00 frames. ], tot_loss[loss=0.1364, simple_loss=0.2166, pruned_loss=0.02809, over 1432932.02 frames. ], batch size: 68, lr: 5.34e-03, grad_scale: 8.0 +2023-03-21 05:59:08,125 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.135e+02 1.756e+02 2.066e+02 2.390e+02 3.507e+02, threshold=4.132e+02, percent-clipped=0.0 +2023-03-21 05:59:19,074 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 05:59:21,231 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=85763.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:59:25,128 INFO [train.py:901] (0/2) Epoch 31, batch 1050, loss[loss=0.1347, simple_loss=0.2141, pruned_loss=0.0276, over 7360.00 frames. ], tot_loss[loss=0.1366, simple_loss=0.2169, pruned_loss=0.02818, over 1435209.55 frames. ], batch size: 73, lr: 5.34e-03, grad_scale: 8.0 +2023-03-21 05:59:25,242 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=85771.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:59:31,714 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=85784.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:59:35,340 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5040, 2.6822, 2.4102, 2.7558, 2.7471, 2.3979, 2.7422, 2.7462], + device='cuda:0'), covar=tensor([0.0722, 0.0781, 0.0860, 0.1011, 0.0736, 0.0599, 0.0549, 0.0820], + device='cuda:0'), in_proj_covar=tensor([0.0056, 0.0056, 0.0064, 0.0055, 0.0054, 0.0057, 0.0055, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:59:40,011 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 05:59:43,938 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 05:59:45,478 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=85811.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:59:50,035 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=85819.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 05:59:50,052 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8795, 4.4217, 4.2593, 4.8260, 4.7213, 4.8108, 4.2733, 4.3789], + device='cuda:0'), covar=tensor([0.0933, 0.2601, 0.2353, 0.1147, 0.0886, 0.1146, 0.0781, 0.1213], + device='cuda:0'), in_proj_covar=tensor([0.0240, 0.0376, 0.0283, 0.0292, 0.0214, 0.0352, 0.0213, 0.0257], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 05:59:50,976 INFO [train.py:901] (0/2) Epoch 31, batch 1100, loss[loss=0.1454, simple_loss=0.2254, pruned_loss=0.03268, over 7342.00 frames. ], tot_loss[loss=0.1367, simple_loss=0.217, pruned_loss=0.02819, over 1440379.45 frames. ], batch size: 73, lr: 5.34e-03, grad_scale: 8.0 +2023-03-21 05:59:59,602 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.280e+02 1.827e+02 2.007e+02 2.454e+02 6.338e+02, threshold=4.013e+02, percent-clipped=1.0 +2023-03-21 06:00:00,692 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=85839.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:00:05,790 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0114, 2.1400, 2.2380, 1.8643, 2.1116, 2.0952, 1.8653, 1.7492], + device='cuda:0'), covar=tensor([0.0451, 0.0473, 0.0281, 0.0299, 0.0533, 0.0444, 0.0382, 0.0356], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0033, 0.0034, 0.0032, 0.0031, 0.0031, 0.0035, 0.0036], + device='cuda:0'), out_proj_covar=tensor([8.7034e-05, 8.5911e-05, 8.5472e-05, 8.2146e-05, 8.2611e-05, 8.1247e-05, + 8.8103e-05, 9.1293e-05], device='cuda:0') +2023-03-21 06:00:13,688 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 06:00:14,147 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 06:00:16,637 INFO [train.py:901] (0/2) Epoch 31, batch 1150, loss[loss=0.1408, simple_loss=0.2108, pruned_loss=0.03534, over 7350.00 frames. ], tot_loss[loss=0.1367, simple_loss=0.217, pruned_loss=0.02821, over 1442204.77 frames. ], batch size: 63, lr: 5.34e-03, grad_scale: 8.0 +2023-03-21 06:00:18,832 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=85875.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:00:24,193 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=85886.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:00:25,678 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 06:00:26,700 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 06:00:42,912 INFO [train.py:901] (0/2) Epoch 31, batch 1200, loss[loss=0.1282, simple_loss=0.216, pruned_loss=0.02021, over 7245.00 frames. ], tot_loss[loss=0.1366, simple_loss=0.217, pruned_loss=0.02806, over 1443024.07 frames. ], batch size: 89, lr: 5.34e-03, grad_scale: 8.0 +2023-03-21 06:00:47,588 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=85930.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 06:00:50,540 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=85936.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:00:50,863 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.241e+02 1.650e+02 2.176e+02 2.512e+02 6.235e+02, threshold=4.352e+02, percent-clipped=4.0 +2023-03-21 06:00:52,276 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.23 vs. limit=2.0 +2023-03-21 06:00:58,494 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 06:01:01,053 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=85957.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:01:07,946 INFO [train.py:901] (0/2) Epoch 31, batch 1250, loss[loss=0.1616, simple_loss=0.2478, pruned_loss=0.03766, over 7111.00 frames. ], tot_loss[loss=0.1369, simple_loss=0.2175, pruned_loss=0.0281, over 1442674.72 frames. ], batch size: 98, lr: 5.34e-03, grad_scale: 8.0 +2023-03-21 06:01:13,053 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=85981.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 06:01:18,817 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=85991.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 06:01:22,809 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 06:01:25,279 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=86002.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:01:26,716 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 06:01:27,721 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 06:01:28,776 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8787, 4.3148, 4.3033, 4.7835, 4.7034, 4.7676, 4.2683, 4.3293], + device='cuda:0'), covar=tensor([0.0707, 0.2401, 0.1994, 0.1090, 0.0888, 0.1054, 0.0868, 0.1079], + device='cuda:0'), in_proj_covar=tensor([0.0240, 0.0377, 0.0282, 0.0293, 0.0216, 0.0352, 0.0212, 0.0258], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 06:01:29,341 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1355, 2.2769, 2.3218, 2.0889, 2.2549, 1.9936, 1.9075, 1.8239], + device='cuda:0'), covar=tensor([0.0342, 0.0336, 0.0219, 0.0320, 0.0606, 0.0515, 0.0299, 0.0315], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0033, 0.0034, 0.0032, 0.0031, 0.0031, 0.0035, 0.0036], + device='cuda:0'), out_proj_covar=tensor([8.7053e-05, 8.6090e-05, 8.5281e-05, 8.1904e-05, 8.2765e-05, 8.1623e-05, + 8.8025e-05, 9.0937e-05], device='cuda:0') +2023-03-21 06:01:33,430 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86018.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:01:34,750 INFO [train.py:901] (0/2) Epoch 31, batch 1300, loss[loss=0.1369, simple_loss=0.2188, pruned_loss=0.02752, over 7351.00 frames. ], tot_loss[loss=0.1371, simple_loss=0.2179, pruned_loss=0.02817, over 1441520.35 frames. ], batch size: 51, lr: 5.33e-03, grad_scale: 8.0 +2023-03-21 06:01:42,740 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.255e+02 1.873e+02 2.179e+02 2.482e+02 5.073e+02, threshold=4.358e+02, percent-clipped=1.0 +2023-03-21 06:01:50,393 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=86052.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:01:51,267 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 06:01:53,813 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 06:01:55,952 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86063.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:01:56,804 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 06:01:58,911 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=86069.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:01:59,808 INFO [train.py:901] (0/2) Epoch 31, batch 1350, loss[loss=0.1422, simple_loss=0.2172, pruned_loss=0.03359, over 7225.00 frames. ], tot_loss[loss=0.1369, simple_loss=0.2174, pruned_loss=0.02822, over 1440419.52 frames. ], batch size: 45, lr: 5.33e-03, grad_scale: 8.0 +2023-03-21 06:02:07,177 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86084.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:02:08,696 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 06:02:19,178 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0373, 2.8617, 3.1235, 2.9721, 3.3679, 3.0386, 2.6759, 3.1833], + device='cuda:0'), covar=tensor([0.1548, 0.0655, 0.1559, 0.1934, 0.0906, 0.1006, 0.2022, 0.1381], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0062, 0.0050, 0.0048, 0.0047, 0.0046, 0.0066, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 06:02:19,232 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5418, 3.5174, 2.5623, 3.9805, 3.0823, 3.7683, 1.5796, 2.4443], + device='cuda:0'), covar=tensor([0.0381, 0.0882, 0.2062, 0.0501, 0.0388, 0.0621, 0.3071, 0.1788], + device='cuda:0'), in_proj_covar=tensor([0.0277, 0.0252, 0.0283, 0.0268, 0.0265, 0.0264, 0.0240, 0.0261], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 06:02:22,233 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86113.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:02:26,059 INFO [train.py:901] (0/2) Epoch 31, batch 1400, loss[loss=0.1368, simple_loss=0.2168, pruned_loss=0.02839, over 7258.00 frames. ], tot_loss[loss=0.1364, simple_loss=0.2166, pruned_loss=0.02811, over 1437804.98 frames. ], batch size: 89, lr: 5.33e-03, grad_scale: 8.0 +2023-03-21 06:02:30,652 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1003, 4.5806, 4.6807, 4.6246, 4.6050, 4.2097, 4.7147, 4.5900], + device='cuda:0'), covar=tensor([0.0572, 0.0457, 0.0419, 0.0531, 0.0301, 0.0418, 0.0360, 0.0414], + device='cuda:0'), in_proj_covar=tensor([0.0213, 0.0247, 0.0189, 0.0191, 0.0150, 0.0220, 0.0193, 0.0144], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 06:02:30,712 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86130.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:02:31,583 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=86132.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:02:34,023 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.336e+02 1.787e+02 2.113e+02 2.434e+02 4.644e+02, threshold=4.226e+02, percent-clipped=1.0 +2023-03-21 06:02:35,147 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86139.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:02:39,669 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 06:02:52,507 INFO [train.py:901] (0/2) Epoch 31, batch 1450, loss[loss=0.1455, simple_loss=0.2266, pruned_loss=0.03223, over 7154.00 frames. ], tot_loss[loss=0.1365, simple_loss=0.2168, pruned_loss=0.02813, over 1438996.05 frames. ], batch size: 98, lr: 5.33e-03, grad_scale: 8.0 +2023-03-21 06:02:58,933 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.84 vs. limit=5.0 +2023-03-21 06:03:00,169 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86186.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:03:00,669 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=86187.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:03:03,959 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-03-21 06:03:04,146 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 06:03:13,685 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3437, 3.2374, 2.4631, 3.8111, 2.9768, 3.6317, 1.5290, 2.4542], + device='cuda:0'), covar=tensor([0.0464, 0.0649, 0.2147, 0.0443, 0.0481, 0.0607, 0.3265, 0.1740], + device='cuda:0'), in_proj_covar=tensor([0.0276, 0.0252, 0.0283, 0.0267, 0.0266, 0.0264, 0.0239, 0.0260], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 06:03:18,055 INFO [train.py:901] (0/2) Epoch 31, batch 1500, loss[loss=0.1183, simple_loss=0.1922, pruned_loss=0.02216, over 7372.00 frames. ], tot_loss[loss=0.1369, simple_loss=0.2172, pruned_loss=0.02836, over 1440150.46 frames. ], batch size: 44, lr: 5.33e-03, grad_scale: 8.0 +2023-03-21 06:03:20,528 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 06:03:21,847 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-03-21 06:03:23,107 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=86231.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:03:24,593 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=86234.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:03:25,991 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.325e+02 1.734e+02 2.049e+02 2.482e+02 4.437e+02, threshold=4.097e+02, percent-clipped=1.0 +2023-03-21 06:03:31,411 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=86246.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:03:44,518 INFO [train.py:901] (0/2) Epoch 31, batch 1550, loss[loss=0.1351, simple_loss=0.2164, pruned_loss=0.02695, over 7268.00 frames. ], tot_loss[loss=0.1368, simple_loss=0.2172, pruned_loss=0.02823, over 1442042.19 frames. ], batch size: 57, lr: 5.33e-03, grad_scale: 8.0 +2023-03-21 06:03:46,026 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 06:03:49,780 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86281.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:03:52,200 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=86286.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 06:04:02,849 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86307.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:04:05,845 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=86313.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:04:09,815 INFO [train.py:901] (0/2) Epoch 31, batch 1600, loss[loss=0.1339, simple_loss=0.2163, pruned_loss=0.02574, over 7280.00 frames. ], tot_loss[loss=0.136, simple_loss=0.2166, pruned_loss=0.02768, over 1443561.92 frames. ], batch size: 70, lr: 5.32e-03, grad_scale: 16.0 +2023-03-21 06:04:13,936 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=86329.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:04:16,956 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 06:04:17,536 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 06:04:18,547 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.253e+02 1.767e+02 1.989e+02 2.470e+02 3.737e+02, threshold=3.978e+02, percent-clipped=0.0 +2023-03-21 06:04:20,574 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 06:04:29,942 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=86358.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:04:30,910 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 06:04:35,362 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 06:04:36,363 INFO [train.py:901] (0/2) Epoch 31, batch 1650, loss[loss=0.1509, simple_loss=0.2329, pruned_loss=0.03444, over 7340.00 frames. ], tot_loss[loss=0.1368, simple_loss=0.2176, pruned_loss=0.02805, over 1443798.73 frames. ], batch size: 54, lr: 5.32e-03, grad_scale: 16.0 +2023-03-21 06:04:39,503 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-03-21 06:04:43,240 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 06:04:55,293 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=86408.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:05:00,292 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 06:05:02,422 INFO [train.py:901] (0/2) Epoch 31, batch 1700, loss[loss=0.1448, simple_loss=0.2271, pruned_loss=0.03126, over 7280.00 frames. ], tot_loss[loss=0.1369, simple_loss=0.2174, pruned_loss=0.02818, over 1443242.21 frames. ], batch size: 77, lr: 5.32e-03, grad_scale: 16.0 +2023-03-21 06:05:04,468 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=86425.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:05:04,919 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 06:05:11,067 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.287e+02 1.933e+02 2.262e+02 2.614e+02 4.811e+02, threshold=4.525e+02, percent-clipped=2.0 +2023-03-21 06:05:12,794 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=86440.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:05:15,277 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=86445.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:05:15,682 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 06:05:28,111 INFO [train.py:901] (0/2) Epoch 31, batch 1750, loss[loss=0.127, simple_loss=0.2103, pruned_loss=0.02182, over 7268.00 frames. ], tot_loss[loss=0.1365, simple_loss=0.2171, pruned_loss=0.02792, over 1444770.79 frames. ], batch size: 47, lr: 5.32e-03, grad_scale: 16.0 +2023-03-21 06:05:31,507 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.65 vs. limit=5.0 +2023-03-21 06:05:35,080 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.58 vs. limit=2.0 +2023-03-21 06:05:40,329 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 06:05:41,387 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2188, 3.8239, 3.8721, 3.9104, 3.8101, 3.6260, 4.0904, 3.5024], + device='cuda:0'), covar=tensor([0.0141, 0.0197, 0.0118, 0.0154, 0.0450, 0.0159, 0.0160, 0.0202], + device='cuda:0'), in_proj_covar=tensor([0.0090, 0.0099, 0.0096, 0.0084, 0.0168, 0.0105, 0.0103, 0.0106], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 06:05:41,757 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 06:05:43,378 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86501.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:05:45,930 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86506.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 06:05:49,520 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2486, 2.3619, 2.1024, 3.5593, 1.7272, 3.3505, 1.2331, 3.0150], + device='cuda:0'), covar=tensor([0.0191, 0.1366, 0.1888, 0.0184, 0.3908, 0.0260, 0.1334, 0.0418], + device='cuda:0'), in_proj_covar=tensor([0.0201, 0.0254, 0.0267, 0.0202, 0.0256, 0.0210, 0.0237, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 06:05:53,962 INFO [train.py:901] (0/2) Epoch 31, batch 1800, loss[loss=0.1247, simple_loss=0.2021, pruned_loss=0.02359, over 7141.00 frames. ], tot_loss[loss=0.1361, simple_loss=0.2168, pruned_loss=0.02772, over 1442209.14 frames. ], batch size: 41, lr: 5.32e-03, grad_scale: 16.0 +2023-03-21 06:05:59,634 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86531.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:06:02,536 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.341e+02 1.702e+02 2.082e+02 2.428e+02 3.685e+02, threshold=4.165e+02, percent-clipped=0.0 +2023-03-21 06:06:02,691 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9681, 2.3876, 2.5441, 2.0612, 2.3060, 2.3829, 1.8855, 1.9917], + device='cuda:0'), covar=tensor([0.0554, 0.0393, 0.0256, 0.0364, 0.0609, 0.0414, 0.0434, 0.0285], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0034, 0.0034, 0.0033, 0.0032, 0.0032, 0.0036, 0.0036], + device='cuda:0'), out_proj_covar=tensor([8.9090e-05, 8.8170e-05, 8.6731e-05, 8.3528e-05, 8.4435e-05, 8.3329e-05, + 9.0710e-05, 9.2020e-05], device='cuda:0') +2023-03-21 06:06:03,540 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 06:06:16,561 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 06:06:19,521 INFO [train.py:901] (0/2) Epoch 31, batch 1850, loss[loss=0.1335, simple_loss=0.2137, pruned_loss=0.02661, over 7273.00 frames. ], tot_loss[loss=0.1363, simple_loss=0.2171, pruned_loss=0.02773, over 1444277.35 frames. ], batch size: 47, lr: 5.32e-03, grad_scale: 16.0 +2023-03-21 06:06:23,590 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=86579.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:06:24,140 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=86580.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:06:26,034 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 06:06:27,105 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86586.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 06:06:36,066 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=86602.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:06:42,330 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86613.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:06:44,705 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 06:06:45,252 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6059, 4.0750, 3.9618, 4.5230, 4.4375, 4.5165, 3.9087, 4.2158], + device='cuda:0'), covar=tensor([0.0925, 0.2747, 0.2763, 0.1314, 0.0956, 0.1417, 0.0948, 0.1181], + device='cuda:0'), in_proj_covar=tensor([0.0242, 0.0376, 0.0287, 0.0292, 0.0214, 0.0356, 0.0215, 0.0260], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 06:06:46,163 INFO [train.py:901] (0/2) Epoch 31, batch 1900, loss[loss=0.1464, simple_loss=0.2325, pruned_loss=0.03011, over 7296.00 frames. ], tot_loss[loss=0.1365, simple_loss=0.2172, pruned_loss=0.02788, over 1441953.55 frames. ], batch size: 66, lr: 5.32e-03, grad_scale: 16.0 +2023-03-21 06:06:52,795 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=86634.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 06:06:54,240 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.252e+02 1.720e+02 1.969e+02 2.412e+02 4.198e+02, threshold=3.938e+02, percent-clipped=1.0 +2023-03-21 06:06:56,430 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86641.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:07:05,024 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86658.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:07:06,542 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=86661.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:07:11,008 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 06:07:11,518 INFO [train.py:901] (0/2) Epoch 31, batch 1950, loss[loss=0.1318, simple_loss=0.2148, pruned_loss=0.02441, over 7349.00 frames. ], tot_loss[loss=0.1362, simple_loss=0.2167, pruned_loss=0.02785, over 1442546.48 frames. ], batch size: 63, lr: 5.31e-03, grad_scale: 16.0 +2023-03-21 06:07:22,653 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 06:07:27,200 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 06:07:27,696 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 06:07:30,448 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=86706.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:07:31,532 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86708.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:07:38,003 INFO [train.py:901] (0/2) Epoch 31, batch 2000, loss[loss=0.1439, simple_loss=0.2246, pruned_loss=0.0316, over 7282.00 frames. ], tot_loss[loss=0.1361, simple_loss=0.2165, pruned_loss=0.02784, over 1441625.65 frames. ], batch size: 86, lr: 5.31e-03, grad_scale: 16.0 +2023-03-21 06:07:40,017 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86725.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:07:43,415 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 06:07:45,923 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.323e+02 1.843e+02 2.081e+02 2.436e+02 4.240e+02, threshold=4.161e+02, percent-clipped=2.0 +2023-03-21 06:07:54,447 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 06:07:55,482 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=86756.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:08:02,022 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 06:08:03,475 INFO [train.py:901] (0/2) Epoch 31, batch 2050, loss[loss=0.1207, simple_loss=0.2042, pruned_loss=0.01864, over 7313.00 frames. ], tot_loss[loss=0.1358, simple_loss=0.2162, pruned_loss=0.02772, over 1443383.80 frames. ], batch size: 44, lr: 5.31e-03, grad_scale: 16.0 +2023-03-21 06:08:04,504 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=86773.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:08:10,286 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.89 vs. limit=5.0 +2023-03-21 06:08:16,782 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=86796.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:08:19,698 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=86801.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 06:08:26,259 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=86814.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:08:29,723 INFO [train.py:901] (0/2) Epoch 31, batch 2100, loss[loss=0.1472, simple_loss=0.2298, pruned_loss=0.03232, over 7242.00 frames. ], tot_loss[loss=0.1359, simple_loss=0.2165, pruned_loss=0.02767, over 1444879.37 frames. ], batch size: 93, lr: 5.31e-03, grad_scale: 16.0 +2023-03-21 06:08:35,864 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 06:08:37,856 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.065e+02 1.745e+02 2.102e+02 2.497e+02 4.075e+02, threshold=4.204e+02, percent-clipped=0.0 +2023-03-21 06:08:38,865 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 06:08:56,245 INFO [train.py:901] (0/2) Epoch 31, batch 2150, loss[loss=0.1319, simple_loss=0.2153, pruned_loss=0.02424, over 7273.00 frames. ], tot_loss[loss=0.1361, simple_loss=0.2168, pruned_loss=0.02774, over 1443276.97 frames. ], batch size: 70, lr: 5.31e-03, grad_scale: 16.0 +2023-03-21 06:08:56,379 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3620, 1.6464, 1.1950, 1.4631, 1.6575, 1.4880, 1.5100, 1.1826], + device='cuda:0'), covar=tensor([0.0152, 0.0114, 0.0271, 0.0178, 0.0100, 0.0138, 0.0101, 0.0151], + device='cuda:0'), in_proj_covar=tensor([0.0033, 0.0030, 0.0030, 0.0031, 0.0030, 0.0028, 0.0031, 0.0040], + device='cuda:0'), out_proj_covar=tensor([3.7182e-05, 3.3435e-05, 3.3731e-05, 3.4261e-05, 3.3283e-05, 3.1409e-05, + 3.5516e-05, 4.5346e-05], device='cuda:0') +2023-03-21 06:08:58,381 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86875.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:08:59,829 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=86878.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:09:11,998 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=86902.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:09:21,498 INFO [train.py:901] (0/2) Epoch 31, batch 2200, loss[loss=0.1319, simple_loss=0.2138, pruned_loss=0.02499, over 7343.00 frames. ], tot_loss[loss=0.1359, simple_loss=0.2165, pruned_loss=0.02768, over 1442680.16 frames. ], batch size: 54, lr: 5.31e-03, grad_scale: 16.0 +2023-03-21 06:09:23,558 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 06:09:29,161 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=86936.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:09:29,580 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.237e+02 1.856e+02 2.260e+02 2.640e+02 6.154e+02, threshold=4.519e+02, percent-clipped=3.0 +2023-03-21 06:09:31,350 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=86939.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:09:36,806 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=86950.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:09:44,768 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-03-21 06:09:47,943 INFO [train.py:901] (0/2) Epoch 31, batch 2250, loss[loss=0.1398, simple_loss=0.2176, pruned_loss=0.03094, over 7371.00 frames. ], tot_loss[loss=0.1359, simple_loss=0.2166, pruned_loss=0.02756, over 1443753.53 frames. ], batch size: 65, lr: 5.30e-03, grad_scale: 16.0 +2023-03-21 06:09:58,623 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 06:09:59,151 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 06:09:59,820 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7942, 2.6150, 2.8736, 2.8896, 2.4138, 2.6150, 2.9873, 2.2004], + device='cuda:0'), covar=tensor([0.0470, 0.0513, 0.0616, 0.0630, 0.0539, 0.0902, 0.0778, 0.1915], + device='cuda:0'), in_proj_covar=tensor([0.0332, 0.0338, 0.0273, 0.0355, 0.0292, 0.0291, 0.0343, 0.0264], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 06:10:04,117 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=87002.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 06:10:11,496 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 06:10:13,526 INFO [train.py:901] (0/2) Epoch 31, batch 2300, loss[loss=0.1047, simple_loss=0.1729, pruned_loss=0.01831, over 5876.00 frames. ], tot_loss[loss=0.1357, simple_loss=0.2164, pruned_loss=0.02749, over 1441373.58 frames. ], batch size: 25, lr: 5.30e-03, grad_scale: 16.0 +2023-03-21 06:10:20,808 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8329, 3.9934, 3.6892, 3.9640, 3.6273, 3.9409, 4.2390, 4.2502], + device='cuda:0'), covar=tensor([0.0271, 0.0206, 0.0285, 0.0206, 0.0413, 0.0351, 0.0305, 0.0244], + device='cuda:0'), in_proj_covar=tensor([0.0118, 0.0121, 0.0114, 0.0118, 0.0110, 0.0100, 0.0096, 0.0094], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 06:10:22,190 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.208e+02 1.765e+02 2.120e+02 2.514e+02 3.861e+02, threshold=4.240e+02, percent-clipped=0.0 +2023-03-21 06:10:36,035 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=87063.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 06:10:39,862 INFO [train.py:901] (0/2) Epoch 31, batch 2350, loss[loss=0.1462, simple_loss=0.2279, pruned_loss=0.03221, over 7282.00 frames. ], tot_loss[loss=0.1358, simple_loss=0.2165, pruned_loss=0.02756, over 1441628.43 frames. ], batch size: 66, lr: 5.30e-03, grad_scale: 16.0 +2023-03-21 06:10:52,322 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=87096.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:10:54,788 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=87101.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 06:10:57,631 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 06:11:03,780 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 06:11:05,270 INFO [train.py:901] (0/2) Epoch 31, batch 2400, loss[loss=0.1399, simple_loss=0.2265, pruned_loss=0.02666, over 7272.00 frames. ], tot_loss[loss=0.1363, simple_loss=0.2172, pruned_loss=0.02772, over 1440631.92 frames. ], batch size: 64, lr: 5.30e-03, grad_scale: 16.0 +2023-03-21 06:11:13,832 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.314e+02 1.833e+02 2.136e+02 2.421e+02 3.611e+02, threshold=4.272e+02, percent-clipped=0.0 +2023-03-21 06:11:14,356 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 06:11:17,483 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 06:11:17,528 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=87144.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:11:20,001 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=87149.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:11:30,747 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=87170.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:11:31,158 INFO [train.py:901] (0/2) Epoch 31, batch 2450, loss[loss=0.1291, simple_loss=0.2181, pruned_loss=0.02005, over 7154.00 frames. ], tot_loss[loss=0.1365, simple_loss=0.2176, pruned_loss=0.02771, over 1442063.27 frames. ], batch size: 98, lr: 5.30e-03, grad_scale: 16.0 +2023-03-21 06:11:34,130 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.65 vs. limit=2.0 +2023-03-21 06:11:43,172 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 06:11:56,828 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.88 vs. limit=2.0 +2023-03-21 06:11:57,941 INFO [train.py:901] (0/2) Epoch 31, batch 2500, loss[loss=0.1383, simple_loss=0.2184, pruned_loss=0.02912, over 7312.00 frames. ], tot_loss[loss=0.1358, simple_loss=0.2168, pruned_loss=0.02736, over 1442156.27 frames. ], batch size: 49, lr: 5.30e-03, grad_scale: 16.0 +2023-03-21 06:12:02,640 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=87230.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:12:04,601 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=87234.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:12:05,602 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=87236.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:12:05,988 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.248e+02 1.655e+02 1.992e+02 2.464e+02 4.160e+02, threshold=3.984e+02, percent-clipped=0.0 +2023-03-21 06:12:09,036 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 06:12:19,779 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3829, 2.9008, 3.2040, 3.1279, 2.7423, 2.8179, 3.2906, 2.5838], + device='cuda:0'), covar=tensor([0.0386, 0.0536, 0.0606, 0.0536, 0.0523, 0.0734, 0.0600, 0.1684], + device='cuda:0'), in_proj_covar=tensor([0.0334, 0.0337, 0.0272, 0.0355, 0.0294, 0.0289, 0.0343, 0.0261], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 06:12:23,126 INFO [train.py:901] (0/2) Epoch 31, batch 2550, loss[loss=0.1078, simple_loss=0.1809, pruned_loss=0.01729, over 7049.00 frames. ], tot_loss[loss=0.1363, simple_loss=0.2171, pruned_loss=0.02768, over 1441741.21 frames. ], batch size: 35, lr: 5.30e-03, grad_scale: 16.0 +2023-03-21 06:12:29,718 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=87284.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:12:33,926 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=87291.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:12:49,293 INFO [train.py:901] (0/2) Epoch 31, batch 2600, loss[loss=0.1478, simple_loss=0.2321, pruned_loss=0.03178, over 7268.00 frames. ], tot_loss[loss=0.1361, simple_loss=0.2167, pruned_loss=0.0277, over 1441237.23 frames. ], batch size: 55, lr: 5.29e-03, grad_scale: 16.0 +2023-03-21 06:12:57,248 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.274e+02 1.776e+02 2.100e+02 2.547e+02 3.881e+02, threshold=4.201e+02, percent-clipped=0.0 +2023-03-21 06:13:01,059 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.86 vs. limit=2.0 +2023-03-21 06:13:07,897 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=87358.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 06:13:09,918 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=87362.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:13:14,186 INFO [train.py:901] (0/2) Epoch 31, batch 2650, loss[loss=0.1226, simple_loss=0.2078, pruned_loss=0.01867, over 7243.00 frames. ], tot_loss[loss=0.1361, simple_loss=0.2168, pruned_loss=0.02768, over 1442803.20 frames. ], batch size: 47, lr: 5.29e-03, grad_scale: 16.0 +2023-03-21 06:13:39,055 INFO [train.py:901] (0/2) Epoch 31, batch 2700, loss[loss=0.1374, simple_loss=0.223, pruned_loss=0.02585, over 7276.00 frames. ], tot_loss[loss=0.1366, simple_loss=0.2171, pruned_loss=0.02806, over 1441585.16 frames. ], batch size: 77, lr: 5.29e-03, grad_scale: 16.0 +2023-03-21 06:13:40,193 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=87423.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:13:46,825 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 9.848e+01 1.823e+02 2.135e+02 2.461e+02 5.467e+02, threshold=4.269e+02, percent-clipped=3.0 +2023-03-21 06:13:58,778 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=87461.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:14:03,344 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=87470.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:14:03,721 INFO [train.py:901] (0/2) Epoch 31, batch 2750, loss[loss=0.1426, simple_loss=0.221, pruned_loss=0.03208, over 7305.00 frames. ], tot_loss[loss=0.1372, simple_loss=0.2179, pruned_loss=0.02828, over 1442440.43 frames. ], batch size: 80, lr: 5.29e-03, grad_scale: 16.0 +2023-03-21 06:14:17,841 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.63 vs. limit=2.0 +2023-03-21 06:14:20,518 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=87505.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:14:26,764 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=87518.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:14:28,556 INFO [train.py:901] (0/2) Epoch 31, batch 2800, loss[loss=0.1302, simple_loss=0.207, pruned_loss=0.0267, over 7270.00 frames. ], tot_loss[loss=0.1376, simple_loss=0.218, pruned_loss=0.0286, over 1441711.53 frames. ], batch size: 47, lr: 5.29e-03, grad_scale: 16.0 +2023-03-21 06:14:29,195 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=87522.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:14:34,979 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=87534.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:14:35,414 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.5876, 5.1159, 4.9212, 5.5198, 5.3650, 5.4933, 4.9022, 5.1882], + device='cuda:0'), covar=tensor([0.0741, 0.2487, 0.2376, 0.0919, 0.0912, 0.1175, 0.0665, 0.0922], + device='cuda:0'), in_proj_covar=tensor([0.0241, 0.0377, 0.0284, 0.0291, 0.0215, 0.0357, 0.0216, 0.0261], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 06:14:36,278 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.300e+02 1.877e+02 2.166e+02 2.662e+02 4.595e+02, threshold=4.333e+02, percent-clipped=1.0 +2023-03-21 06:14:41,527 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-31.pt +2023-03-21 06:14:57,410 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 06:14:57,464 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 06:14:57,490 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 06:14:57,506 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 06:14:57,684 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 06:14:57,700 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 06:14:57,711 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 06:14:57,813 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 06:14:57,888 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 06:14:57,903 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 06:14:57,921 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 06:14:58,431 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 06:14:58,438 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 06:14:58,446 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 06:14:58,590 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 06:14:58,593 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 06:14:58,669 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 06:15:00,755 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5132, 2.9098, 3.3975, 3.5875, 3.5362, 3.5079, 3.2349, 3.4129], + device='cuda:0'), covar=tensor([0.0028, 0.0114, 0.0035, 0.0028, 0.0030, 0.0032, 0.0080, 0.0052], + device='cuda:0'), in_proj_covar=tensor([0.0046, 0.0066, 0.0055, 0.0054, 0.0053, 0.0059, 0.0048, 0.0072], + device='cuda:0'), out_proj_covar=tensor([8.1066e-05, 1.4148e-04, 1.0445e-04, 9.6330e-05, 9.4387e-05, 1.0817e-04, + 9.7628e-05, 1.3990e-04], device='cuda:0') +2023-03-21 06:15:01,127 INFO [train.py:901] (0/2) Epoch 32, batch 0, loss[loss=0.1741, simple_loss=0.2488, pruned_loss=0.04973, over 6720.00 frames. ], tot_loss[loss=0.1741, simple_loss=0.2488, pruned_loss=0.04973, over 6720.00 frames. ], batch size: 106, lr: 5.20e-03, grad_scale: 16.0 +2023-03-21 06:15:01,128 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 06:15:17,757 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9768, 3.6702, 3.6845, 3.6441, 3.6930, 3.5768, 3.8092, 3.3619], + device='cuda:0'), covar=tensor([0.0127, 0.0174, 0.0116, 0.0187, 0.0414, 0.0116, 0.0163, 0.0268], + device='cuda:0'), in_proj_covar=tensor([0.0090, 0.0098, 0.0096, 0.0084, 0.0169, 0.0104, 0.0102, 0.0107], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 06:15:27,537 INFO [train.py:935] (0/2) Epoch 32, validation: loss=0.1652, simple_loss=0.2562, pruned_loss=0.03713, over 1622729.00 frames. +2023-03-21 06:15:27,537 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 06:15:33,130 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 06:15:38,147 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=87566.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 06:15:41,224 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9302, 2.5555, 1.8680, 3.0810, 2.8045, 3.0660, 2.5164, 2.7704], + device='cuda:0'), covar=tensor([0.2361, 0.1051, 0.3770, 0.0885, 0.0276, 0.0387, 0.0379, 0.0504], + device='cuda:0'), in_proj_covar=tensor([0.0256, 0.0233, 0.0252, 0.0258, 0.0189, 0.0190, 0.0212, 0.0223], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 06:15:43,608 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 06:15:44,404 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.79 vs. limit=2.0 +2023-03-21 06:15:46,181 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=87582.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:15:46,807 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0717, 3.3321, 2.9617, 3.1890, 3.1287, 2.7078, 3.2840, 3.1274], + device='cuda:0'), covar=tensor([0.0914, 0.1049, 0.1173, 0.1556, 0.1965, 0.1057, 0.1168, 0.1100], + device='cuda:0'), in_proj_covar=tensor([0.0056, 0.0056, 0.0064, 0.0056, 0.0054, 0.0058, 0.0056, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 06:15:47,338 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=87584.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:15:48,245 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=87586.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:15:49,849 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=87589.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:15:50,243 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 06:15:52,803 INFO [train.py:901] (0/2) Epoch 32, batch 50, loss[loss=0.1263, simple_loss=0.2137, pruned_loss=0.01949, over 7323.00 frames. ], tot_loss[loss=0.1381, simple_loss=0.2187, pruned_loss=0.02871, over 328078.60 frames. ], batch size: 83, lr: 5.20e-03, grad_scale: 16.0 +2023-03-21 06:15:52,827 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 06:15:55,752 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 06:15:58,059 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.86 vs. limit=2.0 +2023-03-21 06:15:58,312 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=87605.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:16:04,021 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.44 vs. limit=2.0 +2023-03-21 06:16:06,647 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.85 vs. limit=2.0 +2023-03-21 06:16:09,955 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3793, 1.6066, 1.2522, 1.5540, 1.6936, 1.2046, 1.4959, 1.0247], + device='cuda:0'), covar=tensor([0.0203, 0.0139, 0.0305, 0.0173, 0.0139, 0.0278, 0.0147, 0.0235], + device='cuda:0'), in_proj_covar=tensor([0.0034, 0.0031, 0.0031, 0.0032, 0.0031, 0.0029, 0.0033, 0.0041], + device='cuda:0'), out_proj_covar=tensor([3.8543e-05, 3.4759e-05, 3.4914e-05, 3.5339e-05, 3.4617e-05, 3.2925e-05, + 3.7896e-05, 4.6441e-05], device='cuda:0') +2023-03-21 06:16:14,068 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 06:16:14,543 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 06:16:15,506 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.185e+02 1.723e+02 1.992e+02 2.354e+02 4.826e+02, threshold=3.984e+02, percent-clipped=1.0 +2023-03-21 06:16:19,615 INFO [train.py:901] (0/2) Epoch 32, batch 100, loss[loss=0.1595, simple_loss=0.2354, pruned_loss=0.04176, over 7233.00 frames. ], tot_loss[loss=0.1367, simple_loss=0.2173, pruned_loss=0.02808, over 574534.35 frames. ], batch size: 45, lr: 5.20e-03, grad_scale: 16.0 +2023-03-21 06:16:19,742 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=87645.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 06:16:22,296 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=87650.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:16:24,875 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.80 vs. limit=2.0 +2023-03-21 06:16:26,093 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=87658.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 06:16:30,201 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=87666.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 06:16:40,777 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7065, 2.5039, 2.3536, 3.6773, 1.8655, 3.4260, 1.5701, 3.1040], + device='cuda:0'), covar=tensor([0.0172, 0.1221, 0.1697, 0.0216, 0.3831, 0.0325, 0.1162, 0.0413], + device='cuda:0'), in_proj_covar=tensor([0.0202, 0.0255, 0.0268, 0.0207, 0.0258, 0.0213, 0.0237, 0.0230], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 06:16:43,971 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-03-21 06:16:44,672 INFO [train.py:901] (0/2) Epoch 32, batch 150, loss[loss=0.1416, simple_loss=0.2225, pruned_loss=0.03036, over 7304.00 frames. ], tot_loss[loss=0.136, simple_loss=0.217, pruned_loss=0.02748, over 767242.52 frames. ], batch size: 80, lr: 5.20e-03, grad_scale: 16.0 +2023-03-21 06:16:50,309 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=87706.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 06:16:56,362 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=87718.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:17:07,051 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.194e+02 1.862e+02 2.197e+02 2.502e+02 5.473e+02, threshold=4.393e+02, percent-clipped=2.0 +2023-03-21 06:17:11,196 INFO [train.py:901] (0/2) Epoch 32, batch 200, loss[loss=0.1479, simple_loss=0.2252, pruned_loss=0.03532, over 7321.00 frames. ], tot_loss[loss=0.1364, simple_loss=0.2172, pruned_loss=0.0278, over 917438.69 frames. ], batch size: 61, lr: 5.20e-03, grad_scale: 16.0 +2023-03-21 06:17:16,210 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 06:17:21,883 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 06:17:27,973 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 06:17:36,377 INFO [train.py:901] (0/2) Epoch 32, batch 250, loss[loss=0.1333, simple_loss=0.219, pruned_loss=0.02381, over 7303.00 frames. ], tot_loss[loss=0.1355, simple_loss=0.2161, pruned_loss=0.02748, over 1031651.88 frames. ], batch size: 80, lr: 5.20e-03, grad_scale: 16.0 +2023-03-21 06:17:40,644 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 06:17:41,410 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.29 vs. limit=2.0 +2023-03-21 06:17:48,949 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=87817.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:17:52,235 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.76 vs. limit=2.0 +2023-03-21 06:17:52,989 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6464, 4.1822, 4.0951, 4.6031, 4.5262, 4.5993, 3.9859, 4.2385], + device='cuda:0'), covar=tensor([0.0776, 0.2370, 0.2142, 0.1127, 0.0919, 0.1172, 0.0955, 0.0978], + device='cuda:0'), in_proj_covar=tensor([0.0240, 0.0370, 0.0278, 0.0286, 0.0212, 0.0352, 0.0212, 0.0257], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 06:17:58,939 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.241e+02 1.893e+02 2.142e+02 2.477e+02 4.691e+02, threshold=4.285e+02, percent-clipped=1.0 +2023-03-21 06:18:03,029 INFO [train.py:901] (0/2) Epoch 32, batch 300, loss[loss=0.1366, simple_loss=0.2198, pruned_loss=0.02669, over 7244.00 frames. ], tot_loss[loss=0.135, simple_loss=0.2154, pruned_loss=0.02728, over 1122523.49 frames. ], batch size: 89, lr: 5.20e-03, grad_scale: 16.0 +2023-03-21 06:18:03,100 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9549, 4.4925, 4.3931, 4.9840, 4.8367, 4.9469, 4.4225, 4.5631], + device='cuda:0'), covar=tensor([0.0802, 0.2426, 0.1971, 0.1095, 0.0885, 0.1117, 0.0758, 0.0987], + device='cuda:0'), in_proj_covar=tensor([0.0241, 0.0371, 0.0279, 0.0286, 0.0213, 0.0353, 0.0213, 0.0258], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 06:18:03,534 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 06:18:11,142 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 06:18:11,199 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=87861.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 06:18:14,890 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.95 vs. limit=2.0 +2023-03-21 06:18:23,740 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=87886.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:18:24,979 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.80 vs. limit=2.0 +2023-03-21 06:18:28,652 INFO [train.py:901] (0/2) Epoch 32, batch 350, loss[loss=0.1233, simple_loss=0.206, pruned_loss=0.02027, over 7267.00 frames. ], tot_loss[loss=0.1353, simple_loss=0.216, pruned_loss=0.02727, over 1195776.31 frames. ], batch size: 70, lr: 5.19e-03, grad_scale: 16.0 +2023-03-21 06:18:45,910 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 06:18:48,967 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=87934.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:18:50,421 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.220e+02 1.703e+02 1.990e+02 2.314e+02 3.102e+02, threshold=3.979e+02, percent-clipped=0.0 +2023-03-21 06:18:52,118 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=87940.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 06:18:54,632 INFO [train.py:901] (0/2) Epoch 32, batch 400, loss[loss=0.14, simple_loss=0.224, pruned_loss=0.028, over 7356.00 frames. ], tot_loss[loss=0.1356, simple_loss=0.2167, pruned_loss=0.0272, over 1252179.10 frames. ], batch size: 73, lr: 5.19e-03, grad_scale: 16.0 +2023-03-21 06:18:54,730 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=87945.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:19:02,740 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=87961.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 06:19:21,172 INFO [train.py:901] (0/2) Epoch 32, batch 450, loss[loss=0.1408, simple_loss=0.2306, pruned_loss=0.02553, over 7344.00 frames. ], tot_loss[loss=0.1352, simple_loss=0.2161, pruned_loss=0.02711, over 1296015.65 frames. ], batch size: 54, lr: 5.19e-03, grad_scale: 8.0 +2023-03-21 06:19:24,062 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-88000.pt +2023-03-21 06:19:31,925 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 06:19:32,455 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 06:19:37,041 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=88018.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:19:47,021 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.188e+02 1.770e+02 2.090e+02 2.498e+02 3.741e+02, threshold=4.181e+02, percent-clipped=0.0 +2023-03-21 06:19:50,542 INFO [train.py:901] (0/2) Epoch 32, batch 500, loss[loss=0.1558, simple_loss=0.2387, pruned_loss=0.03639, over 7292.00 frames. ], tot_loss[loss=0.1346, simple_loss=0.215, pruned_loss=0.02707, over 1326438.91 frames. ], batch size: 68, lr: 5.19e-03, grad_scale: 8.0 +2023-03-21 06:19:54,787 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-03-21 06:19:57,713 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5541, 1.9304, 1.6077, 1.6843, 1.8840, 1.5627, 1.6853, 1.3185], + device='cuda:0'), covar=tensor([0.0201, 0.0136, 0.0266, 0.0154, 0.0115, 0.0164, 0.0152, 0.0159], + device='cuda:0'), in_proj_covar=tensor([0.0034, 0.0031, 0.0030, 0.0031, 0.0030, 0.0029, 0.0033, 0.0040], + device='cuda:0'), out_proj_covar=tensor([3.8291e-05, 3.4395e-05, 3.4527e-05, 3.4725e-05, 3.3974e-05, 3.2668e-05, + 3.7629e-05, 4.5204e-05], device='cuda:0') +2023-03-21 06:20:01,195 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=88066.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:20:06,914 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 06:20:08,473 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 06:20:08,989 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 06:20:11,590 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 06:20:16,114 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 06:20:17,095 INFO [train.py:901] (0/2) Epoch 32, batch 550, loss[loss=0.1373, simple_loss=0.2207, pruned_loss=0.02692, over 7282.00 frames. ], tot_loss[loss=0.1352, simple_loss=0.2154, pruned_loss=0.02751, over 1352862.52 frames. ], batch size: 57, lr: 5.19e-03, grad_scale: 8.0 +2023-03-21 06:20:19,742 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=88100.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:20:24,538 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.28 vs. limit=2.0 +2023-03-21 06:20:26,750 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 06:20:28,392 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=88117.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:20:35,848 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 06:20:38,869 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.305e+02 1.728e+02 2.085e+02 2.594e+02 4.584e+02, threshold=4.170e+02, percent-clipped=1.0 +2023-03-21 06:20:38,890 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 06:20:42,464 INFO [train.py:901] (0/2) Epoch 32, batch 600, loss[loss=0.1216, simple_loss=0.2015, pruned_loss=0.02082, over 7339.00 frames. ], tot_loss[loss=0.1348, simple_loss=0.2155, pruned_loss=0.02702, over 1375588.68 frames. ], batch size: 54, lr: 5.19e-03, grad_scale: 8.0 +2023-03-21 06:20:45,378 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 06:20:51,155 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7906, 5.3525, 5.4685, 5.3779, 5.1556, 4.8148, 5.4608, 5.2337], + device='cuda:0'), covar=tensor([0.0442, 0.0326, 0.0243, 0.0390, 0.0357, 0.0339, 0.0252, 0.0418], + device='cuda:0'), in_proj_covar=tensor([0.0215, 0.0249, 0.0189, 0.0195, 0.0151, 0.0222, 0.0196, 0.0145], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 06:20:51,693 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=88161.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 06:20:51,720 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=88161.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:20:53,660 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=88165.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:21:03,183 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 06:21:08,623 INFO [train.py:901] (0/2) Epoch 32, batch 650, loss[loss=0.1397, simple_loss=0.2184, pruned_loss=0.0305, over 7334.00 frames. ], tot_loss[loss=0.1346, simple_loss=0.2153, pruned_loss=0.02694, over 1390249.24 frames. ], batch size: 75, lr: 5.19e-03, grad_scale: 8.0 +2023-03-21 06:21:12,014 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 06:21:16,009 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=88209.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:21:18,362 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-03-21 06:21:29,003 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 06:21:30,501 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.043e+02 1.734e+02 2.072e+02 2.519e+02 4.714e+02, threshold=4.144e+02, percent-clipped=1.0 +2023-03-21 06:21:31,612 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=88240.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:21:34,604 INFO [train.py:901] (0/2) Epoch 32, batch 700, loss[loss=0.1041, simple_loss=0.1848, pruned_loss=0.01174, over 7297.00 frames. ], tot_loss[loss=0.1344, simple_loss=0.215, pruned_loss=0.02692, over 1400184.67 frames. ], batch size: 42, lr: 5.18e-03, grad_scale: 8.0 +2023-03-21 06:21:35,357 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=88245.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:21:39,253 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 06:21:43,452 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=88261.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 06:21:56,776 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=88288.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:21:58,324 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=88291.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 06:21:59,323 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=88293.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:22:00,249 INFO [train.py:901] (0/2) Epoch 32, batch 750, loss[loss=0.1513, simple_loss=0.2254, pruned_loss=0.03861, over 7288.00 frames. ], tot_loss[loss=0.1348, simple_loss=0.2152, pruned_loss=0.02716, over 1404638.83 frames. ], batch size: 68, lr: 5.18e-03, grad_scale: 8.0 +2023-03-21 06:22:02,191 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 06:22:02,706 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 06:22:05,365 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1004, 1.9856, 2.1984, 2.0267, 2.3158, 2.1703, 1.8159, 1.6159], + device='cuda:0'), covar=tensor([0.0340, 0.0436, 0.0231, 0.0222, 0.0239, 0.0282, 0.0348, 0.0235], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0034, 0.0035, 0.0033, 0.0032, 0.0032, 0.0036, 0.0036], + device='cuda:0'), out_proj_covar=tensor([8.8784e-05, 8.8072e-05, 8.6936e-05, 8.3329e-05, 8.3706e-05, 8.3773e-05, + 8.9622e-05, 9.1441e-05], device='cuda:0') +2023-03-21 06:22:07,279 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=88309.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:22:09,863 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4103, 2.4824, 3.1926, 3.4270, 3.4586, 3.4601, 3.2232, 3.1235], + device='cuda:0'), covar=tensor([0.0042, 0.0198, 0.0062, 0.0047, 0.0042, 0.0042, 0.0116, 0.0085], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0068, 0.0055, 0.0054, 0.0053, 0.0059, 0.0049, 0.0073], + device='cuda:0'), out_proj_covar=tensor([8.2233e-05, 1.4425e-04, 1.0463e-04, 9.6703e-05, 9.5072e-05, 1.0716e-04, + 9.8954e-05, 1.4151e-04], device='cuda:0') +2023-03-21 06:22:15,837 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 06:22:21,089 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 06:22:23,015 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.235e+02 1.782e+02 2.097e+02 2.465e+02 5.037e+02, threshold=4.194e+02, percent-clipped=2.0 +2023-03-21 06:22:26,545 INFO [train.py:901] (0/2) Epoch 32, batch 800, loss[loss=0.1334, simple_loss=0.2107, pruned_loss=0.02807, over 7220.00 frames. ], tot_loss[loss=0.1353, simple_loss=0.2158, pruned_loss=0.02738, over 1411913.35 frames. ], batch size: 50, lr: 5.18e-03, grad_scale: 8.0 +2023-03-21 06:22:26,574 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 06:22:28,069 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 06:22:30,179 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=88352.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 06:22:37,467 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 06:22:51,510 INFO [train.py:901] (0/2) Epoch 32, batch 850, loss[loss=0.1376, simple_loss=0.2193, pruned_loss=0.02796, over 7332.00 frames. ], tot_loss[loss=0.1361, simple_loss=0.2167, pruned_loss=0.02777, over 1418801.60 frames. ], batch size: 54, lr: 5.18e-03, grad_scale: 8.0 +2023-03-21 06:22:51,890 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.86 vs. limit=2.0 +2023-03-21 06:22:55,915 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 06:22:55,925 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 06:23:03,194 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 06:23:07,155 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 06:23:14,807 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.241e+02 1.715e+02 2.087e+02 2.362e+02 4.389e+02, threshold=4.174e+02, percent-clipped=1.0 +2023-03-21 06:23:18,454 INFO [train.py:901] (0/2) Epoch 32, batch 900, loss[loss=0.1438, simple_loss=0.2245, pruned_loss=0.03154, over 7201.00 frames. ], tot_loss[loss=0.1355, simple_loss=0.2157, pruned_loss=0.02761, over 1424885.95 frames. ], batch size: 50, lr: 5.18e-03, grad_scale: 8.0 +2023-03-21 06:23:24,126 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=88456.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:23:27,233 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.51 vs. limit=5.0 +2023-03-21 06:23:43,381 INFO [train.py:901] (0/2) Epoch 32, batch 950, loss[loss=0.1348, simple_loss=0.2124, pruned_loss=0.02857, over 7277.00 frames. ], tot_loss[loss=0.1363, simple_loss=0.2169, pruned_loss=0.02787, over 1429166.05 frames. ], batch size: 86, lr: 5.18e-03, grad_scale: 8.0 +2023-03-21 06:23:43,416 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 06:23:55,017 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4618, 3.3034, 2.3927, 3.8831, 2.7755, 3.2690, 1.7604, 2.4212], + device='cuda:0'), covar=tensor([0.0506, 0.0825, 0.2705, 0.0528, 0.0565, 0.0782, 0.4084, 0.1832], + device='cuda:0'), in_proj_covar=tensor([0.0288, 0.0262, 0.0291, 0.0278, 0.0278, 0.0272, 0.0250, 0.0269], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 06:23:58,584 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1951, 2.9807, 2.2310, 3.6206, 2.5320, 2.9996, 1.6165, 2.3082], + device='cuda:0'), covar=tensor([0.0543, 0.0838, 0.2867, 0.0615, 0.0556, 0.0652, 0.4052, 0.1916], + device='cuda:0'), in_proj_covar=tensor([0.0288, 0.0263, 0.0292, 0.0278, 0.0279, 0.0273, 0.0250, 0.0270], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 06:24:06,410 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.126e+02 1.732e+02 1.981e+02 2.527e+02 4.373e+02, threshold=3.962e+02, percent-clipped=2.0 +2023-03-21 06:24:08,382 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 06:24:09,888 INFO [train.py:901] (0/2) Epoch 32, batch 1000, loss[loss=0.137, simple_loss=0.2184, pruned_loss=0.02783, over 7284.00 frames. ], tot_loss[loss=0.1358, simple_loss=0.2162, pruned_loss=0.02766, over 1431997.66 frames. ], batch size: 66, lr: 5.18e-03, grad_scale: 8.0 +2023-03-21 06:24:21,025 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9956, 2.5604, 3.2207, 3.1675, 3.1796, 2.9190, 2.5041, 3.1347], + device='cuda:0'), covar=tensor([0.1573, 0.0806, 0.0990, 0.1120, 0.0636, 0.1032, 0.2519, 0.0995], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0064, 0.0050, 0.0048, 0.0048, 0.0048, 0.0067, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 06:24:28,559 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 06:24:36,350 INFO [train.py:901] (0/2) Epoch 32, batch 1050, loss[loss=0.1139, simple_loss=0.1987, pruned_loss=0.01453, over 7340.00 frames. ], tot_loss[loss=0.1353, simple_loss=0.2159, pruned_loss=0.02738, over 1433944.41 frames. ], batch size: 44, lr: 5.17e-03, grad_scale: 8.0 +2023-03-21 06:24:50,707 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 06:24:54,756 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 06:24:58,203 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.073e+02 1.638e+02 1.910e+02 2.335e+02 5.404e+02, threshold=3.819e+02, percent-clipped=3.0 +2023-03-21 06:25:00,924 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0243, 3.9886, 3.3967, 3.5461, 3.0228, 2.3139, 1.9987, 3.9935], + device='cuda:0'), covar=tensor([0.0046, 0.0052, 0.0113, 0.0075, 0.0148, 0.0484, 0.0573, 0.0049], + device='cuda:0'), in_proj_covar=tensor([0.0106, 0.0086, 0.0105, 0.0090, 0.0119, 0.0127, 0.0125, 0.0099], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 06:25:01,806 INFO [train.py:901] (0/2) Epoch 32, batch 1100, loss[loss=0.162, simple_loss=0.2408, pruned_loss=0.04156, over 7157.00 frames. ], tot_loss[loss=0.1354, simple_loss=0.2162, pruned_loss=0.02733, over 1436475.48 frames. ], batch size: 98, lr: 5.17e-03, grad_scale: 8.0 +2023-03-21 06:25:02,919 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=88647.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 06:25:24,179 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 06:25:24,636 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 06:25:26,278 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5635, 2.8950, 2.4685, 2.8794, 2.9151, 2.5811, 2.8778, 2.6521], + device='cuda:0'), covar=tensor([0.0738, 0.0556, 0.1145, 0.1241, 0.0639, 0.0739, 0.0531, 0.0949], + device='cuda:0'), in_proj_covar=tensor([0.0055, 0.0056, 0.0064, 0.0055, 0.0053, 0.0057, 0.0055, 0.0051], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 06:25:28,144 INFO [train.py:901] (0/2) Epoch 32, batch 1150, loss[loss=0.1347, simple_loss=0.2168, pruned_loss=0.02628, over 7262.00 frames. ], tot_loss[loss=0.136, simple_loss=0.2168, pruned_loss=0.02765, over 1438161.70 frames. ], batch size: 52, lr: 5.17e-03, grad_scale: 8.0 +2023-03-21 06:25:36,094 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 06:25:36,585 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 06:25:49,709 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.246e+02 1.764e+02 2.030e+02 2.490e+02 8.067e+02, threshold=4.061e+02, percent-clipped=2.0 +2023-03-21 06:25:53,218 INFO [train.py:901] (0/2) Epoch 32, batch 1200, loss[loss=0.1424, simple_loss=0.2209, pruned_loss=0.03191, over 7263.00 frames. ], tot_loss[loss=0.1359, simple_loss=0.2166, pruned_loss=0.02757, over 1438461.90 frames. ], batch size: 47, lr: 5.17e-03, grad_scale: 8.0 +2023-03-21 06:26:00,108 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=88756.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:26:02,403 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.02 vs. limit=5.0 +2023-03-21 06:26:09,532 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 06:26:10,848 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.70 vs. limit=2.0 +2023-03-21 06:26:19,487 INFO [train.py:901] (0/2) Epoch 32, batch 1250, loss[loss=0.09963, simple_loss=0.1679, pruned_loss=0.01566, over 7044.00 frames. ], tot_loss[loss=0.1357, simple_loss=0.2163, pruned_loss=0.02752, over 1439930.90 frames. ], batch size: 35, lr: 5.17e-03, grad_scale: 8.0 +2023-03-21 06:26:19,834 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-03-21 06:26:21,867 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-03-21 06:26:24,499 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=88804.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:26:29,648 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9362, 2.8183, 2.2593, 3.5789, 2.3283, 2.9028, 1.5897, 2.3166], + device='cuda:0'), covar=tensor([0.0414, 0.0902, 0.2427, 0.0594, 0.0498, 0.0465, 0.3701, 0.1894], + device='cuda:0'), in_proj_covar=tensor([0.0286, 0.0262, 0.0290, 0.0277, 0.0275, 0.0271, 0.0248, 0.0268], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 06:26:33,432 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 06:26:37,430 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 06:26:38,444 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 06:26:41,995 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.302e+02 1.883e+02 2.188e+02 2.515e+02 4.579e+02, threshold=4.376e+02, percent-clipped=2.0 +2023-03-21 06:26:46,157 INFO [train.py:901] (0/2) Epoch 32, batch 1300, loss[loss=0.1536, simple_loss=0.2311, pruned_loss=0.03799, over 7273.00 frames. ], tot_loss[loss=0.1355, simple_loss=0.2162, pruned_loss=0.02739, over 1438667.09 frames. ], batch size: 57, lr: 5.17e-03, grad_scale: 8.0 +2023-03-21 06:26:57,194 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6354, 1.3625, 1.7983, 2.1365, 1.8116, 1.9996, 1.3600, 1.9666], + device='cuda:0'), covar=tensor([0.1401, 0.4501, 0.1130, 0.0969, 0.1689, 0.1899, 0.1764, 0.3102], + device='cuda:0'), in_proj_covar=tensor([0.0071, 0.0073, 0.0062, 0.0060, 0.0058, 0.0059, 0.0095, 0.0061], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 06:27:01,568 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 06:27:03,585 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 06:27:06,586 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 06:27:07,206 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2556, 4.1514, 3.5489, 3.7574, 3.3678, 2.3446, 1.9431, 4.2881], + device='cuda:0'), covar=tensor([0.0043, 0.0076, 0.0111, 0.0061, 0.0128, 0.0512, 0.0596, 0.0043], + device='cuda:0'), in_proj_covar=tensor([0.0107, 0.0087, 0.0106, 0.0090, 0.0121, 0.0129, 0.0127, 0.0099], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 06:27:11,021 INFO [train.py:901] (0/2) Epoch 32, batch 1350, loss[loss=0.1253, simple_loss=0.2121, pruned_loss=0.01928, over 7272.00 frames. ], tot_loss[loss=0.1361, simple_loss=0.217, pruned_loss=0.02763, over 1439776.39 frames. ], batch size: 70, lr: 5.17e-03, grad_scale: 8.0 +2023-03-21 06:27:14,457 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.92 vs. limit=2.0 +2023-03-21 06:27:16,100 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 06:27:33,607 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.258e+02 1.816e+02 2.111e+02 2.445e+02 5.910e+02, threshold=4.221e+02, percent-clipped=1.0 +2023-03-21 06:27:37,103 INFO [train.py:901] (0/2) Epoch 32, batch 1400, loss[loss=0.125, simple_loss=0.2085, pruned_loss=0.02072, over 7146.00 frames. ], tot_loss[loss=0.1358, simple_loss=0.2165, pruned_loss=0.02753, over 1438548.56 frames. ], batch size: 41, lr: 5.16e-03, grad_scale: 8.0 +2023-03-21 06:27:38,230 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=88947.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 06:27:50,053 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 06:27:50,850 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.50 vs. limit=5.0 +2023-03-21 06:27:54,272 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4621, 2.7839, 2.4721, 3.7913, 1.9590, 3.5110, 1.6637, 3.0903], + device='cuda:0'), covar=tensor([0.0159, 0.1287, 0.1845, 0.0138, 0.3970, 0.0272, 0.1201, 0.0479], + device='cuda:0'), in_proj_covar=tensor([0.0201, 0.0254, 0.0266, 0.0204, 0.0256, 0.0212, 0.0236, 0.0230], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 06:28:02,051 INFO [train.py:901] (0/2) Epoch 32, batch 1450, loss[loss=0.1272, simple_loss=0.2101, pruned_loss=0.0221, over 7335.00 frames. ], tot_loss[loss=0.1357, simple_loss=0.2162, pruned_loss=0.02758, over 1438194.70 frames. ], batch size: 44, lr: 5.16e-03, grad_scale: 8.0 +2023-03-21 06:28:02,108 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=88995.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 06:28:12,641 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3287, 2.9936, 2.1701, 3.4998, 3.2522, 3.2879, 2.8562, 3.1368], + device='cuda:0'), covar=tensor([0.2070, 0.0857, 0.3534, 0.0547, 0.0318, 0.0230, 0.0379, 0.0513], + device='cuda:0'), in_proj_covar=tensor([0.0259, 0.0237, 0.0256, 0.0263, 0.0192, 0.0191, 0.0213, 0.0225], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 06:28:16,136 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 06:28:19,783 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=89027.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:28:25,190 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.220e+02 1.675e+02 2.024e+02 2.357e+02 4.449e+02, threshold=4.048e+02, percent-clipped=2.0 +2023-03-21 06:28:28,795 INFO [train.py:901] (0/2) Epoch 32, batch 1500, loss[loss=0.1377, simple_loss=0.2303, pruned_loss=0.02257, over 7212.00 frames. ], tot_loss[loss=0.1359, simple_loss=0.2166, pruned_loss=0.02762, over 1440576.03 frames. ], batch size: 93, lr: 5.16e-03, grad_scale: 8.0 +2023-03-21 06:28:31,803 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 06:28:33,976 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5586, 2.6757, 2.4804, 3.7831, 1.9193, 3.5549, 1.5381, 3.0666], + device='cuda:0'), covar=tensor([0.0166, 0.1260, 0.1662, 0.0151, 0.3947, 0.0254, 0.1193, 0.0425], + device='cuda:0'), in_proj_covar=tensor([0.0201, 0.0253, 0.0266, 0.0204, 0.0255, 0.0211, 0.0235, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 06:28:50,416 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=89088.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:28:53,812 INFO [train.py:901] (0/2) Epoch 32, batch 1550, loss[loss=0.1217, simple_loss=0.2105, pruned_loss=0.01642, over 7335.00 frames. ], tot_loss[loss=0.136, simple_loss=0.2166, pruned_loss=0.02774, over 1440462.42 frames. ], batch size: 75, lr: 5.16e-03, grad_scale: 8.0 +2023-03-21 06:28:55,515 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 06:29:16,717 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.184e+02 1.776e+02 2.055e+02 2.401e+02 3.292e+02, threshold=4.110e+02, percent-clipped=0.0 +2023-03-21 06:29:20,245 INFO [train.py:901] (0/2) Epoch 32, batch 1600, loss[loss=0.126, simple_loss=0.2126, pruned_loss=0.01973, over 7301.00 frames. ], tot_loss[loss=0.1357, simple_loss=0.2167, pruned_loss=0.02739, over 1443578.35 frames. ], batch size: 86, lr: 5.16e-03, grad_scale: 8.0 +2023-03-21 06:29:26,326 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 06:29:27,383 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 06:29:30,455 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 06:29:32,607 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=89169.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:29:39,717 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 06:29:44,337 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 06:29:46,767 INFO [train.py:901] (0/2) Epoch 32, batch 1650, loss[loss=0.1368, simple_loss=0.212, pruned_loss=0.03075, over 7211.00 frames. ], tot_loss[loss=0.1365, simple_loss=0.2172, pruned_loss=0.02789, over 1444656.69 frames. ], batch size: 45, lr: 5.16e-03, grad_scale: 8.0 +2023-03-21 06:29:53,109 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 06:30:04,901 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=89230.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:30:07,599 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.50 vs. limit=5.0 +2023-03-21 06:30:08,768 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.241e+02 1.670e+02 2.019e+02 2.409e+02 3.267e+02, threshold=4.038e+02, percent-clipped=0.0 +2023-03-21 06:30:09,780 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 06:30:12,220 INFO [train.py:901] (0/2) Epoch 32, batch 1700, loss[loss=0.134, simple_loss=0.2115, pruned_loss=0.0282, over 7349.00 frames. ], tot_loss[loss=0.136, simple_loss=0.2166, pruned_loss=0.0277, over 1443129.67 frames. ], batch size: 51, lr: 5.15e-03, grad_scale: 8.0 +2023-03-21 06:30:13,693 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 06:30:24,758 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 06:30:29,496 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7837, 5.2905, 5.3290, 5.2775, 5.0613, 4.8060, 5.3768, 5.1653], + device='cuda:0'), covar=tensor([0.0417, 0.0324, 0.0343, 0.0432, 0.0334, 0.0305, 0.0277, 0.0417], + device='cuda:0'), in_proj_covar=tensor([0.0214, 0.0249, 0.0189, 0.0193, 0.0151, 0.0223, 0.0196, 0.0147], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 06:30:32,510 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1727, 2.7693, 1.9939, 3.1541, 3.0626, 2.6431, 2.4267, 2.8035], + device='cuda:0'), covar=tensor([0.1833, 0.0939, 0.3606, 0.0562, 0.0290, 0.0256, 0.0352, 0.0417], + device='cuda:0'), in_proj_covar=tensor([0.0257, 0.0237, 0.0253, 0.0261, 0.0192, 0.0190, 0.0213, 0.0224], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 06:30:34,518 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8023, 1.4549, 2.0144, 2.3039, 2.0833, 2.1128, 1.9423, 2.2174], + device='cuda:0'), covar=tensor([0.2036, 0.4219, 0.1354, 0.1132, 0.1952, 0.3863, 0.1630, 0.3574], + device='cuda:0'), in_proj_covar=tensor([0.0073, 0.0075, 0.0064, 0.0061, 0.0059, 0.0060, 0.0097, 0.0062], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 06:30:38,375 INFO [train.py:901] (0/2) Epoch 32, batch 1750, loss[loss=0.1268, simple_loss=0.2107, pruned_loss=0.02145, over 7302.00 frames. ], tot_loss[loss=0.1353, simple_loss=0.2156, pruned_loss=0.02754, over 1443241.02 frames. ], batch size: 83, lr: 5.15e-03, grad_scale: 8.0 +2023-03-21 06:30:43,518 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6408, 2.9114, 2.4910, 2.6271, 2.8283, 2.4967, 2.6081, 2.6940], + device='cuda:0'), covar=tensor([0.0930, 0.0812, 0.0945, 0.1296, 0.1217, 0.0644, 0.1227, 0.0961], + device='cuda:0'), in_proj_covar=tensor([0.0055, 0.0055, 0.0064, 0.0055, 0.0052, 0.0057, 0.0054, 0.0051], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 06:30:50,335 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 06:30:51,387 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 06:30:51,496 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8371, 3.8343, 3.1360, 3.4601, 2.7410, 2.2494, 1.7599, 3.8561], + device='cuda:0'), covar=tensor([0.0050, 0.0056, 0.0129, 0.0067, 0.0187, 0.0543, 0.0666, 0.0052], + device='cuda:0'), in_proj_covar=tensor([0.0106, 0.0086, 0.0106, 0.0089, 0.0120, 0.0129, 0.0126, 0.0099], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 06:30:59,528 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5384, 1.2913, 1.6410, 2.0103, 1.7110, 1.8216, 1.5302, 1.8910], + device='cuda:0'), covar=tensor([0.1630, 0.4167, 0.1754, 0.1045, 0.1630, 0.2736, 0.2141, 0.2400], + device='cuda:0'), in_proj_covar=tensor([0.0073, 0.0075, 0.0064, 0.0061, 0.0059, 0.0060, 0.0097, 0.0062], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 06:30:59,849 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.271e+02 1.754e+02 1.965e+02 2.362e+02 4.185e+02, threshold=3.929e+02, percent-clipped=1.0 +2023-03-21 06:31:03,401 INFO [train.py:901] (0/2) Epoch 32, batch 1800, loss[loss=0.1312, simple_loss=0.2193, pruned_loss=0.02156, over 7363.00 frames. ], tot_loss[loss=0.1359, simple_loss=0.2163, pruned_loss=0.02779, over 1443440.86 frames. ], batch size: 63, lr: 5.15e-03, grad_scale: 8.0 +2023-03-21 06:31:11,906 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 06:31:19,185 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=89374.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:31:23,736 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=89383.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:31:25,225 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 06:31:26,602 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.70 vs. limit=2.0 +2023-03-21 06:31:29,747 INFO [train.py:901] (0/2) Epoch 32, batch 1850, loss[loss=0.1276, simple_loss=0.2061, pruned_loss=0.02455, over 7262.00 frames. ], tot_loss[loss=0.1355, simple_loss=0.2161, pruned_loss=0.02748, over 1443655.15 frames. ], batch size: 64, lr: 5.15e-03, grad_scale: 8.0 +2023-03-21 06:31:35,128 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 06:31:37,206 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1865, 2.5730, 3.1493, 3.1076, 3.0995, 2.9144, 2.4920, 3.1257], + device='cuda:0'), covar=tensor([0.1079, 0.1213, 0.1110, 0.1193, 0.0807, 0.1244, 0.2140, 0.1248], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0064, 0.0050, 0.0048, 0.0048, 0.0047, 0.0066, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 06:31:45,115 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=89425.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:31:50,107 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=89435.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:31:51,498 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.222e+02 1.843e+02 2.192e+02 2.573e+02 5.735e+02, threshold=4.384e+02, percent-clipped=2.0 +2023-03-21 06:31:51,543 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 06:31:52,097 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=89439.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 06:31:55,573 INFO [train.py:901] (0/2) Epoch 32, batch 1900, loss[loss=0.1502, simple_loss=0.229, pruned_loss=0.03569, over 7354.00 frames. ], tot_loss[loss=0.1356, simple_loss=0.2165, pruned_loss=0.0274, over 1446473.05 frames. ], batch size: 63, lr: 5.15e-03, grad_scale: 8.0 +2023-03-21 06:32:17,062 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=89486.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:32:17,880 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 06:32:21,392 INFO [train.py:901] (0/2) Epoch 32, batch 1950, loss[loss=0.1073, simple_loss=0.1708, pruned_loss=0.02187, over 6964.00 frames. ], tot_loss[loss=0.1354, simple_loss=0.2161, pruned_loss=0.02739, over 1444682.22 frames. ], batch size: 35, lr: 5.15e-03, grad_scale: 8.0 +2023-03-21 06:32:24,010 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=89500.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 06:32:28,456 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=89509.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:32:28,820 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 06:32:32,483 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9570, 2.8027, 2.9739, 3.0126, 2.6865, 2.6616, 3.0817, 2.3127], + device='cuda:0'), covar=tensor([0.0540, 0.0676, 0.0611, 0.0661, 0.0638, 0.0878, 0.0647, 0.1890], + device='cuda:0'), in_proj_covar=tensor([0.0333, 0.0335, 0.0273, 0.0358, 0.0290, 0.0292, 0.0345, 0.0258], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 06:32:33,298 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 06:32:33,776 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 06:32:33,858 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6451, 5.1943, 5.2657, 5.1932, 4.9705, 4.7464, 5.2894, 5.0940], + device='cuda:0'), covar=tensor([0.0437, 0.0375, 0.0352, 0.0453, 0.0341, 0.0335, 0.0282, 0.0428], + device='cuda:0'), in_proj_covar=tensor([0.0214, 0.0252, 0.0189, 0.0194, 0.0151, 0.0224, 0.0198, 0.0149], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 06:32:36,987 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=89525.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:32:44,153 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.288e+02 1.741e+02 2.051e+02 2.439e+02 5.519e+02, threshold=4.102e+02, percent-clipped=1.0 +2023-03-21 06:32:47,693 INFO [train.py:901] (0/2) Epoch 32, batch 2000, loss[loss=0.1312, simple_loss=0.1976, pruned_loss=0.0324, over 6976.00 frames. ], tot_loss[loss=0.1352, simple_loss=0.2156, pruned_loss=0.02734, over 1443704.83 frames. ], batch size: 35, lr: 5.15e-03, grad_scale: 8.0 +2023-03-21 06:32:51,687 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 06:33:00,348 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=89570.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:33:02,725 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 06:33:09,635 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 06:33:12,629 INFO [train.py:901] (0/2) Epoch 32, batch 2050, loss[loss=0.1086, simple_loss=0.1811, pruned_loss=0.01802, over 7074.00 frames. ], tot_loss[loss=0.1352, simple_loss=0.2155, pruned_loss=0.02745, over 1440411.56 frames. ], batch size: 35, lr: 5.14e-03, grad_scale: 8.0 +2023-03-21 06:33:14,138 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.54 vs. limit=2.0 +2023-03-21 06:33:35,812 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.113e+02 1.785e+02 2.119e+02 2.528e+02 6.748e+02, threshold=4.239e+02, percent-clipped=3.0 +2023-03-21 06:33:39,325 INFO [train.py:901] (0/2) Epoch 32, batch 2100, loss[loss=0.1267, simple_loss=0.2161, pruned_loss=0.01868, over 7350.00 frames. ], tot_loss[loss=0.1355, simple_loss=0.2161, pruned_loss=0.02743, over 1441333.02 frames. ], batch size: 63, lr: 5.14e-03, grad_scale: 8.0 +2023-03-21 06:33:45,880 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 06:33:48,368 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 06:33:57,171 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=89680.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:33:58,640 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=89683.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:34:04,479 INFO [train.py:901] (0/2) Epoch 32, batch 2150, loss[loss=0.1209, simple_loss=0.1993, pruned_loss=0.02127, over 7312.00 frames. ], tot_loss[loss=0.135, simple_loss=0.2154, pruned_loss=0.02728, over 1437782.80 frames. ], batch size: 44, lr: 5.14e-03, grad_scale: 8.0 +2023-03-21 06:34:09,759 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6005, 1.4131, 1.8485, 2.1981, 1.9439, 2.1405, 1.8507, 2.0975], + device='cuda:0'), covar=tensor([0.2631, 0.5428, 0.2235, 0.1999, 0.2360, 0.2280, 0.2123, 0.2169], + device='cuda:0'), in_proj_covar=tensor([0.0074, 0.0076, 0.0064, 0.0060, 0.0059, 0.0061, 0.0099, 0.0063], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 06:34:14,783 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.2773, 4.8526, 4.6473, 5.2488, 5.0320, 5.2201, 4.7493, 4.9153], + device='cuda:0'), covar=tensor([0.0703, 0.2284, 0.2183, 0.0913, 0.0810, 0.1025, 0.0677, 0.0872], + device='cuda:0'), in_proj_covar=tensor([0.0248, 0.0384, 0.0291, 0.0298, 0.0223, 0.0361, 0.0218, 0.0265], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 06:34:23,157 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=89730.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:34:23,642 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=89731.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:34:27,154 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.261e+02 1.892e+02 2.197e+02 2.578e+02 4.314e+02, threshold=4.394e+02, percent-clipped=1.0 +2023-03-21 06:34:28,832 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=89741.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:34:30,659 INFO [train.py:901] (0/2) Epoch 32, batch 2200, loss[loss=0.1139, simple_loss=0.186, pruned_loss=0.0209, over 7044.00 frames. ], tot_loss[loss=0.1356, simple_loss=0.2162, pruned_loss=0.02755, over 1439475.76 frames. ], batch size: 35, lr: 5.14e-03, grad_scale: 8.0 +2023-03-21 06:34:33,547 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 06:34:48,495 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=89781.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:34:49,252 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.23 vs. limit=2.0 +2023-03-21 06:34:55,975 INFO [train.py:901] (0/2) Epoch 32, batch 2250, loss[loss=0.1552, simple_loss=0.2277, pruned_loss=0.0413, over 7243.00 frames. ], tot_loss[loss=0.1358, simple_loss=0.2162, pruned_loss=0.0277, over 1441128.24 frames. ], batch size: 89, lr: 5.14e-03, grad_scale: 8.0 +2023-03-21 06:34:56,043 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=89795.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 06:35:00,720 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.85 vs. limit=2.0 +2023-03-21 06:35:06,365 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.75 vs. limit=5.0 +2023-03-21 06:35:07,481 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 06:35:07,990 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 06:35:12,107 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=89825.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:35:13,708 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3548, 3.4363, 2.5577, 3.7872, 3.0115, 3.3914, 1.7629, 2.6805], + device='cuda:0'), covar=tensor([0.0401, 0.0543, 0.2308, 0.0489, 0.0494, 0.0563, 0.3345, 0.1534], + device='cuda:0'), in_proj_covar=tensor([0.0282, 0.0259, 0.0290, 0.0274, 0.0273, 0.0271, 0.0247, 0.0269], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 06:35:18,458 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.217e+02 1.744e+02 2.009e+02 2.399e+02 4.043e+02, threshold=4.017e+02, percent-clipped=0.0 +2023-03-21 06:35:19,981 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 06:35:20,078 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4469, 3.5137, 2.7764, 3.2531, 2.3689, 2.1777, 1.7709, 3.5300], + device='cuda:0'), covar=tensor([0.0090, 0.0072, 0.0239, 0.0091, 0.0313, 0.0668, 0.0737, 0.0083], + device='cuda:0'), in_proj_covar=tensor([0.0108, 0.0087, 0.0108, 0.0091, 0.0123, 0.0131, 0.0128, 0.0100], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 06:35:21,897 INFO [train.py:901] (0/2) Epoch 32, batch 2300, loss[loss=0.1412, simple_loss=0.2186, pruned_loss=0.0319, over 7288.00 frames. ], tot_loss[loss=0.1357, simple_loss=0.2165, pruned_loss=0.02744, over 1441464.79 frames. ], batch size: 66, lr: 5.14e-03, grad_scale: 8.0 +2023-03-21 06:35:31,980 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=89865.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:35:32,028 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=89865.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:35:36,754 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=89873.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:35:48,383 INFO [train.py:901] (0/2) Epoch 32, batch 2350, loss[loss=0.1319, simple_loss=0.2115, pruned_loss=0.02614, over 7310.00 frames. ], tot_loss[loss=0.1354, simple_loss=0.2163, pruned_loss=0.02731, over 1443131.84 frames. ], batch size: 80, lr: 5.14e-03, grad_scale: 8.0 +2023-03-21 06:36:04,000 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=89926.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 06:36:06,443 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 06:36:09,752 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.391e+02 1.845e+02 2.244e+02 2.605e+02 4.330e+02, threshold=4.488e+02, percent-clipped=2.0 +2023-03-21 06:36:12,834 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 06:36:13,308 INFO [train.py:901] (0/2) Epoch 32, batch 2400, loss[loss=0.1357, simple_loss=0.2204, pruned_loss=0.02554, over 7241.00 frames. ], tot_loss[loss=0.1357, simple_loss=0.2159, pruned_loss=0.02771, over 1441502.26 frames. ], batch size: 93, lr: 5.13e-03, grad_scale: 8.0 +2023-03-21 06:36:23,487 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 06:36:26,444 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 06:36:35,937 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.69 vs. limit=2.0 +2023-03-21 06:36:39,645 INFO [train.py:901] (0/2) Epoch 32, batch 2450, loss[loss=0.1395, simple_loss=0.2258, pruned_loss=0.02658, over 7337.00 frames. ], tot_loss[loss=0.1346, simple_loss=0.2147, pruned_loss=0.02721, over 1438213.58 frames. ], batch size: 54, lr: 5.13e-03, grad_scale: 16.0 +2023-03-21 06:36:42,677 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9694, 2.5634, 3.1544, 2.8184, 2.8610, 2.8662, 3.1211, 2.5085], + device='cuda:0'), covar=tensor([0.0280, 0.0244, 0.0611, 0.0352, 0.0507, 0.0669, 0.0408, 0.1588], + device='cuda:0'), in_proj_covar=tensor([0.0334, 0.0338, 0.0274, 0.0360, 0.0293, 0.0293, 0.0346, 0.0260], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 06:36:43,650 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2591, 3.3083, 2.5703, 3.8837, 3.0310, 3.4153, 1.8680, 2.6350], + device='cuda:0'), covar=tensor([0.0401, 0.0799, 0.2345, 0.0489, 0.0417, 0.0582, 0.3471, 0.1729], + device='cuda:0'), in_proj_covar=tensor([0.0282, 0.0259, 0.0289, 0.0272, 0.0271, 0.0270, 0.0244, 0.0268], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 06:36:53,447 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 06:36:57,613 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=90030.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:37:00,552 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=90036.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:37:01,471 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.113e+02 1.713e+02 2.021e+02 2.446e+02 5.091e+02, threshold=4.041e+02, percent-clipped=1.0 +2023-03-21 06:37:05,592 INFO [train.py:901] (0/2) Epoch 32, batch 2500, loss[loss=0.1259, simple_loss=0.2046, pruned_loss=0.02355, over 7279.00 frames. ], tot_loss[loss=0.1343, simple_loss=0.2148, pruned_loss=0.02691, over 1439916.67 frames. ], batch size: 57, lr: 5.13e-03, grad_scale: 16.0 +2023-03-21 06:37:17,682 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5684, 5.0973, 5.1579, 5.1447, 4.9226, 4.5534, 5.2246, 4.9919], + device='cuda:0'), covar=tensor([0.0500, 0.0416, 0.0419, 0.0478, 0.0346, 0.0378, 0.0333, 0.0442], + device='cuda:0'), in_proj_covar=tensor([0.0216, 0.0253, 0.0191, 0.0194, 0.0152, 0.0225, 0.0200, 0.0148], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 06:37:20,143 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 06:37:22,736 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=90078.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:37:24,323 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=90081.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:37:31,183 INFO [train.py:901] (0/2) Epoch 32, batch 2550, loss[loss=0.1043, simple_loss=0.1793, pruned_loss=0.01465, over 7012.00 frames. ], tot_loss[loss=0.1346, simple_loss=0.215, pruned_loss=0.02711, over 1440081.96 frames. ], batch size: 35, lr: 5.13e-03, grad_scale: 16.0 +2023-03-21 06:37:31,294 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=90095.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 06:37:48,872 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=90129.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:37:49,434 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=90130.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:37:53,265 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.307e+02 1.845e+02 2.164e+02 2.494e+02 4.541e+02, threshold=4.327e+02, percent-clipped=1.0 +2023-03-21 06:37:55,814 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=90143.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 06:37:56,710 INFO [train.py:901] (0/2) Epoch 32, batch 2600, loss[loss=0.1592, simple_loss=0.2399, pruned_loss=0.03921, over 6676.00 frames. ], tot_loss[loss=0.1347, simple_loss=0.2152, pruned_loss=0.02715, over 1440765.87 frames. ], batch size: 107, lr: 5.13e-03, grad_scale: 16.0 +2023-03-21 06:38:06,781 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=90165.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:38:20,077 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=90191.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:38:21,899 INFO [train.py:901] (0/2) Epoch 32, batch 2650, loss[loss=0.1435, simple_loss=0.2282, pruned_loss=0.02943, over 7352.00 frames. ], tot_loss[loss=0.1348, simple_loss=0.2153, pruned_loss=0.02714, over 1441579.36 frames. ], batch size: 73, lr: 5.13e-03, grad_scale: 16.0 +2023-03-21 06:38:31,051 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=90213.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:38:34,925 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=90221.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 06:38:40,351 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=90232.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:38:43,107 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.241e+02 1.716e+02 1.987e+02 2.459e+02 4.305e+02, threshold=3.973e+02, percent-clipped=0.0 +2023-03-21 06:38:46,624 INFO [train.py:901] (0/2) Epoch 32, batch 2700, loss[loss=0.1312, simple_loss=0.2139, pruned_loss=0.02426, over 7274.00 frames. ], tot_loss[loss=0.1347, simple_loss=0.215, pruned_loss=0.02717, over 1440079.79 frames. ], batch size: 77, lr: 5.13e-03, grad_scale: 16.0 +2023-03-21 06:39:10,311 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=90293.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:39:11,151 INFO [train.py:901] (0/2) Epoch 32, batch 2750, loss[loss=0.1144, simple_loss=0.2001, pruned_loss=0.01438, over 7331.00 frames. ], tot_loss[loss=0.1349, simple_loss=0.2154, pruned_loss=0.02715, over 1442981.31 frames. ], batch size: 44, lr: 5.12e-03, grad_scale: 16.0 +2023-03-21 06:39:30,961 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9677, 3.9080, 3.4125, 3.6016, 2.9320, 2.3199, 1.9318, 3.9741], + device='cuda:0'), covar=tensor([0.0052, 0.0068, 0.0133, 0.0065, 0.0181, 0.0556, 0.0624, 0.0057], + device='cuda:0'), in_proj_covar=tensor([0.0107, 0.0086, 0.0106, 0.0091, 0.0122, 0.0130, 0.0126, 0.0100], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 06:39:31,470 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=90336.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:39:32,300 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.279e+02 1.708e+02 1.999e+02 2.519e+02 4.921e+02, threshold=3.998e+02, percent-clipped=2.0 +2023-03-21 06:39:35,816 INFO [train.py:901] (0/2) Epoch 32, batch 2800, loss[loss=0.1308, simple_loss=0.2112, pruned_loss=0.02525, over 7302.00 frames. ], tot_loss[loss=0.1346, simple_loss=0.215, pruned_loss=0.02708, over 1439383.21 frames. ], batch size: 80, lr: 5.12e-03, grad_scale: 16.0 +2023-03-21 06:39:37,365 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6264, 2.8118, 3.6102, 3.5859, 3.5943, 3.6122, 3.6762, 3.5771], + device='cuda:0'), covar=tensor([0.0034, 0.0136, 0.0035, 0.0038, 0.0034, 0.0036, 0.0055, 0.0053], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0067, 0.0055, 0.0054, 0.0053, 0.0059, 0.0048, 0.0073], + device='cuda:0'), out_proj_covar=tensor([8.1340e-05, 1.4102e-04, 1.0298e-04, 9.6068e-05, 9.5177e-05, 1.0659e-04, + 9.6041e-05, 1.3992e-04], device='cuda:0') +2023-03-21 06:39:38,864 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=90351.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:39:48,355 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-32.pt +2023-03-21 06:40:03,071 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 06:40:03,127 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 06:40:03,150 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 06:40:03,169 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 06:40:03,346 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 06:40:03,361 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 06:40:03,373 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 06:40:03,814 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 06:40:03,890 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 06:40:03,905 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 06:40:03,923 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 06:40:04,085 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 06:40:04,091 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 06:40:04,099 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 06:40:04,240 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 06:40:04,243 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 06:40:04,301 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 06:40:06,626 INFO [train.py:901] (0/2) Epoch 33, batch 0, loss[loss=0.1445, simple_loss=0.2295, pruned_loss=0.02977, over 7227.00 frames. ], tot_loss[loss=0.1445, simple_loss=0.2295, pruned_loss=0.02977, over 7227.00 frames. ], batch size: 93, lr: 5.05e-03, grad_scale: 16.0 +2023-03-21 06:40:06,627 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 06:40:23,996 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6886, 3.5491, 3.3014, 4.1439, 3.4922, 3.6269, 2.4657, 3.3672], + device='cuda:0'), covar=tensor([0.0406, 0.0705, 0.1658, 0.0487, 0.0620, 0.0648, 0.2417, 0.1462], + device='cuda:0'), in_proj_covar=tensor([0.0281, 0.0258, 0.0287, 0.0271, 0.0271, 0.0269, 0.0244, 0.0266], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 06:40:31,810 INFO [train.py:935] (0/2) Epoch 33, validation: loss=0.165, simple_loss=0.257, pruned_loss=0.03655, over 1622729.00 frames. +2023-03-21 06:40:31,810 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 06:40:38,931 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 06:40:39,945 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=90384.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:40:49,226 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 06:40:51,816 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7408, 3.6126, 3.4721, 3.4848, 2.9296, 3.2311, 3.5940, 3.1760], + device='cuda:0'), covar=tensor([0.0281, 0.0241, 0.0202, 0.0300, 0.0922, 0.0251, 0.0367, 0.0318], + device='cuda:0'), in_proj_covar=tensor([0.0092, 0.0098, 0.0098, 0.0086, 0.0172, 0.0104, 0.0104, 0.0108], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 06:40:54,934 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=90412.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:40:56,343 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 06:40:58,398 INFO [train.py:901] (0/2) Epoch 33, batch 50, loss[loss=0.1408, simple_loss=0.2293, pruned_loss=0.02617, over 7128.00 frames. ], tot_loss[loss=0.1363, simple_loss=0.2176, pruned_loss=0.02756, over 323128.36 frames. ], batch size: 98, lr: 5.04e-03, grad_scale: 16.0 +2023-03-21 06:40:58,405 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 06:41:00,934 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 06:41:02,560 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3834, 2.3925, 2.3288, 3.6138, 1.7204, 3.5540, 1.3324, 3.2164], + device='cuda:0'), covar=tensor([0.0161, 0.1423, 0.1817, 0.0227, 0.3891, 0.0276, 0.1220, 0.0393], + device='cuda:0'), in_proj_covar=tensor([0.0201, 0.0253, 0.0267, 0.0203, 0.0253, 0.0212, 0.0234, 0.0228], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 06:41:05,984 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1116, 3.9769, 3.5375, 3.6516, 3.1447, 2.3679, 1.8700, 4.0856], + device='cuda:0'), covar=tensor([0.0044, 0.0061, 0.0110, 0.0061, 0.0135, 0.0510, 0.0600, 0.0042], + device='cuda:0'), in_proj_covar=tensor([0.0107, 0.0086, 0.0106, 0.0091, 0.0122, 0.0129, 0.0127, 0.0100], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 06:41:07,838 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.447e+02 1.940e+02 2.295e+02 2.780e+02 4.742e+02, threshold=4.589e+02, percent-clipped=7.0 +2023-03-21 06:41:15,607 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0167, 3.9342, 3.5095, 3.5499, 3.0359, 2.2615, 1.8466, 4.0350], + device='cuda:0'), covar=tensor([0.0053, 0.0055, 0.0116, 0.0073, 0.0153, 0.0536, 0.0627, 0.0048], + device='cuda:0'), in_proj_covar=tensor([0.0107, 0.0086, 0.0106, 0.0091, 0.0122, 0.0129, 0.0127, 0.0100], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 06:41:17,989 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 06:41:18,490 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 06:41:19,080 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=90459.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:41:24,006 INFO [train.py:901] (0/2) Epoch 33, batch 100, loss[loss=0.1546, simple_loss=0.2319, pruned_loss=0.0387, over 7322.00 frames. ], tot_loss[loss=0.1348, simple_loss=0.216, pruned_loss=0.02683, over 569136.16 frames. ], batch size: 54, lr: 5.04e-03, grad_scale: 16.0 +2023-03-21 06:41:27,601 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4737, 1.6438, 1.4817, 1.5412, 1.6577, 1.4603, 1.2712, 1.0801], + device='cuda:0'), covar=tensor([0.0175, 0.0203, 0.0252, 0.0140, 0.0166, 0.0137, 0.0196, 0.0198], + device='cuda:0'), in_proj_covar=tensor([0.0034, 0.0032, 0.0031, 0.0032, 0.0031, 0.0030, 0.0034, 0.0041], + device='cuda:0'), out_proj_covar=tensor([3.8893e-05, 3.5632e-05, 3.5486e-05, 3.5836e-05, 3.4404e-05, 3.3234e-05, + 3.8361e-05, 4.5255e-05], device='cuda:0') +2023-03-21 06:41:30,721 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.34 vs. limit=5.0 +2023-03-21 06:41:32,435 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=90486.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:41:38,203 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=90496.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:41:49,637 INFO [train.py:901] (0/2) Epoch 33, batch 150, loss[loss=0.1291, simple_loss=0.2089, pruned_loss=0.02468, over 7333.00 frames. ], tot_loss[loss=0.1357, simple_loss=0.2171, pruned_loss=0.02717, over 764258.32 frames. ], batch size: 44, lr: 5.04e-03, grad_scale: 16.0 +2023-03-21 06:41:50,329 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=90520.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:41:50,789 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=90521.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 06:41:59,733 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.097e+02 1.776e+02 2.057e+02 2.515e+02 4.094e+02, threshold=4.114e+02, percent-clipped=0.0 +2023-03-21 06:42:09,408 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=90557.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:42:15,307 INFO [train.py:901] (0/2) Epoch 33, batch 200, loss[loss=0.1126, simple_loss=0.1979, pruned_loss=0.01369, over 7269.00 frames. ], tot_loss[loss=0.1345, simple_loss=0.2154, pruned_loss=0.02682, over 913014.19 frames. ], batch size: 52, lr: 5.04e-03, grad_scale: 16.0 +2023-03-21 06:42:15,360 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=90569.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:42:18,227 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 06:42:23,195 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 06:42:25,390 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=90588.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:42:29,307 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 06:42:41,047 INFO [train.py:901] (0/2) Epoch 33, batch 250, loss[loss=0.1414, simple_loss=0.2248, pruned_loss=0.02897, over 7352.00 frames. ], tot_loss[loss=0.1349, simple_loss=0.2153, pruned_loss=0.02722, over 1031168.36 frames. ], batch size: 61, lr: 5.04e-03, grad_scale: 16.0 +2023-03-21 06:42:42,096 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 06:42:50,978 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.298e+02 1.739e+02 2.002e+02 2.238e+02 3.675e+02, threshold=4.004e+02, percent-clipped=0.0 +2023-03-21 06:43:01,587 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 06:43:06,545 INFO [train.py:901] (0/2) Epoch 33, batch 300, loss[loss=0.1463, simple_loss=0.2237, pruned_loss=0.03445, over 7273.00 frames. ], tot_loss[loss=0.1351, simple_loss=0.2157, pruned_loss=0.02731, over 1123990.00 frames. ], batch size: 57, lr: 5.04e-03, grad_scale: 8.0 +2023-03-21 06:43:10,671 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 06:43:27,021 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=90707.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:43:32,915 INFO [train.py:901] (0/2) Epoch 33, batch 350, loss[loss=0.1396, simple_loss=0.2247, pruned_loss=0.02723, over 6541.00 frames. ], tot_loss[loss=0.1347, simple_loss=0.2155, pruned_loss=0.02698, over 1194933.17 frames. ], batch size: 106, lr: 5.04e-03, grad_scale: 8.0 +2023-03-21 06:43:42,930 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.149e+02 1.788e+02 2.029e+02 2.568e+02 9.373e+02, threshold=4.059e+02, percent-clipped=4.0 +2023-03-21 06:43:44,538 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 06:43:58,666 INFO [train.py:901] (0/2) Epoch 33, batch 400, loss[loss=0.1353, simple_loss=0.2251, pruned_loss=0.02278, over 7322.00 frames. ], tot_loss[loss=0.1346, simple_loss=0.2156, pruned_loss=0.02677, over 1251950.26 frames. ], batch size: 59, lr: 5.03e-03, grad_scale: 8.0 +2023-03-21 06:43:59,458 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.85 vs. limit=2.0 +2023-03-21 06:44:07,213 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=90786.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:44:08,291 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4169, 3.1398, 2.1178, 3.5367, 3.3434, 3.5198, 3.0415, 3.0690], + device='cuda:0'), covar=tensor([0.2056, 0.0849, 0.3873, 0.0544, 0.0247, 0.0237, 0.0335, 0.0446], + device='cuda:0'), in_proj_covar=tensor([0.0253, 0.0233, 0.0252, 0.0255, 0.0189, 0.0187, 0.0208, 0.0219], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 06:44:22,884 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=90815.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:44:24,764 INFO [train.py:901] (0/2) Epoch 33, batch 450, loss[loss=0.1399, simple_loss=0.2107, pruned_loss=0.03453, over 7204.00 frames. ], tot_loss[loss=0.1347, simple_loss=0.2153, pruned_loss=0.02706, over 1293844.63 frames. ], batch size: 45, lr: 5.03e-03, grad_scale: 8.0 +2023-03-21 06:44:27,030 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.54 vs. limit=5.0 +2023-03-21 06:44:28,707 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 06:44:29,219 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 06:44:32,275 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=90834.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:44:34,680 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.154e+02 1.801e+02 2.162e+02 2.490e+02 7.206e+02, threshold=4.324e+02, percent-clipped=3.0 +2023-03-21 06:44:39,874 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=90849.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:44:40,410 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6422, 2.9572, 2.4856, 2.6968, 2.7615, 2.5564, 2.6815, 2.7712], + device='cuda:0'), covar=tensor([0.0842, 0.0479, 0.1144, 0.1244, 0.1379, 0.0543, 0.1029, 0.0651], + device='cuda:0'), in_proj_covar=tensor([0.0055, 0.0056, 0.0064, 0.0055, 0.0053, 0.0057, 0.0054, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 06:44:40,665 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.61 vs. limit=2.0 +2023-03-21 06:44:41,944 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=90852.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:44:50,280 INFO [train.py:901] (0/2) Epoch 33, batch 500, loss[loss=0.1406, simple_loss=0.2211, pruned_loss=0.03006, over 7290.00 frames. ], tot_loss[loss=0.1347, simple_loss=0.2157, pruned_loss=0.02684, over 1327027.81 frames. ], batch size: 86, lr: 5.03e-03, grad_scale: 8.0 +2023-03-21 06:44:56,531 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4405, 2.9981, 2.1649, 3.4704, 3.3267, 3.4073, 2.9338, 3.0545], + device='cuda:0'), covar=tensor([0.1966, 0.0904, 0.3773, 0.0427, 0.0234, 0.0211, 0.0325, 0.0403], + device='cuda:0'), in_proj_covar=tensor([0.0254, 0.0232, 0.0251, 0.0255, 0.0189, 0.0187, 0.0208, 0.0219], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 06:45:00,489 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=90888.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:45:01,447 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 06:45:02,950 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 06:45:03,491 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 06:45:05,460 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 06:45:10,450 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 06:45:11,599 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=90910.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:45:15,843 INFO [train.py:901] (0/2) Epoch 33, batch 550, loss[loss=0.1292, simple_loss=0.2056, pruned_loss=0.02635, over 7221.00 frames. ], tot_loss[loss=0.1347, simple_loss=0.2159, pruned_loss=0.02678, over 1353744.85 frames. ], batch size: 45, lr: 5.03e-03, grad_scale: 8.0 +2023-03-21 06:45:20,845 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 06:45:25,070 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=90936.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:45:26,468 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.416e+02 1.907e+02 2.191e+02 2.571e+02 3.515e+02, threshold=4.383e+02, percent-clipped=0.0 +2023-03-21 06:45:30,010 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 06:45:32,976 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 06:45:40,049 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 06:45:42,043 INFO [train.py:901] (0/2) Epoch 33, batch 600, loss[loss=0.1329, simple_loss=0.2146, pruned_loss=0.02561, over 7359.00 frames. ], tot_loss[loss=0.1351, simple_loss=0.2164, pruned_loss=0.02695, over 1376081.76 frames. ], batch size: 73, lr: 5.03e-03, grad_scale: 8.0 +2023-03-21 06:45:42,154 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8072, 3.2720, 3.7619, 3.9200, 3.8258, 3.8618, 3.9566, 3.7400], + device='cuda:0'), covar=tensor([0.0030, 0.0105, 0.0033, 0.0030, 0.0034, 0.0033, 0.0039, 0.0053], + device='cuda:0'), in_proj_covar=tensor([0.0046, 0.0067, 0.0055, 0.0054, 0.0053, 0.0059, 0.0048, 0.0074], + device='cuda:0'), out_proj_covar=tensor([8.0506e-05, 1.4090e-04, 1.0352e-04, 9.6217e-05, 9.4734e-05, 1.0698e-04, + 9.4838e-05, 1.4112e-04], device='cuda:0') +2023-03-21 06:45:56,205 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 06:46:01,712 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=91007.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:46:05,691 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 06:46:07,661 INFO [train.py:901] (0/2) Epoch 33, batch 650, loss[loss=0.147, simple_loss=0.2249, pruned_loss=0.03458, over 7353.00 frames. ], tot_loss[loss=0.1347, simple_loss=0.2156, pruned_loss=0.02687, over 1385528.18 frames. ], batch size: 61, lr: 5.03e-03, grad_scale: 8.0 +2023-03-21 06:46:18,397 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.079e+02 1.852e+02 2.052e+02 2.347e+02 4.499e+02, threshold=4.105e+02, percent-clipped=1.0 +2023-03-21 06:46:24,345 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 06:46:26,431 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=91055.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:46:32,949 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 06:46:33,928 INFO [train.py:901] (0/2) Epoch 33, batch 700, loss[loss=0.1399, simple_loss=0.2203, pruned_loss=0.02976, over 7274.00 frames. ], tot_loss[loss=0.1348, simple_loss=0.2155, pruned_loss=0.02706, over 1396208.99 frames. ], batch size: 70, lr: 5.03e-03, grad_scale: 8.0 +2023-03-21 06:46:55,530 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4038, 1.6415, 1.4682, 1.5146, 1.7031, 1.3358, 1.3795, 1.1058], + device='cuda:0'), covar=tensor([0.0186, 0.0146, 0.0289, 0.0174, 0.0124, 0.0218, 0.0187, 0.0208], + device='cuda:0'), in_proj_covar=tensor([0.0034, 0.0031, 0.0031, 0.0032, 0.0030, 0.0029, 0.0033, 0.0040], + device='cuda:0'), out_proj_covar=tensor([3.8129e-05, 3.5242e-05, 3.5092e-05, 3.5278e-05, 3.3980e-05, 3.2679e-05, + 3.7433e-05, 4.4553e-05], device='cuda:0') +2023-03-21 06:46:57,021 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 06:46:57,493 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 06:46:57,609 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=91115.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:46:59,483 INFO [train.py:901] (0/2) Epoch 33, batch 750, loss[loss=0.1388, simple_loss=0.2198, pruned_loss=0.02887, over 7281.00 frames. ], tot_loss[loss=0.1346, simple_loss=0.2153, pruned_loss=0.02692, over 1408156.02 frames. ], batch size: 57, lr: 5.02e-03, grad_scale: 8.0 +2023-03-21 06:47:10,046 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.129e+02 1.773e+02 2.102e+02 2.526e+02 4.184e+02, threshold=4.205e+02, percent-clipped=1.0 +2023-03-21 06:47:11,616 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 06:47:16,054 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 06:47:16,641 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=91152.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:47:21,548 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 06:47:22,086 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=91163.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:47:23,039 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 06:47:25,001 INFO [train.py:901] (0/2) Epoch 33, batch 800, loss[loss=0.1382, simple_loss=0.2271, pruned_loss=0.02466, over 7351.00 frames. ], tot_loss[loss=0.1347, simple_loss=0.2157, pruned_loss=0.02688, over 1417719.02 frames. ], batch size: 63, lr: 5.02e-03, grad_scale: 8.0 +2023-03-21 06:47:32,553 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 06:47:41,708 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=91200.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:47:44,228 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=91205.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:47:51,121 INFO [train.py:901] (0/2) Epoch 33, batch 850, loss[loss=0.1351, simple_loss=0.221, pruned_loss=0.02459, over 7322.00 frames. ], tot_loss[loss=0.1341, simple_loss=0.2151, pruned_loss=0.0266, over 1419843.01 frames. ], batch size: 59, lr: 5.02e-03, grad_scale: 8.0 +2023-03-21 06:47:53,665 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 06:47:53,674 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 06:47:59,376 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 06:48:01,844 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.326e+02 1.733e+02 2.042e+02 2.278e+02 4.031e+02, threshold=4.085e+02, percent-clipped=0.0 +2023-03-21 06:48:02,919 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 06:48:06,532 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8872, 3.2269, 3.8363, 3.9434, 3.9860, 3.8661, 3.9292, 3.8012], + device='cuda:0'), covar=tensor([0.0028, 0.0106, 0.0030, 0.0022, 0.0021, 0.0031, 0.0038, 0.0047], + device='cuda:0'), in_proj_covar=tensor([0.0046, 0.0067, 0.0054, 0.0053, 0.0053, 0.0058, 0.0047, 0.0073], + device='cuda:0'), out_proj_covar=tensor([7.9772e-05, 1.4020e-04, 1.0240e-04, 9.5187e-05, 9.3677e-05, 1.0497e-04, + 9.3747e-05, 1.3883e-04], device='cuda:0') +2023-03-21 06:48:16,665 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-03-21 06:48:16,898 INFO [train.py:901] (0/2) Epoch 33, batch 900, loss[loss=0.1479, simple_loss=0.2289, pruned_loss=0.03343, over 7257.00 frames. ], tot_loss[loss=0.1338, simple_loss=0.2148, pruned_loss=0.0264, over 1425773.67 frames. ], batch size: 52, lr: 5.02e-03, grad_scale: 8.0 +2023-03-21 06:48:23,065 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9876, 2.2987, 1.8161, 2.6406, 2.5484, 2.6920, 2.3755, 2.5091], + device='cuda:0'), covar=tensor([0.2120, 0.0945, 0.3601, 0.0875, 0.0339, 0.0267, 0.0391, 0.0493], + device='cuda:0'), in_proj_covar=tensor([0.0257, 0.0234, 0.0252, 0.0256, 0.0191, 0.0190, 0.0210, 0.0220], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 06:48:25,085 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=91285.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:48:26,573 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2807, 4.7569, 4.8392, 4.7879, 4.7199, 4.3591, 4.8449, 4.7374], + device='cuda:0'), covar=tensor([0.0553, 0.0463, 0.0415, 0.0498, 0.0352, 0.0451, 0.0371, 0.0499], + device='cuda:0'), in_proj_covar=tensor([0.0211, 0.0247, 0.0187, 0.0193, 0.0150, 0.0221, 0.0196, 0.0144], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 06:48:40,884 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 06:48:43,374 INFO [train.py:901] (0/2) Epoch 33, batch 950, loss[loss=0.1257, simple_loss=0.2045, pruned_loss=0.02342, over 7244.00 frames. ], tot_loss[loss=0.1339, simple_loss=0.2148, pruned_loss=0.02647, over 1429518.40 frames. ], batch size: 55, lr: 5.02e-03, grad_scale: 8.0 +2023-03-21 06:48:53,315 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.253e+02 1.750e+02 2.027e+02 2.415e+02 4.015e+02, threshold=4.054e+02, percent-clipped=0.0 +2023-03-21 06:48:56,984 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=91346.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:49:03,936 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 06:49:04,539 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4955, 3.9787, 4.2132, 4.2133, 4.0660, 4.0387, 4.4337, 3.9746], + device='cuda:0'), covar=tensor([0.0139, 0.0167, 0.0112, 0.0139, 0.0471, 0.0112, 0.0129, 0.0150], + device='cuda:0'), in_proj_covar=tensor([0.0092, 0.0100, 0.0099, 0.0088, 0.0173, 0.0105, 0.0104, 0.0109], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 06:49:07,613 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-03-21 06:49:08,354 INFO [train.py:901] (0/2) Epoch 33, batch 1000, loss[loss=0.1652, simple_loss=0.2456, pruned_loss=0.04237, over 7257.00 frames. ], tot_loss[loss=0.1346, simple_loss=0.2154, pruned_loss=0.0269, over 1431804.06 frames. ], batch size: 64, lr: 5.02e-03, grad_scale: 8.0 +2023-03-21 06:49:16,929 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5736, 3.7172, 3.4997, 3.6868, 3.3590, 3.5677, 3.9439, 3.9399], + device='cuda:0'), covar=tensor([0.0224, 0.0167, 0.0247, 0.0179, 0.0362, 0.0338, 0.0233, 0.0184], + device='cuda:0'), in_proj_covar=tensor([0.0116, 0.0120, 0.0112, 0.0115, 0.0106, 0.0096, 0.0094, 0.0091], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 06:49:24,639 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 06:49:31,553 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5237, 2.5317, 2.4095, 3.5503, 1.7627, 3.5502, 1.4915, 3.2852], + device='cuda:0'), covar=tensor([0.0177, 0.1255, 0.1687, 0.0188, 0.4196, 0.0257, 0.1259, 0.0410], + device='cuda:0'), in_proj_covar=tensor([0.0202, 0.0252, 0.0266, 0.0204, 0.0254, 0.0211, 0.0232, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 06:49:34,938 INFO [train.py:901] (0/2) Epoch 33, batch 1050, loss[loss=0.1316, simple_loss=0.2191, pruned_loss=0.02206, over 7217.00 frames. ], tot_loss[loss=0.1346, simple_loss=0.2153, pruned_loss=0.02691, over 1433516.19 frames. ], batch size: 93, lr: 5.02e-03, grad_scale: 8.0 +2023-03-21 06:49:44,961 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.129e+02 1.697e+02 1.949e+02 2.264e+02 4.726e+02, threshold=3.897e+02, percent-clipped=1.0 +2023-03-21 06:49:45,999 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 06:49:50,008 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 06:50:00,708 INFO [train.py:901] (0/2) Epoch 33, batch 1100, loss[loss=0.137, simple_loss=0.2133, pruned_loss=0.03037, over 7328.00 frames. ], tot_loss[loss=0.1339, simple_loss=0.2147, pruned_loss=0.02656, over 1435462.34 frames. ], batch size: 54, lr: 5.01e-03, grad_scale: 8.0 +2023-03-21 06:50:02,010 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-03-21 06:50:19,579 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=91505.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:50:19,988 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 06:50:20,926 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 06:50:25,578 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2885, 3.0052, 2.0191, 3.4725, 3.0318, 3.5290, 2.8467, 2.8903], + device='cuda:0'), covar=tensor([0.2448, 0.1054, 0.4383, 0.0651, 0.0262, 0.0261, 0.0510, 0.0509], + device='cuda:0'), in_proj_covar=tensor([0.0255, 0.0232, 0.0250, 0.0256, 0.0189, 0.0189, 0.0209, 0.0220], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 06:50:26,426 INFO [train.py:901] (0/2) Epoch 33, batch 1150, loss[loss=0.1443, simple_loss=0.232, pruned_loss=0.02834, over 6626.00 frames. ], tot_loss[loss=0.1335, simple_loss=0.2142, pruned_loss=0.02643, over 1436490.35 frames. ], batch size: 106, lr: 5.01e-03, grad_scale: 8.0 +2023-03-21 06:50:33,016 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 06:50:33,518 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 06:50:36,493 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.127e+02 1.779e+02 2.098e+02 2.565e+02 4.143e+02, threshold=4.196e+02, percent-clipped=2.0 +2023-03-21 06:50:38,174 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=91542.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 06:50:44,196 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=91553.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:50:45,434 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.83 vs. limit=2.0 +2023-03-21 06:50:52,781 INFO [train.py:901] (0/2) Epoch 33, batch 1200, loss[loss=0.1359, simple_loss=0.2164, pruned_loss=0.02772, over 7280.00 frames. ], tot_loss[loss=0.1338, simple_loss=0.2144, pruned_loss=0.02659, over 1435879.77 frames. ], batch size: 57, lr: 5.01e-03, grad_scale: 8.0 +2023-03-21 06:51:03,918 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4681, 3.9852, 4.1944, 4.1828, 4.0517, 4.0222, 4.3951, 3.8862], + device='cuda:0'), covar=tensor([0.0119, 0.0149, 0.0101, 0.0145, 0.0408, 0.0111, 0.0121, 0.0160], + device='cuda:0'), in_proj_covar=tensor([0.0090, 0.0097, 0.0097, 0.0087, 0.0170, 0.0103, 0.0102, 0.0107], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 06:51:05,859 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 06:51:10,329 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=91603.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 06:51:15,488 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.09 vs. limit=2.0 +2023-03-21 06:51:18,293 INFO [train.py:901] (0/2) Epoch 33, batch 1250, loss[loss=0.1521, simple_loss=0.2281, pruned_loss=0.03801, over 7237.00 frames. ], tot_loss[loss=0.1343, simple_loss=0.2149, pruned_loss=0.02688, over 1436131.64 frames. ], batch size: 55, lr: 5.01e-03, grad_scale: 8.0 +2023-03-21 06:51:28,864 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.321e+02 1.782e+02 2.023e+02 2.411e+02 3.867e+02, threshold=4.047e+02, percent-clipped=0.0 +2023-03-21 06:51:29,949 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=91641.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:51:30,891 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 06:51:34,998 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 06:51:36,480 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 06:51:44,526 INFO [train.py:901] (0/2) Epoch 33, batch 1300, loss[loss=0.1512, simple_loss=0.2256, pruned_loss=0.03838, over 7274.00 frames. ], tot_loss[loss=0.135, simple_loss=0.2155, pruned_loss=0.02725, over 1438498.26 frames. ], batch size: 57, lr: 5.01e-03, grad_scale: 8.0 +2023-03-21 06:51:59,650 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 06:52:02,125 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 06:52:05,649 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 06:52:09,568 INFO [train.py:901] (0/2) Epoch 33, batch 1350, loss[loss=0.1538, simple_loss=0.235, pruned_loss=0.03634, over 7248.00 frames. ], tot_loss[loss=0.1351, simple_loss=0.2159, pruned_loss=0.02717, over 1441207.86 frames. ], batch size: 89, lr: 5.01e-03, grad_scale: 8.0 +2023-03-21 06:52:15,704 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 06:52:20,213 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.140e+02 1.841e+02 2.177e+02 2.731e+02 4.571e+02, threshold=4.353e+02, percent-clipped=2.0 +2023-03-21 06:52:36,006 INFO [train.py:901] (0/2) Epoch 33, batch 1400, loss[loss=0.1522, simple_loss=0.2241, pruned_loss=0.04018, over 7225.00 frames. ], tot_loss[loss=0.1356, simple_loss=0.2163, pruned_loss=0.02747, over 1441941.03 frames. ], batch size: 45, lr: 5.01e-03, grad_scale: 8.0 +2023-03-21 06:52:48,615 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 06:53:02,052 INFO [train.py:901] (0/2) Epoch 33, batch 1450, loss[loss=0.1423, simple_loss=0.2237, pruned_loss=0.03046, over 7279.00 frames. ], tot_loss[loss=0.1354, simple_loss=0.216, pruned_loss=0.02744, over 1441830.55 frames. ], batch size: 66, lr: 5.01e-03, grad_scale: 8.0 +2023-03-21 06:53:12,890 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.224e+02 1.866e+02 2.159e+02 2.546e+02 4.457e+02, threshold=4.319e+02, percent-clipped=1.0 +2023-03-21 06:53:13,427 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 06:53:27,925 INFO [train.py:901] (0/2) Epoch 33, batch 1500, loss[loss=0.1477, simple_loss=0.2281, pruned_loss=0.03363, over 7360.00 frames. ], tot_loss[loss=0.1347, simple_loss=0.2153, pruned_loss=0.02703, over 1440527.54 frames. ], batch size: 73, lr: 5.00e-03, grad_scale: 8.0 +2023-03-21 06:53:30,497 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 06:53:39,050 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=91891.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:53:43,112 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=91898.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 06:53:53,319 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.2271, 1.5167, 1.2988, 1.3703, 1.5609, 1.3778, 1.3254, 1.1153], + device='cuda:0'), covar=tensor([0.0170, 0.0161, 0.0237, 0.0150, 0.0123, 0.0122, 0.0140, 0.0190], + device='cuda:0'), in_proj_covar=tensor([0.0034, 0.0032, 0.0032, 0.0032, 0.0031, 0.0030, 0.0034, 0.0041], + device='cuda:0'), out_proj_covar=tensor([3.8961e-05, 3.5589e-05, 3.6104e-05, 3.6090e-05, 3.4552e-05, 3.3919e-05, + 3.8269e-05, 4.5324e-05], device='cuda:0') +2023-03-21 06:53:54,181 INFO [train.py:901] (0/2) Epoch 33, batch 1550, loss[loss=0.137, simple_loss=0.2119, pruned_loss=0.03108, over 7264.00 frames. ], tot_loss[loss=0.1347, simple_loss=0.2154, pruned_loss=0.02698, over 1440832.13 frames. ], batch size: 52, lr: 5.00e-03, grad_scale: 8.0 +2023-03-21 06:53:55,727 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 06:53:56,336 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7166, 1.4571, 1.9089, 2.2558, 1.9098, 2.1725, 1.7431, 2.1766], + device='cuda:0'), covar=tensor([0.3156, 0.3581, 0.2162, 0.1375, 0.4720, 0.1247, 0.1631, 0.2760], + device='cuda:0'), in_proj_covar=tensor([0.0073, 0.0075, 0.0064, 0.0060, 0.0060, 0.0060, 0.0098, 0.0061], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 06:54:04,238 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.339e+02 1.806e+02 2.097e+02 2.393e+02 4.199e+02, threshold=4.194e+02, percent-clipped=0.0 +2023-03-21 06:54:05,384 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=91941.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:54:10,998 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=91952.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:54:12,433 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=91955.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:54:19,278 INFO [train.py:901] (0/2) Epoch 33, batch 1600, loss[loss=0.1072, simple_loss=0.1873, pruned_loss=0.01357, over 7133.00 frames. ], tot_loss[loss=0.1341, simple_loss=0.2146, pruned_loss=0.02676, over 1440728.58 frames. ], batch size: 39, lr: 5.00e-03, grad_scale: 8.0 +2023-03-21 06:54:24,910 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 06:54:25,452 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 06:54:25,581 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=91981.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 06:54:29,019 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 06:54:30,085 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=91989.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:54:36,518 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-92000.pt +2023-03-21 06:54:43,104 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 06:54:47,083 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 06:54:48,263 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=92016.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:54:49,707 INFO [train.py:901] (0/2) Epoch 33, batch 1650, loss[loss=0.137, simple_loss=0.2171, pruned_loss=0.0284, over 7325.00 frames. ], tot_loss[loss=0.1343, simple_loss=0.2149, pruned_loss=0.02688, over 1441873.98 frames. ], batch size: 75, lr: 5.00e-03, grad_scale: 8.0 +2023-03-21 06:54:55,337 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 06:54:59,839 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.092e+02 1.781e+02 2.039e+02 2.347e+02 4.306e+02, threshold=4.078e+02, percent-clipped=2.0 +2023-03-21 06:55:01,486 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=92042.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 06:55:11,456 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 06:55:14,853 INFO [train.py:901] (0/2) Epoch 33, batch 1700, loss[loss=0.1397, simple_loss=0.2136, pruned_loss=0.03295, over 7305.00 frames. ], tot_loss[loss=0.1339, simple_loss=0.2146, pruned_loss=0.0266, over 1442513.63 frames. ], batch size: 49, lr: 5.00e-03, grad_scale: 8.0 +2023-03-21 06:55:15,881 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 06:55:26,481 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-03-21 06:55:27,653 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 06:55:41,151 INFO [train.py:901] (0/2) Epoch 33, batch 1750, loss[loss=0.1305, simple_loss=0.2135, pruned_loss=0.02375, over 7276.00 frames. ], tot_loss[loss=0.1339, simple_loss=0.2144, pruned_loss=0.02669, over 1442464.89 frames. ], batch size: 77, lr: 5.00e-03, grad_scale: 8.0 +2023-03-21 06:55:44,367 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1919, 2.9826, 2.2659, 3.3866, 2.4300, 2.9732, 1.5414, 2.3555], + device='cuda:0'), covar=tensor([0.0418, 0.0746, 0.2605, 0.0662, 0.0423, 0.0533, 0.3293, 0.1575], + device='cuda:0'), in_proj_covar=tensor([0.0280, 0.0256, 0.0283, 0.0269, 0.0268, 0.0267, 0.0238, 0.0263], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 06:55:51,183 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.113e+02 1.689e+02 2.025e+02 2.316e+02 3.712e+02, threshold=4.050e+02, percent-clipped=0.0 +2023-03-21 06:55:51,211 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 06:55:52,240 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 06:56:07,478 INFO [train.py:901] (0/2) Epoch 33, batch 1800, loss[loss=0.1411, simple_loss=0.2242, pruned_loss=0.029, over 7307.00 frames. ], tot_loss[loss=0.1331, simple_loss=0.2136, pruned_loss=0.02631, over 1440153.70 frames. ], batch size: 59, lr: 5.00e-03, grad_scale: 8.0 +2023-03-21 06:56:15,428 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 06:56:15,849 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.63 vs. limit=2.0 +2023-03-21 06:56:22,124 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=92198.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 06:56:25,446 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5127, 3.9300, 4.1339, 4.1719, 4.0130, 3.9889, 4.3824, 3.7884], + device='cuda:0'), covar=tensor([0.0114, 0.0177, 0.0116, 0.0149, 0.0462, 0.0119, 0.0139, 0.0200], + device='cuda:0'), in_proj_covar=tensor([0.0091, 0.0100, 0.0098, 0.0088, 0.0174, 0.0104, 0.0103, 0.0108], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 06:56:28,809 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 06:56:32,504 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.60 vs. limit=5.0 +2023-03-21 06:56:32,720 INFO [train.py:901] (0/2) Epoch 33, batch 1850, loss[loss=0.1347, simple_loss=0.2143, pruned_loss=0.02755, over 7293.00 frames. ], tot_loss[loss=0.134, simple_loss=0.2143, pruned_loss=0.02686, over 1438765.54 frames. ], batch size: 66, lr: 4.99e-03, grad_scale: 8.0 +2023-03-21 06:56:38,724 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 06:56:42,785 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.249e+02 1.760e+02 1.984e+02 2.426e+02 4.965e+02, threshold=3.969e+02, percent-clipped=2.0 +2023-03-21 06:56:46,345 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=92246.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 06:56:46,809 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=92247.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:56:56,491 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 06:56:56,602 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5535, 4.7213, 4.5608, 4.7373, 4.3882, 4.7234, 5.0025, 5.0179], + device='cuda:0'), covar=tensor([0.0182, 0.0116, 0.0179, 0.0117, 0.0268, 0.0169, 0.0171, 0.0130], + device='cuda:0'), in_proj_covar=tensor([0.0118, 0.0122, 0.0114, 0.0117, 0.0108, 0.0098, 0.0096, 0.0094], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 06:56:59,088 INFO [train.py:901] (0/2) Epoch 33, batch 1900, loss[loss=0.1377, simple_loss=0.2213, pruned_loss=0.02702, over 7315.00 frames. ], tot_loss[loss=0.1344, simple_loss=0.2144, pruned_loss=0.02721, over 1436983.17 frames. ], batch size: 83, lr: 4.99e-03, grad_scale: 8.0 +2023-03-21 06:57:20,191 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=92311.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:57:21,165 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 06:57:22,413 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.84 vs. limit=2.0 +2023-03-21 06:57:24,139 INFO [train.py:901] (0/2) Epoch 33, batch 1950, loss[loss=0.1394, simple_loss=0.2186, pruned_loss=0.03014, over 7360.00 frames. ], tot_loss[loss=0.1347, simple_loss=0.215, pruned_loss=0.02717, over 1439954.93 frames. ], batch size: 63, lr: 4.99e-03, grad_scale: 8.0 +2023-03-21 06:57:31,742 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 06:57:33,282 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=92337.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 06:57:34,759 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.334e+02 1.739e+02 2.064e+02 2.518e+02 6.766e+02, threshold=4.128e+02, percent-clipped=1.0 +2023-03-21 06:57:35,589 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.83 vs. limit=5.0 +2023-03-21 06:57:36,850 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 06:57:37,361 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 06:57:50,559 INFO [train.py:901] (0/2) Epoch 33, batch 2000, loss[loss=0.1377, simple_loss=0.2201, pruned_loss=0.02768, over 7270.00 frames. ], tot_loss[loss=0.1348, simple_loss=0.2155, pruned_loss=0.02702, over 1441663.73 frames. ], batch size: 52, lr: 4.99e-03, grad_scale: 8.0 +2023-03-21 06:57:54,101 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 06:57:54,264 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4292, 2.8587, 2.1572, 3.0650, 3.2041, 3.1781, 2.9916, 2.8860], + device='cuda:0'), covar=tensor([0.1958, 0.0963, 0.3604, 0.0535, 0.0288, 0.0196, 0.0367, 0.0444], + device='cuda:0'), in_proj_covar=tensor([0.0253, 0.0231, 0.0249, 0.0254, 0.0190, 0.0189, 0.0208, 0.0219], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 06:58:04,657 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 06:58:13,440 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 06:58:15,966 INFO [train.py:901] (0/2) Epoch 33, batch 2050, loss[loss=0.1289, simple_loss=0.2064, pruned_loss=0.02567, over 7250.00 frames. ], tot_loss[loss=0.1344, simple_loss=0.2146, pruned_loss=0.02703, over 1439316.91 frames. ], batch size: 89, lr: 4.99e-03, grad_scale: 8.0 +2023-03-21 06:58:27,229 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.309e+02 1.918e+02 2.279e+02 2.660e+02 4.216e+02, threshold=4.557e+02, percent-clipped=1.0 +2023-03-21 06:58:42,294 INFO [train.py:901] (0/2) Epoch 33, batch 2100, loss[loss=0.1472, simple_loss=0.2251, pruned_loss=0.03459, over 7266.00 frames. ], tot_loss[loss=0.1345, simple_loss=0.2152, pruned_loss=0.02694, over 1440092.53 frames. ], batch size: 52, lr: 4.99e-03, grad_scale: 8.0 +2023-03-21 06:58:46,279 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 06:58:49,274 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 06:58:50,956 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6950, 1.5342, 1.9393, 2.2747, 1.9470, 2.1293, 1.7709, 2.1643], + device='cuda:0'), covar=tensor([0.1369, 0.3520, 0.1244, 0.0666, 0.4607, 0.2466, 0.1641, 0.1329], + device='cuda:0'), in_proj_covar=tensor([0.0072, 0.0075, 0.0064, 0.0060, 0.0060, 0.0060, 0.0099, 0.0061], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 06:59:08,986 INFO [train.py:901] (0/2) Epoch 33, batch 2150, loss[loss=0.1414, simple_loss=0.2204, pruned_loss=0.03126, over 7296.00 frames. ], tot_loss[loss=0.1345, simple_loss=0.2151, pruned_loss=0.02691, over 1439824.21 frames. ], batch size: 68, lr: 4.99e-03, grad_scale: 8.0 +2023-03-21 06:59:10,623 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8653, 3.7805, 3.1162, 3.3767, 2.8229, 2.1443, 1.7796, 3.8940], + device='cuda:0'), covar=tensor([0.0053, 0.0063, 0.0147, 0.0086, 0.0173, 0.0541, 0.0644, 0.0051], + device='cuda:0'), in_proj_covar=tensor([0.0109, 0.0088, 0.0109, 0.0094, 0.0123, 0.0130, 0.0127, 0.0100], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 06:59:19,163 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.208e+02 1.797e+02 1.982e+02 2.424e+02 4.305e+02, threshold=3.964e+02, percent-clipped=0.0 +2023-03-21 06:59:23,365 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=92547.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:59:34,086 INFO [train.py:901] (0/2) Epoch 33, batch 2200, loss[loss=0.1167, simple_loss=0.1882, pruned_loss=0.02255, over 6996.00 frames. ], tot_loss[loss=0.1349, simple_loss=0.2158, pruned_loss=0.02699, over 1441064.86 frames. ], batch size: 35, lr: 4.98e-03, grad_scale: 8.0 +2023-03-21 06:59:35,595 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 06:59:47,991 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=92595.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:59:56,969 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=92611.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 06:59:59,608 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-03-21 07:00:00,849 INFO [train.py:901] (0/2) Epoch 33, batch 2250, loss[loss=0.1314, simple_loss=0.2109, pruned_loss=0.02592, over 7311.00 frames. ], tot_loss[loss=0.1348, simple_loss=0.2156, pruned_loss=0.02704, over 1439866.73 frames. ], batch size: 59, lr: 4.98e-03, grad_scale: 8.0 +2023-03-21 07:00:04,074 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=92625.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:00:09,042 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 07:00:09,500 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 07:00:10,123 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=92637.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 07:00:10,964 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.180e+02 1.747e+02 2.128e+02 2.442e+02 8.863e+02, threshold=4.255e+02, percent-clipped=5.0 +2023-03-21 07:00:21,206 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=92659.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:00:21,683 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 07:00:26,181 INFO [train.py:901] (0/2) Epoch 33, batch 2300, loss[loss=0.1242, simple_loss=0.2074, pruned_loss=0.02051, over 7275.00 frames. ], tot_loss[loss=0.1345, simple_loss=0.2153, pruned_loss=0.0268, over 1441731.61 frames. ], batch size: 77, lr: 4.98e-03, grad_scale: 16.0 +2023-03-21 07:00:35,511 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=92685.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 07:00:36,055 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=92686.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:00:52,478 INFO [train.py:901] (0/2) Epoch 33, batch 2350, loss[loss=0.1431, simple_loss=0.2301, pruned_loss=0.02807, over 6735.00 frames. ], tot_loss[loss=0.1343, simple_loss=0.2155, pruned_loss=0.02658, over 1442944.63 frames. ], batch size: 107, lr: 4.98e-03, grad_scale: 16.0 +2023-03-21 07:00:53,129 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=92720.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:00:56,225 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6831, 3.4376, 2.6440, 4.0179, 3.0836, 3.5800, 1.8887, 2.7017], + device='cuda:0'), covar=tensor([0.0506, 0.1015, 0.2465, 0.0439, 0.0438, 0.0663, 0.3301, 0.1881], + device='cuda:0'), in_proj_covar=tensor([0.0278, 0.0254, 0.0280, 0.0268, 0.0269, 0.0265, 0.0236, 0.0260], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 07:01:02,575 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.140e+02 1.641e+02 1.954e+02 2.231e+02 4.206e+02, threshold=3.909e+02, percent-clipped=0.0 +2023-03-21 07:01:10,218 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 07:01:16,357 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 07:01:18,388 INFO [train.py:901] (0/2) Epoch 33, batch 2400, loss[loss=0.1249, simple_loss=0.2112, pruned_loss=0.01925, over 7307.00 frames. ], tot_loss[loss=0.1338, simple_loss=0.2149, pruned_loss=0.02636, over 1441548.01 frames. ], batch size: 83, lr: 4.98e-03, grad_scale: 16.0 +2023-03-21 07:01:20,710 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4445, 2.8108, 2.3858, 2.5635, 2.6848, 2.2433, 2.6850, 2.5069], + device='cuda:0'), covar=tensor([0.0719, 0.0500, 0.0926, 0.0901, 0.1216, 0.0943, 0.0622, 0.0943], + device='cuda:0'), in_proj_covar=tensor([0.0057, 0.0056, 0.0065, 0.0057, 0.0054, 0.0059, 0.0055, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:01:25,396 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=92781.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:01:28,366 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 07:01:31,899 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 07:01:44,635 INFO [train.py:901] (0/2) Epoch 33, batch 2450, loss[loss=0.1569, simple_loss=0.2347, pruned_loss=0.0395, over 7329.00 frames. ], tot_loss[loss=0.1335, simple_loss=0.2144, pruned_loss=0.02632, over 1443023.11 frames. ], batch size: 54, lr: 4.98e-03, grad_scale: 16.0 +2023-03-21 07:01:50,915 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=92831.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:01:54,891 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.351e+02 1.774e+02 2.037e+02 2.369e+02 4.194e+02, threshold=4.073e+02, percent-clipped=1.0 +2023-03-21 07:01:58,993 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 07:01:59,229 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.93 vs. limit=5.0 +2023-03-21 07:02:11,266 INFO [train.py:901] (0/2) Epoch 33, batch 2500, loss[loss=0.1259, simple_loss=0.2087, pruned_loss=0.02158, over 7311.00 frames. ], tot_loss[loss=0.1332, simple_loss=0.2141, pruned_loss=0.02618, over 1443575.90 frames. ], batch size: 80, lr: 4.98e-03, grad_scale: 16.0 +2023-03-21 07:02:20,530 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9539, 4.5034, 4.5761, 4.4639, 4.4625, 4.0540, 4.5860, 4.4408], + device='cuda:0'), covar=tensor([0.0540, 0.0427, 0.0356, 0.0535, 0.0325, 0.0436, 0.0357, 0.0441], + device='cuda:0'), in_proj_covar=tensor([0.0208, 0.0246, 0.0188, 0.0189, 0.0149, 0.0219, 0.0196, 0.0144], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:02:23,017 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=92892.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:02:24,858 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 07:02:26,491 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=92899.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:02:36,328 INFO [train.py:901] (0/2) Epoch 33, batch 2550, loss[loss=0.134, simple_loss=0.2193, pruned_loss=0.02434, over 7326.00 frames. ], tot_loss[loss=0.1339, simple_loss=0.2149, pruned_loss=0.0264, over 1442805.68 frames. ], batch size: 61, lr: 4.98e-03, grad_scale: 8.0 +2023-03-21 07:02:39,496 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9876, 1.9894, 2.4239, 2.7662, 1.8652, 2.2327, 2.5378, 2.4213], + device='cuda:0'), covar=tensor([0.3250, 0.4435, 0.1584, 0.1951, 0.7968, 0.5634, 0.2655, 0.5386], + device='cuda:0'), in_proj_covar=tensor([0.0072, 0.0075, 0.0064, 0.0060, 0.0060, 0.0060, 0.0100, 0.0061], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:02:42,433 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=92931.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:02:46,897 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.311e+02 1.757e+02 2.040e+02 2.452e+02 4.787e+02, threshold=4.080e+02, percent-clipped=2.0 +2023-03-21 07:02:58,325 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=92960.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:03:02,612 INFO [train.py:901] (0/2) Epoch 33, batch 2600, loss[loss=0.1345, simple_loss=0.2214, pruned_loss=0.02384, over 7285.00 frames. ], tot_loss[loss=0.1342, simple_loss=0.215, pruned_loss=0.02673, over 1442736.07 frames. ], batch size: 66, lr: 4.97e-03, grad_scale: 8.0 +2023-03-21 07:03:05,797 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=92975.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:03:08,616 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=92981.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:03:14,090 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=92992.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:03:23,315 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=93010.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:03:27,667 INFO [train.py:901] (0/2) Epoch 33, batch 2650, loss[loss=0.1671, simple_loss=0.2467, pruned_loss=0.04375, over 6614.00 frames. ], tot_loss[loss=0.1338, simple_loss=0.2146, pruned_loss=0.02652, over 1441289.68 frames. ], batch size: 106, lr: 4.97e-03, grad_scale: 8.0 +2023-03-21 07:03:36,391 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=93036.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:03:36,406 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=93036.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:03:38,270 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.296e+02 1.709e+02 2.039e+02 2.473e+02 4.429e+02, threshold=4.078e+02, percent-clipped=1.0 +2023-03-21 07:03:50,871 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.89 vs. limit=2.0 +2023-03-21 07:03:52,585 INFO [train.py:901] (0/2) Epoch 33, batch 2700, loss[loss=0.1532, simple_loss=0.2326, pruned_loss=0.03687, over 7254.00 frames. ], tot_loss[loss=0.1337, simple_loss=0.2146, pruned_loss=0.02635, over 1441694.13 frames. ], batch size: 89, lr: 4.97e-03, grad_scale: 8.0 +2023-03-21 07:03:53,705 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=93071.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:03:54,672 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1922, 4.3579, 4.0248, 4.3548, 3.9446, 4.3161, 4.5691, 4.5931], + device='cuda:0'), covar=tensor([0.0205, 0.0132, 0.0245, 0.0142, 0.0374, 0.0212, 0.0266, 0.0186], + device='cuda:0'), in_proj_covar=tensor([0.0115, 0.0119, 0.0112, 0.0114, 0.0106, 0.0097, 0.0094, 0.0093], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:03:56,177 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=93076.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:04:06,572 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=93097.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 07:04:15,358 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=93115.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:04:17,246 INFO [train.py:901] (0/2) Epoch 33, batch 2750, loss[loss=0.1355, simple_loss=0.2235, pruned_loss=0.02381, over 7282.00 frames. ], tot_loss[loss=0.134, simple_loss=0.2151, pruned_loss=0.02649, over 1441864.12 frames. ], batch size: 77, lr: 4.97e-03, grad_scale: 8.0 +2023-03-21 07:04:27,761 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.140e+02 1.702e+02 2.040e+02 2.360e+02 5.193e+02, threshold=4.081e+02, percent-clipped=1.0 +2023-03-21 07:04:28,590 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.87 vs. limit=2.0 +2023-03-21 07:04:41,914 INFO [train.py:901] (0/2) Epoch 33, batch 2800, loss[loss=0.139, simple_loss=0.2148, pruned_loss=0.03165, over 7210.00 frames. ], tot_loss[loss=0.1343, simple_loss=0.2152, pruned_loss=0.02671, over 1442913.91 frames. ], batch size: 50, lr: 4.97e-03, grad_scale: 8.0 +2023-03-21 07:04:45,802 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=93176.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:04:51,035 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=93187.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:04:54,813 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-33.pt +2023-03-21 07:05:10,428 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 07:05:10,817 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 07:05:10,841 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 07:05:10,859 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 07:05:11,051 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 07:05:11,067 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 07:05:11,079 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 07:05:11,182 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 07:05:11,262 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 07:05:11,277 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 07:05:11,295 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 07:05:11,471 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 07:05:11,477 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 07:05:11,487 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 07:05:11,981 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 07:05:11,985 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 07:05:12,047 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 07:05:14,366 INFO [train.py:901] (0/2) Epoch 34, batch 0, loss[loss=0.1562, simple_loss=0.2353, pruned_loss=0.03852, over 7260.00 frames. ], tot_loss[loss=0.1562, simple_loss=0.2353, pruned_loss=0.03852, over 7260.00 frames. ], batch size: 55, lr: 4.89e-03, grad_scale: 8.0 +2023-03-21 07:05:14,367 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 07:05:21,656 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5444, 4.1833, 3.8359, 4.5244, 4.3257, 4.5430, 4.3242, 4.2978], + device='cuda:0'), covar=tensor([0.0616, 0.2109, 0.2120, 0.1175, 0.0846, 0.0988, 0.0581, 0.0959], + device='cuda:0'), in_proj_covar=tensor([0.0249, 0.0381, 0.0290, 0.0299, 0.0222, 0.0358, 0.0218, 0.0267], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:05:24,593 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0169, 3.4542, 2.9919, 4.1805, 2.2604, 3.9873, 2.1873, 3.6703], + device='cuda:0'), covar=tensor([0.0157, 0.0697, 0.1473, 0.0157, 0.3820, 0.0178, 0.1006, 0.0402], + device='cuda:0'), in_proj_covar=tensor([0.0201, 0.0252, 0.0263, 0.0204, 0.0252, 0.0210, 0.0231, 0.0225], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 07:05:40,677 INFO [train.py:935] (0/2) Epoch 34, validation: loss=0.1638, simple_loss=0.2553, pruned_loss=0.03615, over 1622729.00 frames. +2023-03-21 07:05:40,677 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 07:05:41,331 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1684, 2.5135, 2.6645, 2.2834, 2.3800, 2.4912, 2.2852, 1.8456], + device='cuda:0'), covar=tensor([0.0451, 0.0507, 0.0382, 0.0349, 0.0732, 0.0373, 0.0298, 0.0416], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0034, 0.0036, 0.0034, 0.0033, 0.0032, 0.0036, 0.0036], + device='cuda:0'), out_proj_covar=tensor([8.9868e-05, 8.8643e-05, 9.0090e-05, 8.5280e-05, 8.6486e-05, 8.4884e-05, + 9.0996e-05, 9.2258e-05], device='cuda:0') +2023-03-21 07:05:46,496 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4682, 3.3745, 2.5356, 3.8961, 2.9303, 3.4369, 1.7109, 2.4296], + device='cuda:0'), covar=tensor([0.0505, 0.0733, 0.2321, 0.0481, 0.0442, 0.0685, 0.3281, 0.1783], + device='cuda:0'), in_proj_covar=tensor([0.0281, 0.0257, 0.0282, 0.0271, 0.0270, 0.0266, 0.0240, 0.0264], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 07:05:49,823 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 07:05:51,839 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0344, 3.4493, 4.1011, 4.2724, 4.2147, 4.1832, 4.2000, 4.0494], + device='cuda:0'), covar=tensor([0.0034, 0.0123, 0.0039, 0.0038, 0.0035, 0.0036, 0.0032, 0.0054], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0067, 0.0056, 0.0053, 0.0054, 0.0059, 0.0048, 0.0074], + device='cuda:0'), out_proj_covar=tensor([8.1159e-05, 1.4076e-04, 1.0418e-04, 9.4647e-05, 9.4346e-05, 1.0587e-04, + 9.4824e-05, 1.4128e-04], device='cuda:0') +2023-03-21 07:05:59,709 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 07:06:05,299 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.104e+02 1.686e+02 1.995e+02 2.522e+02 6.389e+02, threshold=3.989e+02, percent-clipped=1.0 +2023-03-21 07:06:06,308 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 07:06:06,791 INFO [train.py:901] (0/2) Epoch 34, batch 50, loss[loss=0.1326, simple_loss=0.214, pruned_loss=0.02558, over 7283.00 frames. ], tot_loss[loss=0.1337, simple_loss=0.2155, pruned_loss=0.02592, over 324654.25 frames. ], batch size: 68, lr: 4.89e-03, grad_scale: 8.0 +2023-03-21 07:06:08,816 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 07:06:11,412 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 07:06:12,985 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=93255.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:06:25,938 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=93281.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:06:26,449 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7682, 2.3723, 2.7912, 3.0154, 2.8216, 2.7121, 2.3253, 2.7803], + device='cuda:0'), covar=tensor([0.1927, 0.1230, 0.1300, 0.0776, 0.1018, 0.1174, 0.2469, 0.1780], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0064, 0.0049, 0.0048, 0.0048, 0.0048, 0.0065, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 07:06:29,508 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=93287.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:06:29,969 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 07:06:30,469 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 07:06:33,139 INFO [train.py:901] (0/2) Epoch 34, batch 100, loss[loss=0.1617, simple_loss=0.2524, pruned_loss=0.03549, over 6749.00 frames. ], tot_loss[loss=0.1345, simple_loss=0.2158, pruned_loss=0.0266, over 572902.72 frames. ], batch size: 106, lr: 4.89e-03, grad_scale: 8.0 +2023-03-21 07:06:43,532 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.18 vs. limit=2.0 +2023-03-21 07:06:51,319 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=93329.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:06:52,375 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=93331.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:06:56,806 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.175e+02 1.737e+02 2.034e+02 2.402e+02 4.556e+02, threshold=4.067e+02, percent-clipped=1.0 +2023-03-21 07:06:58,280 INFO [train.py:901] (0/2) Epoch 34, batch 150, loss[loss=0.1582, simple_loss=0.2354, pruned_loss=0.04049, over 6759.00 frames. ], tot_loss[loss=0.1347, simple_loss=0.2158, pruned_loss=0.02685, over 763520.83 frames. ], batch size: 107, lr: 4.89e-03, grad_scale: 8.0 +2023-03-21 07:07:09,980 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=93366.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:07:16,219 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=93376.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:07:18,962 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.55 vs. limit=5.0 +2023-03-21 07:07:24,199 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=93392.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 07:07:24,606 INFO [train.py:901] (0/2) Epoch 34, batch 200, loss[loss=0.1337, simple_loss=0.213, pruned_loss=0.02714, over 7294.00 frames. ], tot_loss[loss=0.1332, simple_loss=0.2145, pruned_loss=0.02597, over 916888.07 frames. ], batch size: 86, lr: 4.89e-03, grad_scale: 8.0 +2023-03-21 07:07:29,120 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 07:07:33,768 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 07:07:39,291 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 07:07:40,835 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=93424.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:07:42,421 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7802, 3.3159, 3.6545, 3.6362, 3.2563, 3.0997, 3.4392, 2.7871], + device='cuda:0'), covar=tensor([0.0361, 0.0407, 0.0529, 0.0617, 0.0781, 0.1002, 0.0705, 0.1925], + device='cuda:0'), in_proj_covar=tensor([0.0331, 0.0338, 0.0273, 0.0358, 0.0294, 0.0292, 0.0345, 0.0258], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:07:48,649 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.73 vs. limit=2.0 +2023-03-21 07:07:48,749 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.225e+02 1.705e+02 1.990e+02 2.336e+02 5.774e+02, threshold=3.979e+02, percent-clipped=1.0 +2023-03-21 07:07:48,857 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.0016, 4.5994, 4.4105, 4.9950, 4.8560, 4.9734, 4.4844, 4.6116], + device='cuda:0'), covar=tensor([0.0905, 0.2380, 0.2079, 0.1027, 0.0808, 0.1006, 0.0696, 0.1107], + device='cuda:0'), in_proj_covar=tensor([0.0250, 0.0379, 0.0287, 0.0297, 0.0219, 0.0356, 0.0216, 0.0265], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:07:50,290 INFO [train.py:901] (0/2) Epoch 34, batch 250, loss[loss=0.1528, simple_loss=0.2338, pruned_loss=0.0359, over 7346.00 frames. ], tot_loss[loss=0.134, simple_loss=0.2154, pruned_loss=0.0263, over 1032675.76 frames. ], batch size: 63, lr: 4.89e-03, grad_scale: 8.0 +2023-03-21 07:07:52,264 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 07:08:05,534 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=93471.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:08:12,982 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 07:08:13,558 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=93487.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:08:16,427 INFO [train.py:901] (0/2) Epoch 34, batch 300, loss[loss=0.1362, simple_loss=0.2197, pruned_loss=0.02633, over 7263.00 frames. ], tot_loss[loss=0.1349, simple_loss=0.2158, pruned_loss=0.02698, over 1124308.89 frames. ], batch size: 89, lr: 4.89e-03, grad_scale: 8.0 +2023-03-21 07:08:21,993 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 07:08:28,272 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.81 vs. limit=5.0 +2023-03-21 07:08:30,072 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4885, 1.8815, 1.6069, 1.6511, 1.7429, 1.5487, 1.5720, 1.3323], + device='cuda:0'), covar=tensor([0.0157, 0.0103, 0.0288, 0.0148, 0.0124, 0.0116, 0.0155, 0.0169], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0032, 0.0032, 0.0033, 0.0032, 0.0031, 0.0034, 0.0041], + device='cuda:0'), out_proj_covar=tensor([3.9107e-05, 3.5999e-05, 3.5934e-05, 3.6909e-05, 3.5469e-05, 3.4316e-05, + 3.8825e-05, 4.5022e-05], device='cuda:0') +2023-03-21 07:08:35,166 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8819, 3.0033, 2.5887, 4.0364, 1.9986, 3.7988, 1.7307, 3.5360], + device='cuda:0'), covar=tensor([0.0195, 0.1012, 0.1573, 0.0156, 0.3400, 0.0238, 0.1001, 0.0389], + device='cuda:0'), in_proj_covar=tensor([0.0201, 0.0251, 0.0264, 0.0203, 0.0251, 0.0208, 0.0230, 0.0225], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 07:08:37,521 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=93535.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:08:39,924 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.325e+02 2.018e+02 2.343e+02 2.755e+02 4.268e+02, threshold=4.687e+02, percent-clipped=3.0 +2023-03-21 07:08:42,065 INFO [train.py:901] (0/2) Epoch 34, batch 350, loss[loss=0.1345, simple_loss=0.2183, pruned_loss=0.02538, over 7306.00 frames. ], tot_loss[loss=0.1349, simple_loss=0.2156, pruned_loss=0.02715, over 1193661.74 frames. ], batch size: 86, lr: 4.89e-03, grad_scale: 8.0 +2023-03-21 07:08:48,922 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=93555.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:08:54,963 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3462, 3.8519, 4.1125, 4.3294, 4.4054, 4.3319, 4.4353, 4.2330], + device='cuda:0'), covar=tensor([0.0023, 0.0082, 0.0037, 0.0028, 0.0024, 0.0032, 0.0029, 0.0056], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0069, 0.0057, 0.0055, 0.0055, 0.0060, 0.0049, 0.0076], + device='cuda:0'), out_proj_covar=tensor([8.1774e-05, 1.4329e-04, 1.0737e-04, 9.7115e-05, 9.6589e-05, 1.0784e-04, + 9.7121e-05, 1.4363e-04], device='cuda:0') +2023-03-21 07:08:57,355 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 07:09:04,900 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=93587.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:09:07,746 INFO [train.py:901] (0/2) Epoch 34, batch 400, loss[loss=0.1436, simple_loss=0.2249, pruned_loss=0.0312, over 7293.00 frames. ], tot_loss[loss=0.1346, simple_loss=0.2152, pruned_loss=0.02698, over 1248791.79 frames. ], batch size: 77, lr: 4.88e-03, grad_scale: 8.0 +2023-03-21 07:09:13,159 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=93603.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:09:28,721 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=93631.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:09:30,683 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=93635.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:09:31,436 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.71 vs. limit=2.0 +2023-03-21 07:09:32,770 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8603, 1.7101, 2.2509, 2.5328, 2.2562, 2.3632, 2.0634, 2.3672], + device='cuda:0'), covar=tensor([0.2040, 0.4181, 0.1561, 0.1060, 0.1606, 0.3024, 0.2236, 0.1585], + device='cuda:0'), in_proj_covar=tensor([0.0072, 0.0075, 0.0064, 0.0059, 0.0058, 0.0059, 0.0098, 0.0061], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:09:33,109 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.211e+02 1.661e+02 1.987e+02 2.340e+02 3.942e+02, threshold=3.975e+02, percent-clipped=0.0 +2023-03-21 07:09:34,666 INFO [train.py:901] (0/2) Epoch 34, batch 450, loss[loss=0.129, simple_loss=0.2101, pruned_loss=0.02402, over 7231.00 frames. ], tot_loss[loss=0.1336, simple_loss=0.2143, pruned_loss=0.02647, over 1291600.49 frames. ], batch size: 45, lr: 4.88e-03, grad_scale: 8.0 +2023-03-21 07:09:39,721 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 07:09:40,224 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 07:09:46,295 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=93666.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:09:52,791 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=93679.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:09:59,413 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=93692.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:09:59,820 INFO [train.py:901] (0/2) Epoch 34, batch 500, loss[loss=0.136, simple_loss=0.2205, pruned_loss=0.02577, over 7272.00 frames. ], tot_loss[loss=0.1333, simple_loss=0.214, pruned_loss=0.02631, over 1322827.08 frames. ], batch size: 70, lr: 4.88e-03, grad_scale: 8.0 +2023-03-21 07:10:10,393 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=93714.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:10:12,011 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 07:10:13,101 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.1082, 5.6423, 5.6771, 5.5535, 5.3881, 5.1904, 5.7511, 5.5076], + device='cuda:0'), covar=tensor([0.0440, 0.0332, 0.0323, 0.0529, 0.0322, 0.0322, 0.0255, 0.0425], + device='cuda:0'), in_proj_covar=tensor([0.0209, 0.0253, 0.0191, 0.0192, 0.0150, 0.0222, 0.0197, 0.0148], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:10:13,544 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 07:10:13,681 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4507, 2.4581, 2.2408, 3.6396, 1.8191, 3.5938, 1.4888, 3.1801], + device='cuda:0'), covar=tensor([0.0208, 0.1417, 0.1809, 0.0209, 0.3820, 0.0247, 0.1218, 0.0400], + device='cuda:0'), in_proj_covar=tensor([0.0201, 0.0252, 0.0265, 0.0204, 0.0252, 0.0209, 0.0232, 0.0224], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 07:10:14,079 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 07:10:16,744 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 07:10:21,171 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 07:10:24,672 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.204e+02 1.702e+02 2.032e+02 2.310e+02 3.971e+02, threshold=4.063e+02, percent-clipped=0.0 +2023-03-21 07:10:24,758 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=93740.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:10:26,215 INFO [train.py:901] (0/2) Epoch 34, batch 550, loss[loss=0.1254, simple_loss=0.2149, pruned_loss=0.01795, over 7217.00 frames. ], tot_loss[loss=0.1336, simple_loss=0.2145, pruned_loss=0.02637, over 1348819.32 frames. ], batch size: 93, lr: 4.88e-03, grad_scale: 8.0 +2023-03-21 07:10:32,098 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 07:10:39,785 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=93770.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:10:40,162 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 07:10:40,241 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=93771.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:10:43,681 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 07:10:50,614 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 07:10:51,077 INFO [train.py:901] (0/2) Epoch 34, batch 600, loss[loss=0.1349, simple_loss=0.2141, pruned_loss=0.02783, over 7225.00 frames. ], tot_loss[loss=0.1335, simple_loss=0.2143, pruned_loss=0.02633, over 1365524.10 frames. ], batch size: 45, lr: 4.88e-03, grad_scale: 8.0 +2023-03-21 07:11:05,772 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=93819.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:11:08,866 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 07:11:12,136 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=93831.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:11:16,354 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.78 vs. limit=2.0 +2023-03-21 07:11:16,509 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.397e+02 1.841e+02 2.134e+02 2.382e+02 3.944e+02, threshold=4.268e+02, percent-clipped=0.0 +2023-03-21 07:11:18,122 INFO [train.py:901] (0/2) Epoch 34, batch 650, loss[loss=0.1363, simple_loss=0.219, pruned_loss=0.02681, over 7238.00 frames. ], tot_loss[loss=0.1332, simple_loss=0.2139, pruned_loss=0.02624, over 1379673.82 frames. ], batch size: 55, lr: 4.88e-03, grad_scale: 8.0 +2023-03-21 07:11:18,148 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 07:11:27,299 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6144, 2.8795, 3.5238, 3.5906, 3.6751, 3.6173, 3.5737, 3.5458], + device='cuda:0'), covar=tensor([0.0027, 0.0118, 0.0040, 0.0036, 0.0032, 0.0037, 0.0047, 0.0059], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0068, 0.0057, 0.0055, 0.0054, 0.0059, 0.0048, 0.0075], + device='cuda:0'), out_proj_covar=tensor([8.1330e-05, 1.4152e-04, 1.0609e-04, 9.6848e-05, 9.5455e-05, 1.0649e-04, + 9.5154e-05, 1.4206e-04], device='cuda:0') +2023-03-21 07:11:28,536 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-03-21 07:11:34,698 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 07:11:43,889 INFO [train.py:901] (0/2) Epoch 34, batch 700, loss[loss=0.1242, simple_loss=0.2081, pruned_loss=0.02009, over 7300.00 frames. ], tot_loss[loss=0.133, simple_loss=0.2136, pruned_loss=0.02616, over 1394180.49 frames. ], batch size: 80, lr: 4.88e-03, grad_scale: 8.0 +2023-03-21 07:11:43,908 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 07:11:52,724 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8241, 1.4709, 1.9393, 2.3561, 2.0350, 2.1846, 1.7460, 2.2475], + device='cuda:0'), covar=tensor([0.1879, 0.4255, 0.1510, 0.0961, 0.2057, 0.1469, 0.2070, 0.1787], + device='cuda:0'), in_proj_covar=tensor([0.0073, 0.0075, 0.0064, 0.0060, 0.0059, 0.0059, 0.0099, 0.0061], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:12:08,218 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.210e+02 1.694e+02 2.006e+02 2.347e+02 4.324e+02, threshold=4.012e+02, percent-clipped=1.0 +2023-03-21 07:12:08,242 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 07:12:08,772 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 07:12:09,767 INFO [train.py:901] (0/2) Epoch 34, batch 750, loss[loss=0.1284, simple_loss=0.2189, pruned_loss=0.01896, over 7329.00 frames. ], tot_loss[loss=0.1326, simple_loss=0.2136, pruned_loss=0.02584, over 1406613.77 frames. ], batch size: 75, lr: 4.88e-03, grad_scale: 8.0 +2023-03-21 07:12:22,922 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 07:12:27,004 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 07:12:30,422 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=93981.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:12:34,815 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 07:12:35,803 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 07:12:36,286 INFO [train.py:901] (0/2) Epoch 34, batch 800, loss[loss=0.1277, simple_loss=0.2109, pruned_loss=0.02227, over 7344.00 frames. ], tot_loss[loss=0.1323, simple_loss=0.2129, pruned_loss=0.02588, over 1412906.15 frames. ], batch size: 44, lr: 4.87e-03, grad_scale: 8.0 +2023-03-21 07:12:46,868 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 07:13:00,409 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.299e+02 1.747e+02 2.063e+02 2.429e+02 4.246e+02, threshold=4.125e+02, percent-clipped=1.0 +2023-03-21 07:13:01,539 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=94042.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:13:01,927 INFO [train.py:901] (0/2) Epoch 34, batch 850, loss[loss=0.1365, simple_loss=0.2181, pruned_loss=0.0275, over 7296.00 frames. ], tot_loss[loss=0.1321, simple_loss=0.2128, pruned_loss=0.02575, over 1417185.32 frames. ], batch size: 68, lr: 4.87e-03, grad_scale: 8.0 +2023-03-21 07:13:05,500 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 07:13:06,018 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 07:13:10,126 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2490, 4.1530, 3.4680, 3.7416, 3.3017, 2.4989, 2.0821, 4.2034], + device='cuda:0'), covar=tensor([0.0044, 0.0043, 0.0129, 0.0069, 0.0137, 0.0454, 0.0550, 0.0048], + device='cuda:0'), in_proj_covar=tensor([0.0108, 0.0088, 0.0110, 0.0093, 0.0122, 0.0131, 0.0126, 0.0100], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 07:13:10,557 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 07:13:15,736 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 07:13:28,328 INFO [train.py:901] (0/2) Epoch 34, batch 900, loss[loss=0.1499, simple_loss=0.2303, pruned_loss=0.03472, over 6844.00 frames. ], tot_loss[loss=0.1326, simple_loss=0.2133, pruned_loss=0.02596, over 1422009.49 frames. ], batch size: 107, lr: 4.87e-03, grad_scale: 8.0 +2023-03-21 07:13:45,074 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=94126.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:13:52,194 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.195e+02 1.701e+02 1.978e+02 2.381e+02 4.265e+02, threshold=3.957e+02, percent-clipped=2.0 +2023-03-21 07:13:53,734 INFO [train.py:901] (0/2) Epoch 34, batch 950, loss[loss=0.123, simple_loss=0.2075, pruned_loss=0.01923, over 7331.00 frames. ], tot_loss[loss=0.1325, simple_loss=0.2133, pruned_loss=0.02587, over 1427366.79 frames. ], batch size: 59, lr: 4.87e-03, grad_scale: 8.0 +2023-03-21 07:13:54,230 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 07:14:15,111 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2516, 3.3024, 2.3082, 3.6990, 2.9335, 3.4362, 1.6588, 2.2598], + device='cuda:0'), covar=tensor([0.0465, 0.0957, 0.2790, 0.0659, 0.0520, 0.0588, 0.3630, 0.1945], + device='cuda:0'), in_proj_covar=tensor([0.0282, 0.0258, 0.0284, 0.0271, 0.0272, 0.0266, 0.0239, 0.0265], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 07:14:18,010 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 07:14:20,034 INFO [train.py:901] (0/2) Epoch 34, batch 1000, loss[loss=0.1393, simple_loss=0.2161, pruned_loss=0.03126, over 7291.00 frames. ], tot_loss[loss=0.1332, simple_loss=0.2142, pruned_loss=0.02606, over 1433283.02 frames. ], batch size: 66, lr: 4.87e-03, grad_scale: 8.0 +2023-03-21 07:14:25,041 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.2010, 5.6783, 5.7029, 5.6179, 5.3977, 5.2882, 5.7524, 5.5731], + device='cuda:0'), covar=tensor([0.0370, 0.0315, 0.0315, 0.0440, 0.0334, 0.0289, 0.0255, 0.0382], + device='cuda:0'), in_proj_covar=tensor([0.0208, 0.0251, 0.0190, 0.0190, 0.0150, 0.0220, 0.0197, 0.0146], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:14:35,702 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=94223.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 07:14:38,125 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 07:14:44,817 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.337e+02 1.717e+02 1.958e+02 2.370e+02 4.573e+02, threshold=3.915e+02, percent-clipped=2.0 +2023-03-21 07:14:45,451 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0441, 3.2888, 3.9992, 3.9652, 4.1499, 4.1039, 4.1495, 3.9566], + device='cuda:0'), covar=tensor([0.0026, 0.0109, 0.0028, 0.0034, 0.0025, 0.0028, 0.0037, 0.0052], + device='cuda:0'), in_proj_covar=tensor([0.0046, 0.0067, 0.0056, 0.0054, 0.0053, 0.0058, 0.0048, 0.0075], + device='cuda:0'), out_proj_covar=tensor([7.9985e-05, 1.3977e-04, 1.0385e-04, 9.5163e-05, 9.3436e-05, 1.0414e-04, + 9.4246e-05, 1.4142e-04], device='cuda:0') +2023-03-21 07:14:46,365 INFO [train.py:901] (0/2) Epoch 34, batch 1050, loss[loss=0.1323, simple_loss=0.2135, pruned_loss=0.02553, over 7349.00 frames. ], tot_loss[loss=0.1339, simple_loss=0.215, pruned_loss=0.0264, over 1436656.40 frames. ], batch size: 63, lr: 4.87e-03, grad_scale: 8.0 +2023-03-21 07:15:00,009 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 07:15:04,470 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 07:15:07,692 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=94284.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 07:15:12,023 INFO [train.py:901] (0/2) Epoch 34, batch 1100, loss[loss=0.1325, simple_loss=0.2197, pruned_loss=0.02261, over 7245.00 frames. ], tot_loss[loss=0.1335, simple_loss=0.2147, pruned_loss=0.0262, over 1436959.62 frames. ], batch size: 93, lr: 4.87e-03, grad_scale: 8.0 +2023-03-21 07:15:32,025 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 07:15:32,584 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 07:15:35,622 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=94337.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:15:37,014 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.224e+02 1.720e+02 1.980e+02 2.346e+02 4.157e+02, threshold=3.960e+02, percent-clipped=1.0 +2023-03-21 07:15:38,528 INFO [train.py:901] (0/2) Epoch 34, batch 1150, loss[loss=0.1421, simple_loss=0.2229, pruned_loss=0.03061, over 7325.00 frames. ], tot_loss[loss=0.1332, simple_loss=0.2143, pruned_loss=0.02599, over 1439049.52 frames. ], batch size: 49, lr: 4.87e-03, grad_scale: 8.0 +2023-03-21 07:15:41,115 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=94348.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:15:44,481 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 07:15:44,983 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 07:15:50,140 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1737, 2.2915, 2.1285, 3.4131, 1.7324, 3.2374, 1.4209, 3.2050], + device='cuda:0'), covar=tensor([0.0206, 0.1391, 0.1854, 0.0194, 0.3856, 0.0173, 0.1323, 0.0265], + device='cuda:0'), in_proj_covar=tensor([0.0201, 0.0250, 0.0264, 0.0205, 0.0252, 0.0208, 0.0233, 0.0224], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 07:16:03,411 INFO [train.py:901] (0/2) Epoch 34, batch 1200, loss[loss=0.1707, simple_loss=0.2411, pruned_loss=0.05016, over 7281.00 frames. ], tot_loss[loss=0.1338, simple_loss=0.2146, pruned_loss=0.02654, over 1439353.48 frames. ], batch size: 47, lr: 4.86e-03, grad_scale: 8.0 +2023-03-21 07:16:12,588 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=94409.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 07:16:18,522 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 07:16:19,638 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4892, 1.6303, 1.5639, 1.5124, 1.8054, 1.4865, 1.4901, 1.2680], + device='cuda:0'), covar=tensor([0.0148, 0.0187, 0.0203, 0.0182, 0.0138, 0.0161, 0.0298, 0.0192], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0033, 0.0032, 0.0034, 0.0032, 0.0031, 0.0035, 0.0040], + device='cuda:0'), out_proj_covar=tensor([3.9118e-05, 3.6496e-05, 3.6454e-05, 3.7231e-05, 3.5927e-05, 3.4889e-05, + 3.9039e-05, 4.4731e-05], device='cuda:0') +2023-03-21 07:16:21,543 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=94426.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:16:28,467 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.293e+02 1.769e+02 2.087e+02 2.412e+02 4.266e+02, threshold=4.175e+02, percent-clipped=1.0 +2023-03-21 07:16:29,986 INFO [train.py:901] (0/2) Epoch 34, batch 1250, loss[loss=0.1416, simple_loss=0.2188, pruned_loss=0.03219, over 7338.00 frames. ], tot_loss[loss=0.1336, simple_loss=0.2144, pruned_loss=0.0264, over 1442746.56 frames. ], batch size: 61, lr: 4.86e-03, grad_scale: 8.0 +2023-03-21 07:16:42,504 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 07:16:45,598 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=94474.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:16:47,070 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 07:16:48,085 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 07:16:55,088 INFO [train.py:901] (0/2) Epoch 34, batch 1300, loss[loss=0.1289, simple_loss=0.2127, pruned_loss=0.0226, over 7291.00 frames. ], tot_loss[loss=0.1338, simple_loss=0.2147, pruned_loss=0.0265, over 1441780.50 frames. ], batch size: 68, lr: 4.86e-03, grad_scale: 8.0 +2023-03-21 07:17:12,397 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 07:17:14,903 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 07:17:18,400 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 07:17:19,910 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.017e+02 1.844e+02 2.157e+02 2.540e+02 4.184e+02, threshold=4.314e+02, percent-clipped=1.0 +2023-03-21 07:17:21,502 INFO [train.py:901] (0/2) Epoch 34, batch 1350, loss[loss=0.142, simple_loss=0.2253, pruned_loss=0.02934, over 7342.00 frames. ], tot_loss[loss=0.1331, simple_loss=0.2137, pruned_loss=0.02629, over 1439879.83 frames. ], batch size: 54, lr: 4.86e-03, grad_scale: 8.0 +2023-03-21 07:17:28,041 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 07:17:33,578 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0850, 3.4196, 3.9820, 3.9387, 4.0199, 4.1749, 4.3165, 3.8918], + device='cuda:0'), covar=tensor([0.0042, 0.0149, 0.0049, 0.0047, 0.0045, 0.0044, 0.0035, 0.0076], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0068, 0.0056, 0.0055, 0.0054, 0.0059, 0.0048, 0.0076], + device='cuda:0'), out_proj_covar=tensor([8.1229e-05, 1.4170e-04, 1.0514e-04, 9.6581e-05, 9.4594e-05, 1.0555e-04, + 9.5458e-05, 1.4311e-04], device='cuda:0') +2023-03-21 07:17:39,608 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=94579.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 07:17:47,199 INFO [train.py:901] (0/2) Epoch 34, batch 1400, loss[loss=0.1322, simple_loss=0.2181, pruned_loss=0.02312, over 7272.00 frames. ], tot_loss[loss=0.1337, simple_loss=0.2146, pruned_loss=0.02643, over 1442687.83 frames. ], batch size: 89, lr: 4.86e-03, grad_scale: 8.0 +2023-03-21 07:18:01,604 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 07:18:10,195 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=94637.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:18:11,546 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.216e+02 1.753e+02 2.058e+02 2.553e+02 5.277e+02, threshold=4.117e+02, percent-clipped=1.0 +2023-03-21 07:18:13,070 INFO [train.py:901] (0/2) Epoch 34, batch 1450, loss[loss=0.1416, simple_loss=0.2277, pruned_loss=0.02775, over 6758.00 frames. ], tot_loss[loss=0.1333, simple_loss=0.214, pruned_loss=0.02629, over 1440873.98 frames. ], batch size: 106, lr: 4.86e-03, grad_scale: 8.0 +2023-03-21 07:18:24,543 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 07:18:35,336 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=94685.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:18:39,199 INFO [train.py:901] (0/2) Epoch 34, batch 1500, loss[loss=0.0972, simple_loss=0.171, pruned_loss=0.01168, over 6992.00 frames. ], tot_loss[loss=0.1338, simple_loss=0.2143, pruned_loss=0.02666, over 1441708.42 frames. ], batch size: 35, lr: 4.86e-03, grad_scale: 8.0 +2023-03-21 07:18:41,387 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7140, 3.3362, 3.4808, 3.6140, 3.1407, 3.0279, 3.6238, 2.6718], + device='cuda:0'), covar=tensor([0.0419, 0.0522, 0.0604, 0.0537, 0.0600, 0.0780, 0.0465, 0.1939], + device='cuda:0'), in_proj_covar=tensor([0.0333, 0.0340, 0.0274, 0.0359, 0.0295, 0.0291, 0.0347, 0.0258], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:18:41,780 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 07:18:44,896 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=94704.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 07:19:02,849 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.092e+02 1.786e+02 2.193e+02 2.482e+02 4.700e+02, threshold=4.385e+02, percent-clipped=1.0 +2023-03-21 07:19:04,331 INFO [train.py:901] (0/2) Epoch 34, batch 1550, loss[loss=0.1116, simple_loss=0.1906, pruned_loss=0.01628, over 7181.00 frames. ], tot_loss[loss=0.1335, simple_loss=0.214, pruned_loss=0.02653, over 1441392.73 frames. ], batch size: 39, lr: 4.85e-03, grad_scale: 8.0 +2023-03-21 07:19:05,385 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 07:19:30,396 INFO [train.py:901] (0/2) Epoch 34, batch 1600, loss[loss=0.143, simple_loss=0.217, pruned_loss=0.03453, over 7352.00 frames. ], tot_loss[loss=0.1341, simple_loss=0.2146, pruned_loss=0.02683, over 1444054.75 frames. ], batch size: 73, lr: 4.85e-03, grad_scale: 8.0 +2023-03-21 07:19:37,399 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 07:19:38,404 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 07:19:41,402 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 07:19:50,971 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 07:19:54,409 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.349e+02 1.787e+02 1.989e+02 2.299e+02 4.264e+02, threshold=3.978e+02, percent-clipped=0.0 +2023-03-21 07:19:54,453 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 07:19:56,591 INFO [train.py:901] (0/2) Epoch 34, batch 1650, loss[loss=0.1339, simple_loss=0.2146, pruned_loss=0.02659, over 7348.00 frames. ], tot_loss[loss=0.1345, simple_loss=0.2151, pruned_loss=0.02694, over 1442242.24 frames. ], batch size: 61, lr: 4.85e-03, grad_scale: 8.0 +2023-03-21 07:20:03,176 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 07:20:07,759 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.61 vs. limit=2.0 +2023-03-21 07:20:08,445 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4274, 4.1019, 3.9795, 3.8744, 3.7040, 2.8168, 2.0841, 4.3703], + device='cuda:0'), covar=tensor([0.0039, 0.0079, 0.0078, 0.0055, 0.0103, 0.0417, 0.0591, 0.0045], + device='cuda:0'), in_proj_covar=tensor([0.0109, 0.0089, 0.0111, 0.0093, 0.0125, 0.0133, 0.0129, 0.0101], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 07:20:15,500 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=94879.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 07:20:20,980 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 07:20:22,476 INFO [train.py:901] (0/2) Epoch 34, batch 1700, loss[loss=0.1446, simple_loss=0.2276, pruned_loss=0.03084, over 7269.00 frames. ], tot_loss[loss=0.134, simple_loss=0.2147, pruned_loss=0.02662, over 1442815.31 frames. ], batch size: 52, lr: 4.85e-03, grad_scale: 16.0 +2023-03-21 07:20:25,089 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 07:20:25,485 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.61 vs. limit=2.0 +2023-03-21 07:20:30,716 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9486, 2.5776, 3.1956, 2.9077, 3.1418, 2.8418, 2.5118, 3.1108], + device='cuda:0'), covar=tensor([0.1545, 0.0976, 0.0947, 0.1552, 0.0793, 0.1193, 0.2152, 0.1363], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0065, 0.0049, 0.0049, 0.0047, 0.0049, 0.0066, 0.0049], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 07:20:32,672 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8677, 3.0097, 3.7275, 3.6685, 3.8281, 3.8305, 3.9366, 3.5803], + device='cuda:0'), covar=tensor([0.0043, 0.0178, 0.0052, 0.0059, 0.0041, 0.0050, 0.0055, 0.0083], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0068, 0.0056, 0.0054, 0.0054, 0.0059, 0.0048, 0.0075], + device='cuda:0'), out_proj_covar=tensor([8.1202e-05, 1.4155e-04, 1.0467e-04, 9.6071e-05, 9.4692e-05, 1.0613e-04, + 9.4574e-05, 1.4330e-04], device='cuda:0') +2023-03-21 07:20:35,552 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 07:20:39,573 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=94927.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 07:20:46,780 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.214e+02 1.668e+02 1.961e+02 2.375e+02 4.314e+02, threshold=3.923e+02, percent-clipped=2.0 +2023-03-21 07:20:48,256 INFO [train.py:901] (0/2) Epoch 34, batch 1750, loss[loss=0.1177, simple_loss=0.1925, pruned_loss=0.02142, over 6969.00 frames. ], tot_loss[loss=0.1338, simple_loss=0.2146, pruned_loss=0.02644, over 1444625.30 frames. ], batch size: 35, lr: 4.85e-03, grad_scale: 16.0 +2023-03-21 07:21:01,622 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 07:21:02,584 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 07:21:13,882 INFO [train.py:901] (0/2) Epoch 34, batch 1800, loss[loss=0.1457, simple_loss=0.2262, pruned_loss=0.03264, over 7308.00 frames. ], tot_loss[loss=0.1338, simple_loss=0.2147, pruned_loss=0.02642, over 1445402.46 frames. ], batch size: 80, lr: 4.85e-03, grad_scale: 8.0 +2023-03-21 07:21:18,856 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1961, 3.8029, 4.2868, 4.1707, 4.3828, 4.3396, 4.3111, 4.2134], + device='cuda:0'), covar=tensor([0.0034, 0.0095, 0.0040, 0.0040, 0.0033, 0.0034, 0.0035, 0.0051], + device='cuda:0'), in_proj_covar=tensor([0.0046, 0.0068, 0.0056, 0.0054, 0.0053, 0.0059, 0.0048, 0.0075], + device='cuda:0'), out_proj_covar=tensor([8.0816e-05, 1.4080e-04, 1.0378e-04, 9.5762e-05, 9.4141e-05, 1.0599e-04, + 9.4925e-05, 1.4235e-04], device='cuda:0') +2023-03-21 07:21:19,878 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=95004.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:21:24,852 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 07:21:39,031 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 07:21:39,513 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.226e+02 1.793e+02 2.137e+02 2.599e+02 4.200e+02, threshold=4.274e+02, percent-clipped=3.0 +2023-03-21 07:21:40,603 INFO [train.py:901] (0/2) Epoch 34, batch 1850, loss[loss=0.1375, simple_loss=0.2198, pruned_loss=0.02764, over 7337.00 frames. ], tot_loss[loss=0.1336, simple_loss=0.2146, pruned_loss=0.0263, over 1445810.87 frames. ], batch size: 54, lr: 4.85e-03, grad_scale: 8.0 +2023-03-21 07:21:45,126 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=95052.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:21:48,006 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 07:21:49,627 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0907, 2.1633, 2.3953, 2.1283, 2.2187, 2.1915, 2.0394, 1.7268], + device='cuda:0'), covar=tensor([0.0442, 0.0363, 0.0170, 0.0301, 0.0384, 0.0405, 0.0319, 0.0424], + device='cuda:0'), in_proj_covar=tensor([0.0036, 0.0034, 0.0036, 0.0035, 0.0034, 0.0033, 0.0037, 0.0037], + device='cuda:0'), out_proj_covar=tensor([9.1157e-05, 8.8956e-05, 9.0533e-05, 8.7501e-05, 8.7752e-05, 8.6507e-05, + 9.2303e-05, 9.3928e-05], device='cuda:0') +2023-03-21 07:22:04,948 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 07:22:05,430 INFO [train.py:901] (0/2) Epoch 34, batch 1900, loss[loss=0.1218, simple_loss=0.1997, pruned_loss=0.02198, over 7336.00 frames. ], tot_loss[loss=0.134, simple_loss=0.2148, pruned_loss=0.0266, over 1447079.37 frames. ], batch size: 44, lr: 4.85e-03, grad_scale: 8.0 +2023-03-21 07:22:12,137 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=95105.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 07:22:30,150 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 07:22:30,536 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.091e+02 1.839e+02 2.066e+02 2.358e+02 4.845e+02, threshold=4.133e+02, percent-clipped=1.0 +2023-03-21 07:22:31,616 INFO [train.py:901] (0/2) Epoch 34, batch 1950, loss[loss=0.1353, simple_loss=0.2191, pruned_loss=0.02579, over 7323.00 frames. ], tot_loss[loss=0.1342, simple_loss=0.2151, pruned_loss=0.02668, over 1447417.87 frames. ], batch size: 83, lr: 4.84e-03, grad_scale: 8.0 +2023-03-21 07:22:41,067 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 07:22:43,189 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=95166.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 07:22:46,063 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 07:22:46,572 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 07:22:57,221 INFO [train.py:901] (0/2) Epoch 34, batch 2000, loss[loss=0.1322, simple_loss=0.2061, pruned_loss=0.02912, over 7221.00 frames. ], tot_loss[loss=0.1341, simple_loss=0.2147, pruned_loss=0.02678, over 1447222.93 frames. ], batch size: 45, lr: 4.84e-03, grad_scale: 8.0 +2023-03-21 07:23:03,150 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 07:23:11,891 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9738, 5.4924, 5.5499, 5.4721, 5.2286, 5.0366, 5.5995, 5.3813], + device='cuda:0'), covar=tensor([0.0344, 0.0360, 0.0326, 0.0400, 0.0295, 0.0341, 0.0271, 0.0407], + device='cuda:0'), in_proj_covar=tensor([0.0206, 0.0246, 0.0187, 0.0186, 0.0146, 0.0219, 0.0194, 0.0143], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:23:14,843 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 07:23:22,279 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.193e+02 1.722e+02 2.097e+02 2.467e+02 4.921e+02, threshold=4.194e+02, percent-clipped=3.0 +2023-03-21 07:23:22,820 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 07:23:23,286 INFO [train.py:901] (0/2) Epoch 34, batch 2050, loss[loss=0.1363, simple_loss=0.2179, pruned_loss=0.02738, over 7341.00 frames. ], tot_loss[loss=0.1343, simple_loss=0.2151, pruned_loss=0.02675, over 1447568.66 frames. ], batch size: 61, lr: 4.84e-03, grad_scale: 8.0 +2023-03-21 07:23:31,981 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=95260.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:23:36,108 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=95268.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:23:43,223 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9007, 4.0701, 3.8380, 4.0537, 3.7338, 4.0249, 4.3524, 4.3507], + device='cuda:0'), covar=tensor([0.0230, 0.0152, 0.0213, 0.0164, 0.0338, 0.0265, 0.0204, 0.0177], + device='cuda:0'), in_proj_covar=tensor([0.0119, 0.0123, 0.0116, 0.0118, 0.0111, 0.0100, 0.0096, 0.0097], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:23:49,090 INFO [train.py:901] (0/2) Epoch 34, batch 2100, loss[loss=0.1258, simple_loss=0.2059, pruned_loss=0.02283, over 7364.00 frames. ], tot_loss[loss=0.1344, simple_loss=0.2154, pruned_loss=0.02671, over 1448108.95 frames. ], batch size: 73, lr: 4.84e-03, grad_scale: 8.0 +2023-03-21 07:23:56,736 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 07:23:59,729 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 07:24:04,000 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=95321.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:24:08,035 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=95329.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:24:13,002 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=95339.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:24:13,820 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.277e+02 1.746e+02 2.026e+02 2.348e+02 4.750e+02, threshold=4.051e+02, percent-clipped=2.0 +2023-03-21 07:24:14,730 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-03-21 07:24:14,890 INFO [train.py:901] (0/2) Epoch 34, batch 2150, loss[loss=0.1289, simple_loss=0.2126, pruned_loss=0.02253, over 7300.00 frames. ], tot_loss[loss=0.1342, simple_loss=0.2148, pruned_loss=0.02675, over 1445830.56 frames. ], batch size: 77, lr: 4.84e-03, grad_scale: 8.0 +2023-03-21 07:24:37,328 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7114, 3.9314, 3.6151, 3.8879, 3.5696, 3.7856, 4.1524, 4.1687], + device='cuda:0'), covar=tensor([0.0253, 0.0154, 0.0264, 0.0185, 0.0349, 0.0394, 0.0270, 0.0217], + device='cuda:0'), in_proj_covar=tensor([0.0120, 0.0124, 0.0118, 0.0120, 0.0112, 0.0101, 0.0098, 0.0099], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:24:41,433 INFO [train.py:901] (0/2) Epoch 34, batch 2200, loss[loss=0.111, simple_loss=0.1807, pruned_loss=0.0206, over 6963.00 frames. ], tot_loss[loss=0.1338, simple_loss=0.2145, pruned_loss=0.02654, over 1444944.71 frames. ], batch size: 35, lr: 4.84e-03, grad_scale: 8.0 +2023-03-21 07:24:45,471 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=95400.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:24:45,846 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 07:24:52,093 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.27 vs. limit=2.0 +2023-03-21 07:25:05,777 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.237e+02 1.719e+02 1.925e+02 2.336e+02 3.673e+02, threshold=3.850e+02, percent-clipped=0.0 +2023-03-21 07:25:06,822 INFO [train.py:901] (0/2) Epoch 34, batch 2250, loss[loss=0.1418, simple_loss=0.2267, pruned_loss=0.02842, over 6587.00 frames. ], tot_loss[loss=0.1338, simple_loss=0.2145, pruned_loss=0.0265, over 1444347.18 frames. ], batch size: 106, lr: 4.84e-03, grad_scale: 8.0 +2023-03-21 07:25:16,508 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=95461.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 07:25:20,921 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 07:25:20,932 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 07:25:32,908 INFO [train.py:901] (0/2) Epoch 34, batch 2300, loss[loss=0.1323, simple_loss=0.2122, pruned_loss=0.02621, over 7256.00 frames. ], tot_loss[loss=0.1342, simple_loss=0.2148, pruned_loss=0.02678, over 1444358.94 frames. ], batch size: 55, lr: 4.84e-03, grad_scale: 8.0 +2023-03-21 07:25:33,880 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 07:25:57,306 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.302e+02 1.783e+02 2.106e+02 2.494e+02 4.244e+02, threshold=4.213e+02, percent-clipped=2.0 +2023-03-21 07:25:58,309 INFO [train.py:901] (0/2) Epoch 34, batch 2350, loss[loss=0.1217, simple_loss=0.2093, pruned_loss=0.01701, over 7271.00 frames. ], tot_loss[loss=0.1339, simple_loss=0.2143, pruned_loss=0.02673, over 1442125.52 frames. ], batch size: 64, lr: 4.83e-03, grad_scale: 8.0 +2023-03-21 07:26:03,697 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-21 07:26:21,155 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 07:26:24,174 INFO [train.py:901] (0/2) Epoch 34, batch 2400, loss[loss=0.1445, simple_loss=0.2275, pruned_loss=0.03073, over 7360.00 frames. ], tot_loss[loss=0.134, simple_loss=0.2145, pruned_loss=0.02674, over 1443289.93 frames. ], batch size: 54, lr: 4.83e-03, grad_scale: 8.0 +2023-03-21 07:26:27,196 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 07:26:36,106 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=95616.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:26:36,693 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=95617.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:26:37,599 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 07:26:37,689 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0846, 3.6096, 3.7499, 3.7756, 3.7237, 3.6041, 3.9523, 3.4783], + device='cuda:0'), covar=tensor([0.0119, 0.0205, 0.0124, 0.0178, 0.0415, 0.0125, 0.0149, 0.0200], + device='cuda:0'), in_proj_covar=tensor([0.0092, 0.0100, 0.0099, 0.0087, 0.0172, 0.0105, 0.0104, 0.0109], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:26:40,721 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 07:26:40,774 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=95624.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:26:49,095 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.337e+02 1.742e+02 2.021e+02 2.360e+02 4.581e+02, threshold=4.042e+02, percent-clipped=1.0 +2023-03-21 07:26:50,155 INFO [train.py:901] (0/2) Epoch 34, batch 2450, loss[loss=0.1291, simple_loss=0.2111, pruned_loss=0.02356, over 7312.00 frames. ], tot_loss[loss=0.1339, simple_loss=0.2147, pruned_loss=0.02658, over 1444045.80 frames. ], batch size: 83, lr: 4.83e-03, grad_scale: 8.0 +2023-03-21 07:26:59,357 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2221, 3.5709, 3.0700, 3.3226, 3.1495, 2.8915, 3.5068, 3.2992], + device='cuda:0'), covar=tensor([0.0915, 0.1348, 0.1235, 0.1275, 0.2105, 0.0765, 0.0818, 0.0819], + device='cuda:0'), in_proj_covar=tensor([0.0057, 0.0057, 0.0065, 0.0057, 0.0055, 0.0060, 0.0055, 0.0054], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:26:59,861 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7616, 1.9932, 2.2355, 2.1356, 2.1092, 2.0923, 2.0257, 1.8567], + device='cuda:0'), covar=tensor([0.1287, 0.0641, 0.0404, 0.0330, 0.0662, 0.0593, 0.0401, 0.0432], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0035, 0.0037, 0.0036, 0.0034, 0.0034, 0.0039, 0.0038], + device='cuda:0'), out_proj_covar=tensor([9.5254e-05, 9.1917e-05, 9.3758e-05, 9.0302e-05, 9.0246e-05, 8.8911e-05, + 9.6160e-05, 9.7181e-05], device='cuda:0') +2023-03-21 07:27:06,771 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 07:27:08,410 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=95678.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:27:15,808 INFO [train.py:901] (0/2) Epoch 34, batch 2500, loss[loss=0.137, simple_loss=0.2185, pruned_loss=0.02773, over 7298.00 frames. ], tot_loss[loss=0.1339, simple_loss=0.215, pruned_loss=0.02644, over 1443344.05 frames. ], batch size: 80, lr: 4.83e-03, grad_scale: 8.0 +2023-03-21 07:27:16,815 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=95695.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:27:25,701 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.64 vs. limit=2.0 +2023-03-21 07:27:31,218 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 07:27:40,891 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.147e+02 1.816e+02 2.143e+02 2.532e+02 3.542e+02, threshold=4.287e+02, percent-clipped=0.0 +2023-03-21 07:27:41,903 INFO [train.py:901] (0/2) Epoch 34, batch 2550, loss[loss=0.1302, simple_loss=0.2093, pruned_loss=0.02552, over 7302.00 frames. ], tot_loss[loss=0.1342, simple_loss=0.2151, pruned_loss=0.02664, over 1443046.55 frames. ], batch size: 49, lr: 4.83e-03, grad_scale: 8.0 +2023-03-21 07:27:51,005 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=95761.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 07:28:06,903 INFO [train.py:901] (0/2) Epoch 34, batch 2600, loss[loss=0.1443, simple_loss=0.2312, pruned_loss=0.0287, over 6593.00 frames. ], tot_loss[loss=0.134, simple_loss=0.215, pruned_loss=0.02648, over 1444213.60 frames. ], batch size: 106, lr: 4.83e-03, grad_scale: 8.0 +2023-03-21 07:28:08,155 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.23 vs. limit=2.0 +2023-03-21 07:28:15,793 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=95809.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 07:28:17,391 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=95812.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 07:28:31,405 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.165e+02 1.727e+02 2.112e+02 2.461e+02 5.012e+02, threshold=4.224e+02, percent-clipped=2.0 +2023-03-21 07:28:32,389 INFO [train.py:901] (0/2) Epoch 34, batch 2650, loss[loss=0.141, simple_loss=0.2231, pruned_loss=0.02942, over 7328.00 frames. ], tot_loss[loss=0.1331, simple_loss=0.2143, pruned_loss=0.02598, over 1443421.67 frames. ], batch size: 83, lr: 4.83e-03, grad_scale: 8.0 +2023-03-21 07:28:34,423 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=95847.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:28:45,564 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2348, 2.5542, 2.3212, 2.5061, 2.5689, 2.0623, 2.6282, 2.4146], + device='cuda:0'), covar=tensor([0.0933, 0.0722, 0.0991, 0.1083, 0.0721, 0.0996, 0.0568, 0.1212], + device='cuda:0'), in_proj_covar=tensor([0.0057, 0.0058, 0.0066, 0.0057, 0.0055, 0.0059, 0.0055, 0.0054], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:28:47,109 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=95873.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 07:28:51,462 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6807, 5.2016, 5.2967, 5.1770, 4.9817, 4.7050, 5.2456, 5.0654], + device='cuda:0'), covar=tensor([0.0418, 0.0357, 0.0328, 0.0434, 0.0346, 0.0330, 0.0322, 0.0415], + device='cuda:0'), in_proj_covar=tensor([0.0208, 0.0246, 0.0188, 0.0188, 0.0149, 0.0222, 0.0197, 0.0143], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:28:56,651 INFO [train.py:901] (0/2) Epoch 34, batch 2700, loss[loss=0.1417, simple_loss=0.2212, pruned_loss=0.03105, over 7252.00 frames. ], tot_loss[loss=0.1337, simple_loss=0.215, pruned_loss=0.0262, over 1443725.51 frames. ], batch size: 89, lr: 4.83e-03, grad_scale: 8.0 +2023-03-21 07:28:59,930 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.96 vs. limit=2.0 +2023-03-21 07:29:04,729 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=95908.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 07:29:08,647 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=95916.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:29:12,597 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=95924.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:29:20,856 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.273e+02 1.771e+02 2.102e+02 2.649e+02 3.915e+02, threshold=4.205e+02, percent-clipped=0.0 +2023-03-21 07:29:21,973 INFO [train.py:901] (0/2) Epoch 34, batch 2750, loss[loss=0.127, simple_loss=0.2084, pruned_loss=0.02278, over 7175.00 frames. ], tot_loss[loss=0.1342, simple_loss=0.2154, pruned_loss=0.02648, over 1445362.04 frames. ], batch size: 39, lr: 4.82e-03, grad_scale: 8.0 +2023-03-21 07:29:32,410 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=95964.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:29:36,408 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=95972.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:29:36,894 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=95973.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:29:40,388 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=95980.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:29:46,643 INFO [train.py:901] (0/2) Epoch 34, batch 2800, loss[loss=0.1366, simple_loss=0.2274, pruned_loss=0.0229, over 7137.00 frames. ], tot_loss[loss=0.1345, simple_loss=0.2155, pruned_loss=0.02674, over 1444035.66 frames. ], batch size: 98, lr: 4.82e-03, grad_scale: 8.0 +2023-03-21 07:29:47,682 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=95995.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:29:49,124 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9822, 4.5375, 4.3691, 4.9365, 4.8083, 4.9354, 4.4852, 4.5714], + device='cuda:0'), covar=tensor([0.0834, 0.2388, 0.2251, 0.1103, 0.0851, 0.1089, 0.0700, 0.1141], + device='cuda:0'), in_proj_covar=tensor([0.0251, 0.0380, 0.0292, 0.0303, 0.0224, 0.0358, 0.0224, 0.0267], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:29:50,395 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-96000.pt +2023-03-21 07:29:59,161 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5613, 2.4931, 2.3977, 3.7878, 1.8567, 3.4640, 1.4440, 3.2752], + device='cuda:0'), covar=tensor([0.0190, 0.1275, 0.1768, 0.0181, 0.3915, 0.0191, 0.1195, 0.0314], + device='cuda:0'), in_proj_covar=tensor([0.0203, 0.0249, 0.0263, 0.0207, 0.0251, 0.0210, 0.0229, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 07:30:03,280 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-34.pt +2023-03-21 07:30:20,967 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 07:30:21,022 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 07:30:21,048 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 07:30:21,065 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 07:30:21,240 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 07:30:21,256 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 07:30:21,266 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 07:30:21,366 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 07:30:21,440 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 07:30:21,801 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 07:30:21,818 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 07:30:21,981 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 07:30:21,986 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 07:30:21,994 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 07:30:22,133 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 07:30:22,136 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 07:30:22,194 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 07:30:24,332 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.69 vs. limit=2.0 +2023-03-21 07:30:24,387 INFO [train.py:901] (0/2) Epoch 35, batch 0, loss[loss=0.1245, simple_loss=0.2128, pruned_loss=0.01807, over 7308.00 frames. ], tot_loss[loss=0.1245, simple_loss=0.2128, pruned_loss=0.01807, over 7308.00 frames. ], batch size: 59, lr: 4.75e-03, grad_scale: 8.0 +2023-03-21 07:30:24,388 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 07:30:50,188 INFO [train.py:935] (0/2) Epoch 35, validation: loss=0.1643, simple_loss=0.2559, pruned_loss=0.03642, over 1622729.00 frames. +2023-03-21 07:30:50,189 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 07:30:56,276 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 07:31:02,948 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.206e+02 1.782e+02 1.999e+02 2.365e+02 3.842e+02, threshold=3.998e+02, percent-clipped=0.0 +2023-03-21 07:31:03,134 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=96041.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:31:04,032 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=96043.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:31:07,478 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 07:31:15,047 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 07:31:16,032 INFO [train.py:901] (0/2) Epoch 35, batch 50, loss[loss=0.1202, simple_loss=0.2064, pruned_loss=0.01701, over 7294.00 frames. ], tot_loss[loss=0.1319, simple_loss=0.213, pruned_loss=0.02537, over 324112.45 frames. ], batch size: 86, lr: 4.75e-03, grad_scale: 8.0 +2023-03-21 07:31:17,090 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 07:31:20,615 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 07:31:22,799 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=96080.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:31:37,956 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 07:31:38,436 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 07:31:41,927 INFO [train.py:901] (0/2) Epoch 35, batch 100, loss[loss=0.1406, simple_loss=0.2175, pruned_loss=0.03182, over 7276.00 frames. ], tot_loss[loss=0.1338, simple_loss=0.2146, pruned_loss=0.02654, over 572495.54 frames. ], batch size: 47, lr: 4.75e-03, grad_scale: 8.0 +2023-03-21 07:31:54,360 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.190e+02 1.736e+02 2.019e+02 2.220e+02 3.801e+02, threshold=4.038e+02, percent-clipped=0.0 +2023-03-21 07:31:54,560 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=96141.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:32:06,567 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=96165.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:32:07,446 INFO [train.py:901] (0/2) Epoch 35, batch 150, loss[loss=0.1534, simple_loss=0.231, pruned_loss=0.03787, over 7330.00 frames. ], tot_loss[loss=0.1345, simple_loss=0.2152, pruned_loss=0.0269, over 767582.37 frames. ], batch size: 54, lr: 4.75e-03, grad_scale: 8.0 +2023-03-21 07:32:08,028 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=96168.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 07:32:26,413 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=96203.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 07:32:33,889 INFO [train.py:901] (0/2) Epoch 35, batch 200, loss[loss=0.1172, simple_loss=0.2015, pruned_loss=0.01638, over 7284.00 frames. ], tot_loss[loss=0.134, simple_loss=0.215, pruned_loss=0.02656, over 917456.02 frames. ], batch size: 66, lr: 4.75e-03, grad_scale: 8.0 +2023-03-21 07:32:38,609 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=96226.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:32:41,063 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 07:32:45,097 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 07:32:46,084 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.234e+02 1.725e+02 2.081e+02 2.460e+02 5.560e+02, threshold=4.161e+02, percent-clipped=1.0 +2023-03-21 07:32:52,149 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 07:32:59,847 INFO [train.py:901] (0/2) Epoch 35, batch 250, loss[loss=0.1262, simple_loss=0.2101, pruned_loss=0.02113, over 7305.00 frames. ], tot_loss[loss=0.1336, simple_loss=0.2145, pruned_loss=0.02638, over 1032684.34 frames. ], batch size: 80, lr: 4.75e-03, grad_scale: 8.0 +2023-03-21 07:33:03,151 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=96273.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:33:05,567 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 07:33:25,520 INFO [train.py:901] (0/2) Epoch 35, batch 300, loss[loss=0.1275, simple_loss=0.2171, pruned_loss=0.01895, over 7365.00 frames. ], tot_loss[loss=0.1333, simple_loss=0.214, pruned_loss=0.02636, over 1124190.82 frames. ], batch size: 63, lr: 4.75e-03, grad_scale: 8.0 +2023-03-21 07:33:26,006 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 07:33:27,587 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=96321.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:33:29,210 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=96324.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:33:35,141 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=96336.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:33:35,587 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 07:33:37,565 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.130e+02 1.756e+02 2.059e+02 2.406e+02 3.966e+02, threshold=4.119e+02, percent-clipped=0.0 +2023-03-21 07:33:38,775 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2899, 3.0541, 3.1138, 3.1910, 2.8026, 2.6807, 3.2260, 2.4128], + device='cuda:0'), covar=tensor([0.0600, 0.0529, 0.0654, 0.0661, 0.0658, 0.0832, 0.0596, 0.2051], + device='cuda:0'), in_proj_covar=tensor([0.0336, 0.0336, 0.0274, 0.0357, 0.0293, 0.0291, 0.0347, 0.0255], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:33:50,929 INFO [train.py:901] (0/2) Epoch 35, batch 350, loss[loss=0.1312, simple_loss=0.2151, pruned_loss=0.02368, over 7266.00 frames. ], tot_loss[loss=0.1331, simple_loss=0.2132, pruned_loss=0.02646, over 1194254.13 frames. ], batch size: 77, lr: 4.74e-03, grad_scale: 8.0 +2023-03-21 07:33:59,801 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=96385.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:34:00,097 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.63 vs. limit=2.0 +2023-03-21 07:34:02,287 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=96390.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:34:09,992 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 07:34:16,481 INFO [train.py:901] (0/2) Epoch 35, batch 400, loss[loss=0.142, simple_loss=0.2252, pruned_loss=0.02942, over 7277.00 frames. ], tot_loss[loss=0.1342, simple_loss=0.2147, pruned_loss=0.02684, over 1251214.70 frames. ], batch size: 77, lr: 4.74e-03, grad_scale: 8.0 +2023-03-21 07:34:26,180 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=96436.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:34:28,584 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.391e+02 1.843e+02 2.067e+02 2.365e+02 4.098e+02, threshold=4.134e+02, percent-clipped=0.0 +2023-03-21 07:34:33,674 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4019, 3.9873, 4.0501, 4.0501, 4.0426, 3.9860, 4.2630, 3.7806], + device='cuda:0'), covar=tensor([0.0115, 0.0164, 0.0110, 0.0181, 0.0411, 0.0111, 0.0132, 0.0174], + device='cuda:0'), in_proj_covar=tensor([0.0093, 0.0100, 0.0100, 0.0089, 0.0173, 0.0107, 0.0105, 0.0110], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:34:34,239 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=96451.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:34:38,415 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.74 vs. limit=2.0 +2023-03-21 07:34:38,934 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-03-21 07:34:42,143 INFO [train.py:901] (0/2) Epoch 35, batch 450, loss[loss=0.1249, simple_loss=0.2016, pruned_loss=0.02409, over 7282.00 frames. ], tot_loss[loss=0.1334, simple_loss=0.2138, pruned_loss=0.02651, over 1295211.40 frames. ], batch size: 47, lr: 4.74e-03, grad_scale: 8.0 +2023-03-21 07:34:42,781 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=96468.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 07:34:51,680 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 07:34:52,190 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 07:34:52,789 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5448, 4.1274, 4.2578, 4.2426, 4.2633, 4.2141, 4.5025, 3.9191], + device='cuda:0'), covar=tensor([0.0130, 0.0146, 0.0106, 0.0173, 0.0396, 0.0121, 0.0131, 0.0187], + device='cuda:0'), in_proj_covar=tensor([0.0093, 0.0100, 0.0100, 0.0089, 0.0173, 0.0107, 0.0105, 0.0110], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:35:00,872 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=96503.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:35:07,385 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=96516.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 07:35:07,800 INFO [train.py:901] (0/2) Epoch 35, batch 500, loss[loss=0.1195, simple_loss=0.2061, pruned_loss=0.01646, over 7140.00 frames. ], tot_loss[loss=0.1335, simple_loss=0.2141, pruned_loss=0.02645, over 1329146.40 frames. ], batch size: 41, lr: 4.74e-03, grad_scale: 8.0 +2023-03-21 07:35:09,823 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=96521.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:35:20,463 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.185e+02 1.758e+02 2.053e+02 2.649e+02 4.888e+02, threshold=4.106e+02, percent-clipped=1.0 +2023-03-21 07:35:23,980 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6540, 4.2521, 4.4001, 4.0788, 3.9105, 2.7752, 2.4390, 4.6413], + device='cuda:0'), covar=tensor([0.0040, 0.0115, 0.0060, 0.0061, 0.0093, 0.0468, 0.0509, 0.0039], + device='cuda:0'), in_proj_covar=tensor([0.0110, 0.0090, 0.0111, 0.0095, 0.0126, 0.0133, 0.0130, 0.0103], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 07:35:24,185 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.67 vs. limit=2.0 +2023-03-21 07:35:24,401 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 07:35:25,425 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=96551.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:35:25,872 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 07:35:26,853 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 07:35:28,346 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 07:35:33,899 INFO [train.py:901] (0/2) Epoch 35, batch 550, loss[loss=0.1257, simple_loss=0.2098, pruned_loss=0.02078, over 7289.00 frames. ], tot_loss[loss=0.1335, simple_loss=0.214, pruned_loss=0.02649, over 1353660.02 frames. ], batch size: 66, lr: 4.74e-03, grad_scale: 8.0 +2023-03-21 07:35:33,929 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 07:35:44,355 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 07:35:52,729 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 07:35:55,651 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 07:35:59,823 INFO [train.py:901] (0/2) Epoch 35, batch 600, loss[loss=0.1411, simple_loss=0.2223, pruned_loss=0.03002, over 7277.00 frames. ], tot_loss[loss=0.1337, simple_loss=0.2145, pruned_loss=0.02647, over 1375862.60 frames. ], batch size: 66, lr: 4.74e-03, grad_scale: 8.0 +2023-03-21 07:36:03,286 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 07:36:09,401 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=96636.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:36:11,803 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.247e+02 1.802e+02 2.034e+02 2.397e+02 4.474e+02, threshold=4.067e+02, percent-clipped=1.0 +2023-03-21 07:36:19,412 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 07:36:25,456 INFO [train.py:901] (0/2) Epoch 35, batch 650, loss[loss=0.1181, simple_loss=0.2063, pruned_loss=0.01494, over 7322.00 frames. ], tot_loss[loss=0.1331, simple_loss=0.214, pruned_loss=0.02614, over 1391595.30 frames. ], batch size: 59, lr: 4.74e-03, grad_scale: 8.0 +2023-03-21 07:36:27,954 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 07:36:32,107 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=96680.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:36:34,078 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=96684.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:36:46,095 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 07:36:48,184 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5955, 3.8064, 3.5838, 3.7880, 3.4250, 3.7212, 4.0459, 4.0764], + device='cuda:0'), covar=tensor([0.0248, 0.0161, 0.0229, 0.0190, 0.0422, 0.0413, 0.0213, 0.0161], + device='cuda:0'), in_proj_covar=tensor([0.0118, 0.0123, 0.0116, 0.0118, 0.0111, 0.0101, 0.0096, 0.0096], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:36:49,797 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.65 vs. limit=5.0 +2023-03-21 07:36:51,027 INFO [train.py:901] (0/2) Epoch 35, batch 700, loss[loss=0.1401, simple_loss=0.2324, pruned_loss=0.02389, over 7288.00 frames. ], tot_loss[loss=0.1329, simple_loss=0.2136, pruned_loss=0.02607, over 1399526.92 frames. ], batch size: 68, lr: 4.74e-03, grad_scale: 8.0 +2023-03-21 07:36:54,512 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 07:37:01,204 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=96736.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:37:03,571 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.274e+02 1.655e+02 1.913e+02 2.352e+02 3.396e+02, threshold=3.826e+02, percent-clipped=0.0 +2023-03-21 07:37:04,646 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.2460, 4.8296, 4.6246, 5.2731, 5.0507, 5.2052, 4.5930, 4.8163], + device='cuda:0'), covar=tensor([0.0599, 0.1974, 0.1898, 0.0756, 0.0809, 0.0914, 0.0619, 0.1020], + device='cuda:0'), in_proj_covar=tensor([0.0246, 0.0378, 0.0289, 0.0298, 0.0222, 0.0355, 0.0219, 0.0263], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:37:06,179 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=96746.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:37:07,217 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6088, 2.8417, 2.4650, 3.6809, 2.0369, 3.5808, 1.4949, 3.0594], + device='cuda:0'), covar=tensor([0.0154, 0.1086, 0.1827, 0.0204, 0.3791, 0.0277, 0.1337, 0.0618], + device='cuda:0'), in_proj_covar=tensor([0.0205, 0.0253, 0.0268, 0.0212, 0.0254, 0.0215, 0.0232, 0.0231], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 07:37:16,481 INFO [train.py:901] (0/2) Epoch 35, batch 750, loss[loss=0.1311, simple_loss=0.2162, pruned_loss=0.02296, over 7353.00 frames. ], tot_loss[loss=0.1328, simple_loss=0.2136, pruned_loss=0.02601, over 1409913.11 frames. ], batch size: 73, lr: 4.73e-03, grad_scale: 8.0 +2023-03-21 07:37:19,009 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 07:37:19,533 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 07:37:21,162 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=96776.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:37:25,051 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=96784.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:37:30,614 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8973, 2.2968, 1.7180, 2.0031, 2.0616, 1.9172, 1.9203, 1.6865], + device='cuda:0'), covar=tensor([0.0128, 0.0091, 0.0258, 0.0126, 0.0104, 0.0137, 0.0148, 0.0140], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0033, 0.0032, 0.0034, 0.0033, 0.0031, 0.0035, 0.0040], + device='cuda:0'), out_proj_covar=tensor([3.9415e-05, 3.6281e-05, 3.6798e-05, 3.7692e-05, 3.6393e-05, 3.5067e-05, + 3.9136e-05, 4.4871e-05], device='cuda:0') +2023-03-21 07:37:33,913 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 07:37:38,810 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 07:37:42,329 INFO [train.py:901] (0/2) Epoch 35, batch 800, loss[loss=0.1471, simple_loss=0.229, pruned_loss=0.03258, over 7120.00 frames. ], tot_loss[loss=0.1323, simple_loss=0.2129, pruned_loss=0.02584, over 1415836.29 frames. ], batch size: 98, lr: 4.73e-03, grad_scale: 8.0 +2023-03-21 07:37:45,018 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=96821.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:37:45,403 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 07:37:46,410 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 07:37:51,506 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6725, 1.6285, 1.8591, 2.1771, 1.8156, 1.9364, 1.3868, 1.9762], + device='cuda:0'), covar=tensor([0.2003, 0.3662, 0.1293, 0.1271, 0.1543, 0.1802, 0.1646, 0.1704], + device='cuda:0'), in_proj_covar=tensor([0.0073, 0.0076, 0.0065, 0.0059, 0.0058, 0.0060, 0.0100, 0.0061], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:37:52,982 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=96837.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:37:54,836 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.143e+02 1.675e+02 2.012e+02 2.364e+02 4.591e+02, threshold=4.024e+02, percent-clipped=3.0 +2023-03-21 07:37:57,387 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 07:38:07,908 INFO [train.py:901] (0/2) Epoch 35, batch 850, loss[loss=0.1303, simple_loss=0.2172, pruned_loss=0.02169, over 7257.00 frames. ], tot_loss[loss=0.1325, simple_loss=0.2129, pruned_loss=0.02608, over 1420070.16 frames. ], batch size: 64, lr: 4.73e-03, grad_scale: 8.0 +2023-03-21 07:38:08,941 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=96869.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:38:14,948 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 07:38:15,438 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 07:38:21,509 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.37 vs. limit=5.0 +2023-03-21 07:38:21,681 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 07:38:24,663 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 07:38:34,191 INFO [train.py:901] (0/2) Epoch 35, batch 900, loss[loss=0.1543, simple_loss=0.2285, pruned_loss=0.04002, over 7262.00 frames. ], tot_loss[loss=0.133, simple_loss=0.2133, pruned_loss=0.02633, over 1424809.74 frames. ], batch size: 52, lr: 4.73e-03, grad_scale: 8.0 +2023-03-21 07:38:46,262 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.358e+02 1.734e+02 2.102e+02 2.438e+02 4.011e+02, threshold=4.204e+02, percent-clipped=0.0 +2023-03-21 07:38:58,065 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3998, 1.6090, 1.3961, 1.5176, 1.5475, 1.4381, 1.4772, 1.1899], + device='cuda:0'), covar=tensor([0.0162, 0.0142, 0.0386, 0.0160, 0.0167, 0.0168, 0.0158, 0.0221], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0033, 0.0033, 0.0034, 0.0033, 0.0031, 0.0035, 0.0041], + device='cuda:0'), out_proj_covar=tensor([3.9578e-05, 3.6479e-05, 3.7066e-05, 3.8044e-05, 3.6679e-05, 3.5050e-05, + 3.9658e-05, 4.5231e-05], device='cuda:0') +2023-03-21 07:38:59,396 INFO [train.py:901] (0/2) Epoch 35, batch 950, loss[loss=0.1151, simple_loss=0.1955, pruned_loss=0.01736, over 7128.00 frames. ], tot_loss[loss=0.1329, simple_loss=0.2133, pruned_loss=0.0262, over 1429424.34 frames. ], batch size: 41, lr: 4.73e-03, grad_scale: 8.0 +2023-03-21 07:39:01,525 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 07:39:06,390 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.56 vs. limit=2.0 +2023-03-21 07:39:06,603 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=96980.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:39:17,639 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=97000.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:39:25,838 INFO [train.py:901] (0/2) Epoch 35, batch 1000, loss[loss=0.1457, simple_loss=0.2304, pruned_loss=0.03056, over 7271.00 frames. ], tot_loss[loss=0.133, simple_loss=0.2137, pruned_loss=0.02619, over 1434479.35 frames. ], batch size: 66, lr: 4.73e-03, grad_scale: 16.0 +2023-03-21 07:39:25,866 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 07:39:31,270 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.60 vs. limit=2.0 +2023-03-21 07:39:31,466 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=97028.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:39:38,013 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.105e+02 1.712e+02 1.972e+02 2.279e+02 4.692e+02, threshold=3.943e+02, percent-clipped=1.0 +2023-03-21 07:39:40,693 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=97046.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:39:46,231 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 07:39:48,891 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=97061.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:39:51,429 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9480, 2.0723, 2.3008, 1.9520, 2.2549, 2.2136, 1.8426, 1.7360], + device='cuda:0'), covar=tensor([0.0388, 0.0348, 0.0199, 0.0407, 0.0318, 0.0449, 0.0442, 0.0341], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0036, 0.0037, 0.0036, 0.0034, 0.0034, 0.0039, 0.0039], + device='cuda:0'), out_proj_covar=tensor([9.4679e-05, 9.3341e-05, 9.4045e-05, 9.0773e-05, 9.0451e-05, 8.8753e-05, + 9.7368e-05, 9.8532e-05], device='cuda:0') +2023-03-21 07:39:51,807 INFO [train.py:901] (0/2) Epoch 35, batch 1050, loss[loss=0.1589, simple_loss=0.2474, pruned_loss=0.03525, over 7102.00 frames. ], tot_loss[loss=0.1329, simple_loss=0.2135, pruned_loss=0.02609, over 1436215.66 frames. ], batch size: 98, lr: 4.73e-03, grad_scale: 16.0 +2023-03-21 07:40:05,941 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=97094.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:40:07,892 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 07:40:11,897 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 07:40:17,357 INFO [train.py:901] (0/2) Epoch 35, batch 1100, loss[loss=0.1565, simple_loss=0.2412, pruned_loss=0.03586, over 7288.00 frames. ], tot_loss[loss=0.1331, simple_loss=0.2137, pruned_loss=0.02621, over 1439137.20 frames. ], batch size: 68, lr: 4.73e-03, grad_scale: 16.0 +2023-03-21 07:40:24,938 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=97132.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:40:29,384 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.067e+02 1.690e+02 1.951e+02 2.274e+02 4.378e+02, threshold=3.903e+02, percent-clipped=1.0 +2023-03-21 07:40:40,481 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 07:40:40,955 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 07:40:42,907 INFO [train.py:901] (0/2) Epoch 35, batch 1150, loss[loss=0.1255, simple_loss=0.2108, pruned_loss=0.02014, over 7364.00 frames. ], tot_loss[loss=0.1329, simple_loss=0.2135, pruned_loss=0.02616, over 1439878.90 frames. ], batch size: 63, lr: 4.73e-03, grad_scale: 16.0 +2023-03-21 07:40:45,654 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3585, 1.5922, 1.3734, 1.4166, 1.5177, 1.4488, 1.3498, 1.1181], + device='cuda:0'), covar=tensor([0.0152, 0.0117, 0.0213, 0.0142, 0.0119, 0.0154, 0.0203, 0.0175], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0033, 0.0033, 0.0035, 0.0033, 0.0032, 0.0035, 0.0041], + device='cuda:0'), out_proj_covar=tensor([3.9892e-05, 3.6473e-05, 3.6979e-05, 3.8148e-05, 3.6882e-05, 3.5176e-05, + 3.9685e-05, 4.5494e-05], device='cuda:0') +2023-03-21 07:40:54,044 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 07:40:54,547 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 07:40:56,934 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.84 vs. limit=2.0 +2023-03-21 07:41:08,767 INFO [train.py:901] (0/2) Epoch 35, batch 1200, loss[loss=0.1477, simple_loss=0.229, pruned_loss=0.03318, over 7236.00 frames. ], tot_loss[loss=0.1334, simple_loss=0.214, pruned_loss=0.02635, over 1438366.47 frames. ], batch size: 89, lr: 4.72e-03, grad_scale: 16.0 +2023-03-21 07:41:20,036 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0842, 1.9365, 2.4050, 2.1104, 2.4642, 2.3186, 2.0372, 1.8557], + device='cuda:0'), covar=tensor([0.0476, 0.0612, 0.0252, 0.0344, 0.0451, 0.0470, 0.0333, 0.0343], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0037, 0.0037, 0.0036, 0.0034, 0.0034, 0.0040, 0.0039], + device='cuda:0'), out_proj_covar=tensor([9.6091e-05, 9.4807e-05, 9.4018e-05, 9.1752e-05, 9.0873e-05, 8.9310e-05, + 9.7893e-05, 9.9558e-05], device='cuda:0') +2023-03-21 07:41:21,385 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.240e+02 1.833e+02 2.092e+02 2.573e+02 3.839e+02, threshold=4.184e+02, percent-clipped=0.0 +2023-03-21 07:41:26,325 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 07:41:31,455 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=97260.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:41:34,833 INFO [train.py:901] (0/2) Epoch 35, batch 1250, loss[loss=0.1317, simple_loss=0.2149, pruned_loss=0.02425, over 7338.00 frames. ], tot_loss[loss=0.1336, simple_loss=0.2145, pruned_loss=0.02633, over 1441161.48 frames. ], batch size: 54, lr: 4.72e-03, grad_scale: 16.0 +2023-03-21 07:41:49,375 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 07:41:53,976 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 07:41:55,513 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 07:42:00,730 INFO [train.py:901] (0/2) Epoch 35, batch 1300, loss[loss=0.1487, simple_loss=0.2339, pruned_loss=0.03176, over 7336.00 frames. ], tot_loss[loss=0.1336, simple_loss=0.2148, pruned_loss=0.02626, over 1442861.13 frames. ], batch size: 54, lr: 4.72e-03, grad_scale: 16.0 +2023-03-21 07:42:02,929 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=97321.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:42:13,456 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.213e+02 1.729e+02 1.997e+02 2.434e+02 4.053e+02, threshold=3.995e+02, percent-clipped=0.0 +2023-03-21 07:42:14,081 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=97342.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:42:16,601 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=97347.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:42:17,206 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.08 vs. limit=5.0 +2023-03-21 07:42:18,999 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 07:42:19,667 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6513, 2.4360, 2.3928, 3.6041, 1.8967, 3.5217, 1.4765, 3.2845], + device='cuda:0'), covar=tensor([0.0206, 0.1406, 0.1850, 0.0236, 0.3923, 0.0315, 0.1213, 0.0396], + device='cuda:0'), in_proj_covar=tensor([0.0205, 0.0252, 0.0269, 0.0210, 0.0255, 0.0215, 0.0232, 0.0231], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 07:42:21,020 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 07:42:21,071 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=97356.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:42:25,029 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 07:42:26,537 INFO [train.py:901] (0/2) Epoch 35, batch 1350, loss[loss=0.1231, simple_loss=0.209, pruned_loss=0.01858, over 7318.00 frames. ], tot_loss[loss=0.1332, simple_loss=0.2143, pruned_loss=0.02608, over 1441974.64 frames. ], batch size: 83, lr: 4.72e-03, grad_scale: 16.0 +2023-03-21 07:42:30,210 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=97374.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:42:30,764 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8794, 2.6983, 2.4491, 3.7905, 1.9782, 3.6771, 1.5521, 3.3741], + device='cuda:0'), covar=tensor([0.0169, 0.1234, 0.1779, 0.0224, 0.3710, 0.0235, 0.1225, 0.0378], + device='cuda:0'), in_proj_covar=tensor([0.0205, 0.0253, 0.0269, 0.0211, 0.0255, 0.0216, 0.0233, 0.0231], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 07:42:35,128 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 07:42:41,776 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.70 vs. limit=2.0 +2023-03-21 07:42:45,977 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=97403.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 07:42:47,638 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.72 vs. limit=2.0 +2023-03-21 07:42:48,457 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=97408.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:42:52,843 INFO [train.py:901] (0/2) Epoch 35, batch 1400, loss[loss=0.1341, simple_loss=0.2158, pruned_loss=0.02616, over 7278.00 frames. ], tot_loss[loss=0.1323, simple_loss=0.2132, pruned_loss=0.02572, over 1438591.24 frames. ], batch size: 66, lr: 4.72e-03, grad_scale: 16.0 +2023-03-21 07:42:57,083 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7288, 2.5725, 2.9275, 2.9145, 2.7012, 2.6527, 2.9787, 2.2271], + device='cuda:0'), covar=tensor([0.0474, 0.0503, 0.0665, 0.0595, 0.0573, 0.0782, 0.0523, 0.2008], + device='cuda:0'), in_proj_covar=tensor([0.0337, 0.0338, 0.0275, 0.0356, 0.0293, 0.0291, 0.0345, 0.0257], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:43:01,110 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=97432.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:43:02,668 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=97435.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:43:06,067 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.085e+02 1.642e+02 1.848e+02 2.311e+02 4.158e+02, threshold=3.695e+02, percent-clipped=1.0 +2023-03-21 07:43:10,603 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 07:43:18,615 INFO [train.py:901] (0/2) Epoch 35, batch 1450, loss[loss=0.1342, simple_loss=0.2166, pruned_loss=0.02585, over 7327.00 frames. ], tot_loss[loss=0.132, simple_loss=0.213, pruned_loss=0.02554, over 1439139.08 frames. ], batch size: 54, lr: 4.72e-03, grad_scale: 8.0 +2023-03-21 07:43:25,230 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=97480.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:43:34,278 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 07:43:44,141 INFO [train.py:901] (0/2) Epoch 35, batch 1500, loss[loss=0.1298, simple_loss=0.2127, pruned_loss=0.0235, over 7279.00 frames. ], tot_loss[loss=0.1323, simple_loss=0.2132, pruned_loss=0.02571, over 1437190.01 frames. ], batch size: 77, lr: 4.72e-03, grad_scale: 8.0 +2023-03-21 07:43:50,317 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 07:43:50,855 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.1090, 4.5887, 4.4034, 5.0280, 4.8810, 4.9821, 4.2743, 4.5885], + device='cuda:0'), covar=tensor([0.0796, 0.2420, 0.2165, 0.0950, 0.0853, 0.1092, 0.0755, 0.1129], + device='cuda:0'), in_proj_covar=tensor([0.0247, 0.0377, 0.0290, 0.0296, 0.0221, 0.0353, 0.0217, 0.0260], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:43:57,134 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.381e+02 1.775e+02 2.183e+02 2.591e+02 6.177e+02, threshold=4.365e+02, percent-clipped=2.0 +2023-03-21 07:44:03,851 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=97555.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:44:09,685 INFO [train.py:901] (0/2) Epoch 35, batch 1550, loss[loss=0.1344, simple_loss=0.2166, pruned_loss=0.02607, over 7362.00 frames. ], tot_loss[loss=0.1321, simple_loss=0.2129, pruned_loss=0.02568, over 1438879.74 frames. ], batch size: 54, lr: 4.72e-03, grad_scale: 8.0 +2023-03-21 07:44:13,280 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 07:44:26,489 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.0182, 4.2868, 4.6947, 4.6144, 4.6518, 4.6040, 4.9688, 4.5022], + device='cuda:0'), covar=tensor([0.0137, 0.0162, 0.0088, 0.0163, 0.0325, 0.0081, 0.0098, 0.0124], + device='cuda:0'), in_proj_covar=tensor([0.0092, 0.0100, 0.0099, 0.0089, 0.0173, 0.0106, 0.0103, 0.0110], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:44:35,924 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=97616.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:44:36,001 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=97616.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:44:36,387 INFO [train.py:901] (0/2) Epoch 35, batch 1600, loss[loss=0.1417, simple_loss=0.2276, pruned_loss=0.0279, over 7329.00 frames. ], tot_loss[loss=0.1316, simple_loss=0.2122, pruned_loss=0.0255, over 1435889.10 frames. ], batch size: 61, lr: 4.71e-03, grad_scale: 8.0 +2023-03-21 07:44:38,518 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3866, 3.8979, 4.0406, 4.0398, 4.0387, 3.9908, 4.2915, 3.8038], + device='cuda:0'), covar=tensor([0.0158, 0.0185, 0.0116, 0.0185, 0.0420, 0.0108, 0.0128, 0.0181], + device='cuda:0'), in_proj_covar=tensor([0.0092, 0.0100, 0.0099, 0.0089, 0.0173, 0.0106, 0.0103, 0.0109], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:44:44,833 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 07:44:45,851 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 07:44:48,742 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.345e+02 1.881e+02 2.141e+02 2.690e+02 4.588e+02, threshold=4.282e+02, percent-clipped=0.0 +2023-03-21 07:44:48,794 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 07:44:48,874 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0929, 3.9310, 3.8900, 3.8814, 3.3434, 3.8618, 4.1099, 3.7399], + device='cuda:0'), covar=tensor([0.0303, 0.0231, 0.0184, 0.0261, 0.0899, 0.0191, 0.0252, 0.0262], + device='cuda:0'), in_proj_covar=tensor([0.0092, 0.0100, 0.0099, 0.0089, 0.0173, 0.0106, 0.0103, 0.0110], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:44:55,907 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=97656.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:44:57,219 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.64 vs. limit=2.0 +2023-03-21 07:44:57,831 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 07:45:01,965 INFO [train.py:901] (0/2) Epoch 35, batch 1650, loss[loss=0.134, simple_loss=0.2167, pruned_loss=0.02566, over 7293.00 frames. ], tot_loss[loss=0.1324, simple_loss=0.2132, pruned_loss=0.02584, over 1438763.86 frames. ], batch size: 68, lr: 4.71e-03, grad_scale: 8.0 +2023-03-21 07:45:02,489 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 07:45:02,588 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8157, 3.0241, 3.7096, 3.7483, 3.7899, 3.8760, 3.7786, 3.6401], + device='cuda:0'), covar=tensor([0.0026, 0.0115, 0.0032, 0.0033, 0.0030, 0.0025, 0.0044, 0.0057], + device='cuda:0'), in_proj_covar=tensor([0.0045, 0.0065, 0.0054, 0.0052, 0.0052, 0.0057, 0.0046, 0.0072], + device='cuda:0'), out_proj_covar=tensor([7.8069e-05, 1.3444e-04, 1.0082e-04, 9.2032e-05, 9.0723e-05, 1.0086e-04, + 8.8868e-05, 1.3593e-04], device='cuda:0') +2023-03-21 07:45:04,857 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.71 vs. limit=5.0 +2023-03-21 07:45:05,663 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=97674.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 07:45:10,592 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 07:45:18,330 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=97698.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 07:45:20,772 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=97703.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:45:21,258 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=97704.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:45:25,479 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-21 07:45:26,744 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9071, 2.9705, 3.2167, 3.0356, 3.2034, 3.1589, 2.8696, 3.2589], + device='cuda:0'), covar=tensor([0.2064, 0.0761, 0.1082, 0.1430, 0.0964, 0.1046, 0.1440, 0.1448], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0065, 0.0050, 0.0049, 0.0048, 0.0048, 0.0066, 0.0051], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 07:45:27,139 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 07:45:27,643 INFO [train.py:901] (0/2) Epoch 35, batch 1700, loss[loss=0.1242, simple_loss=0.2062, pruned_loss=0.02105, over 7343.00 frames. ], tot_loss[loss=0.1325, simple_loss=0.2134, pruned_loss=0.02582, over 1439151.66 frames. ], batch size: 54, lr: 4.71e-03, grad_scale: 8.0 +2023-03-21 07:45:31,167 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 07:45:34,252 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=97730.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:45:36,754 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=97735.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 07:45:40,193 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.243e+02 1.701e+02 2.015e+02 2.345e+02 4.849e+02, threshold=4.029e+02, percent-clipped=2.0 +2023-03-21 07:45:41,236 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 07:45:53,427 INFO [train.py:901] (0/2) Epoch 35, batch 1750, loss[loss=0.1353, simple_loss=0.2205, pruned_loss=0.02502, over 7280.00 frames. ], tot_loss[loss=0.1333, simple_loss=0.2142, pruned_loss=0.02621, over 1438462.84 frames. ], batch size: 70, lr: 4.71e-03, grad_scale: 8.0 +2023-03-21 07:46:01,260 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1067, 2.2806, 1.7819, 1.7643, 2.1535, 1.8932, 1.8811, 1.8584], + device='cuda:0'), covar=tensor([0.0121, 0.0123, 0.0381, 0.0242, 0.0100, 0.0140, 0.0185, 0.0166], + device='cuda:0'), in_proj_covar=tensor([0.0035, 0.0033, 0.0032, 0.0034, 0.0033, 0.0032, 0.0035, 0.0041], + device='cuda:0'), out_proj_covar=tensor([3.9603e-05, 3.6548e-05, 3.6800e-05, 3.7771e-05, 3.6690e-05, 3.5226e-05, + 3.9492e-05, 4.5964e-05], device='cuda:0') +2023-03-21 07:46:06,152 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 07:46:07,175 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 07:46:19,584 INFO [train.py:901] (0/2) Epoch 35, batch 1800, loss[loss=0.1481, simple_loss=0.2319, pruned_loss=0.03211, over 7220.00 frames. ], tot_loss[loss=0.1329, simple_loss=0.2139, pruned_loss=0.026, over 1439013.73 frames. ], batch size: 93, lr: 4.71e-03, grad_scale: 8.0 +2023-03-21 07:46:29,775 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 07:46:32,368 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3575, 4.5733, 4.3227, 4.5029, 4.1282, 4.4811, 4.8055, 4.8111], + device='cuda:0'), covar=tensor([0.0213, 0.0119, 0.0161, 0.0131, 0.0293, 0.0215, 0.0207, 0.0183], + device='cuda:0'), in_proj_covar=tensor([0.0120, 0.0125, 0.0117, 0.0121, 0.0111, 0.0102, 0.0098, 0.0098], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:46:32,763 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.185e+02 1.651e+02 2.008e+02 2.380e+02 5.127e+02, threshold=4.015e+02, percent-clipped=1.0 +2023-03-21 07:46:44,440 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 07:46:45,915 INFO [train.py:901] (0/2) Epoch 35, batch 1850, loss[loss=0.1378, simple_loss=0.2179, pruned_loss=0.0289, over 7267.00 frames. ], tot_loss[loss=0.1327, simple_loss=0.2135, pruned_loss=0.02593, over 1437473.82 frames. ], batch size: 52, lr: 4.71e-03, grad_scale: 8.0 +2023-03-21 07:46:48,042 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7023, 3.8650, 3.6719, 3.7775, 3.4841, 3.7590, 4.1030, 4.1158], + device='cuda:0'), covar=tensor([0.0246, 0.0166, 0.0227, 0.0198, 0.0408, 0.0368, 0.0303, 0.0250], + device='cuda:0'), in_proj_covar=tensor([0.0120, 0.0125, 0.0118, 0.0121, 0.0112, 0.0103, 0.0098, 0.0099], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:46:54,437 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 07:46:56,409 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=97888.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:47:07,832 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=97911.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:47:10,385 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=97916.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:47:10,771 INFO [train.py:901] (0/2) Epoch 35, batch 1900, loss[loss=0.1193, simple_loss=0.1936, pruned_loss=0.02249, over 7177.00 frames. ], tot_loss[loss=0.1324, simple_loss=0.2129, pruned_loss=0.02593, over 1437208.17 frames. ], batch size: 39, lr: 4.71e-03, grad_scale: 8.0 +2023-03-21 07:47:11,307 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 07:47:21,193 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0917, 2.0416, 2.3281, 2.1817, 2.2717, 2.2234, 2.1709, 1.7313], + device='cuda:0'), covar=tensor([0.0338, 0.0341, 0.0205, 0.0233, 0.0350, 0.0320, 0.0241, 0.0230], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0036, 0.0037, 0.0035, 0.0034, 0.0034, 0.0039, 0.0039], + device='cuda:0'), out_proj_covar=tensor([9.4811e-05, 9.3766e-05, 9.3009e-05, 8.9942e-05, 8.9785e-05, 8.8973e-05, + 9.6585e-05, 9.7669e-05], device='cuda:0') +2023-03-21 07:47:24,070 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.234e+02 1.816e+02 2.128e+02 2.562e+02 3.920e+02, threshold=4.257e+02, percent-clipped=0.0 +2023-03-21 07:47:28,351 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=97949.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:47:35,754 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=97964.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:47:36,213 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 07:47:37,197 INFO [train.py:901] (0/2) Epoch 35, batch 1950, loss[loss=0.1417, simple_loss=0.2214, pruned_loss=0.031, over 7360.00 frames. ], tot_loss[loss=0.1328, simple_loss=0.2133, pruned_loss=0.02619, over 1436487.88 frames. ], batch size: 51, lr: 4.71e-03, grad_scale: 8.0 +2023-03-21 07:47:46,522 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0282, 3.7101, 3.7277, 3.7192, 3.7333, 3.6142, 3.8909, 3.4694], + device='cuda:0'), covar=tensor([0.0155, 0.0175, 0.0124, 0.0184, 0.0420, 0.0119, 0.0153, 0.0178], + device='cuda:0'), in_proj_covar=tensor([0.0093, 0.0101, 0.0099, 0.0088, 0.0173, 0.0106, 0.0103, 0.0110], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:47:47,459 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 07:47:51,550 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 07:47:52,527 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 07:47:53,150 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=97998.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:47:56,009 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=98003.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:48:03,657 INFO [train.py:901] (0/2) Epoch 35, batch 2000, loss[loss=0.146, simple_loss=0.2238, pruned_loss=0.03412, over 7249.00 frames. ], tot_loss[loss=0.1328, simple_loss=0.2131, pruned_loss=0.02622, over 1437496.78 frames. ], batch size: 55, lr: 4.70e-03, grad_scale: 8.0 +2023-03-21 07:48:05,267 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1051, 4.0091, 3.9123, 3.8855, 3.4051, 3.9084, 4.1320, 3.6390], + device='cuda:0'), covar=tensor([0.0369, 0.0206, 0.0183, 0.0264, 0.0787, 0.0184, 0.0261, 0.0324], + device='cuda:0'), in_proj_covar=tensor([0.0092, 0.0100, 0.0098, 0.0088, 0.0172, 0.0106, 0.0103, 0.0110], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:48:09,701 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 07:48:10,917 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=98030.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 07:48:10,938 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=98030.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:48:13,500 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5353, 1.8357, 1.5061, 1.7876, 1.6372, 1.5429, 1.6244, 1.2274], + device='cuda:0'), covar=tensor([0.0155, 0.0200, 0.0353, 0.0159, 0.0141, 0.0213, 0.0492, 0.0243], + device='cuda:0'), in_proj_covar=tensor([0.0034, 0.0033, 0.0032, 0.0034, 0.0032, 0.0031, 0.0034, 0.0041], + device='cuda:0'), out_proj_covar=tensor([3.9042e-05, 3.6234e-05, 3.6337e-05, 3.7177e-05, 3.6160e-05, 3.4924e-05, + 3.8879e-05, 4.5291e-05], device='cuda:0') +2023-03-21 07:48:16,925 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.119e+02 1.660e+02 1.963e+02 2.436e+02 6.666e+02, threshold=3.925e+02, percent-clipped=3.0 +2023-03-21 07:48:18,130 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1237, 3.1625, 2.2976, 3.7541, 2.9083, 3.2287, 1.6755, 2.4188], + device='cuda:0'), covar=tensor([0.0450, 0.0876, 0.2518, 0.0625, 0.0410, 0.0591, 0.3550, 0.1803], + device='cuda:0'), in_proj_covar=tensor([0.0283, 0.0260, 0.0284, 0.0270, 0.0273, 0.0266, 0.0239, 0.0262], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 07:48:19,030 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=98046.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:48:20,951 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 07:48:21,496 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=98051.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:48:28,933 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 07:48:29,416 INFO [train.py:901] (0/2) Epoch 35, batch 2050, loss[loss=0.1214, simple_loss=0.2068, pruned_loss=0.01795, over 7212.00 frames. ], tot_loss[loss=0.1325, simple_loss=0.2129, pruned_loss=0.02599, over 1439139.98 frames. ], batch size: 93, lr: 4.70e-03, grad_scale: 8.0 +2023-03-21 07:48:35,046 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=98078.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:48:44,917 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.73 vs. limit=2.0 +2023-03-21 07:48:45,648 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2605, 2.4602, 2.5713, 2.4744, 2.3574, 2.3330, 2.3197, 2.1773], + device='cuda:0'), covar=tensor([0.0498, 0.0382, 0.0291, 0.0284, 0.0762, 0.0648, 0.0403, 0.0269], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0036, 0.0037, 0.0036, 0.0034, 0.0034, 0.0039, 0.0039], + device='cuda:0'), out_proj_covar=tensor([9.5726e-05, 9.4055e-05, 9.4234e-05, 9.0507e-05, 9.0815e-05, 8.9740e-05, + 9.7703e-05, 9.8189e-05], device='cuda:0') +2023-03-21 07:48:49,257 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=98106.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:48:52,407 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2870, 2.9991, 3.1942, 3.1041, 2.8448, 2.7098, 3.3012, 2.3463], + device='cuda:0'), covar=tensor([0.0433, 0.0568, 0.0620, 0.0692, 0.0723, 0.1031, 0.0695, 0.2061], + device='cuda:0'), in_proj_covar=tensor([0.0338, 0.0341, 0.0278, 0.0358, 0.0295, 0.0294, 0.0347, 0.0257], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:48:54,977 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0614, 2.4778, 1.8466, 2.7474, 2.6750, 2.9613, 2.1912, 2.4167], + device='cuda:0'), covar=tensor([0.2250, 0.1074, 0.3534, 0.0615, 0.0306, 0.0311, 0.0305, 0.0364], + device='cuda:0'), in_proj_covar=tensor([0.0257, 0.0233, 0.0250, 0.0257, 0.0197, 0.0196, 0.0212, 0.0224], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 07:48:55,922 INFO [train.py:901] (0/2) Epoch 35, batch 2100, loss[loss=0.1477, simple_loss=0.2279, pruned_loss=0.03372, over 7272.00 frames. ], tot_loss[loss=0.1324, simple_loss=0.2129, pruned_loss=0.02594, over 1440835.60 frames. ], batch size: 70, lr: 4.70e-03, grad_scale: 8.0 +2023-03-21 07:48:58,031 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8721, 3.1710, 3.7696, 3.7626, 3.7701, 3.8298, 3.7705, 3.6868], + device='cuda:0'), covar=tensor([0.0026, 0.0103, 0.0027, 0.0030, 0.0030, 0.0027, 0.0036, 0.0055], + device='cuda:0'), in_proj_covar=tensor([0.0045, 0.0065, 0.0054, 0.0052, 0.0052, 0.0057, 0.0045, 0.0072], + device='cuda:0'), out_proj_covar=tensor([7.7482e-05, 1.3432e-04, 1.0060e-04, 9.1751e-05, 9.0667e-05, 1.0134e-04, + 8.8043e-05, 1.3514e-04], device='cuda:0') +2023-03-21 07:49:04,007 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 07:49:06,956 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 07:49:08,438 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.192e+02 1.595e+02 1.829e+02 2.259e+02 6.867e+02, threshold=3.658e+02, percent-clipped=3.0 +2023-03-21 07:49:21,030 INFO [train.py:901] (0/2) Epoch 35, batch 2150, loss[loss=0.104, simple_loss=0.1876, pruned_loss=0.01015, over 7150.00 frames. ], tot_loss[loss=0.1323, simple_loss=0.213, pruned_loss=0.02584, over 1441108.92 frames. ], batch size: 41, lr: 4.70e-03, grad_scale: 8.0 +2023-03-21 07:49:21,186 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=98167.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:49:30,198 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9639, 5.4981, 5.5330, 5.4920, 5.2562, 4.9693, 5.5939, 5.3840], + device='cuda:0'), covar=tensor([0.0437, 0.0359, 0.0399, 0.0405, 0.0344, 0.0395, 0.0309, 0.0423], + device='cuda:0'), in_proj_covar=tensor([0.0215, 0.0254, 0.0198, 0.0197, 0.0155, 0.0230, 0.0204, 0.0148], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:49:43,055 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-03-21 07:49:44,952 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=98211.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:49:47,842 INFO [train.py:901] (0/2) Epoch 35, batch 2200, loss[loss=0.1307, simple_loss=0.2146, pruned_loss=0.0234, over 7303.00 frames. ], tot_loss[loss=0.1327, simple_loss=0.2136, pruned_loss=0.02595, over 1442752.00 frames. ], batch size: 86, lr: 4.70e-03, grad_scale: 8.0 +2023-03-21 07:49:53,873 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 07:50:00,348 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.165e+02 1.754e+02 2.083e+02 2.530e+02 4.661e+02, threshold=4.167e+02, percent-clipped=2.0 +2023-03-21 07:50:01,378 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=98244.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:50:08,885 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=98259.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:50:12,932 INFO [train.py:901] (0/2) Epoch 35, batch 2250, loss[loss=0.144, simple_loss=0.224, pruned_loss=0.03199, over 7266.00 frames. ], tot_loss[loss=0.1325, simple_loss=0.2136, pruned_loss=0.02573, over 1444691.34 frames. ], batch size: 64, lr: 4.70e-03, grad_scale: 8.0 +2023-03-21 07:50:26,767 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 07:50:27,223 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 07:50:39,634 INFO [train.py:901] (0/2) Epoch 35, batch 2300, loss[loss=0.1272, simple_loss=0.2046, pruned_loss=0.02497, over 7243.00 frames. ], tot_loss[loss=0.1327, simple_loss=0.2136, pruned_loss=0.02586, over 1441300.86 frames. ], batch size: 45, lr: 4.70e-03, grad_scale: 8.0 +2023-03-21 07:50:41,170 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 07:50:46,217 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=98330.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 07:50:52,068 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.334e+02 1.745e+02 2.069e+02 2.499e+02 4.009e+02, threshold=4.139e+02, percent-clipped=0.0 +2023-03-21 07:51:04,685 INFO [train.py:901] (0/2) Epoch 35, batch 2350, loss[loss=0.1266, simple_loss=0.2091, pruned_loss=0.02206, over 7212.00 frames. ], tot_loss[loss=0.1329, simple_loss=0.2139, pruned_loss=0.02597, over 1440370.03 frames. ], batch size: 45, lr: 4.70e-03, grad_scale: 4.0 +2023-03-21 07:51:09,498 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8641, 3.2574, 2.8032, 3.2099, 3.2215, 2.9076, 3.2597, 3.1213], + device='cuda:0'), covar=tensor([0.1197, 0.0640, 0.1183, 0.1190, 0.0768, 0.0808, 0.0898, 0.0976], + device='cuda:0'), in_proj_covar=tensor([0.0057, 0.0057, 0.0067, 0.0058, 0.0055, 0.0060, 0.0056, 0.0054], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:51:10,959 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=98378.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 07:51:27,848 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=98410.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:51:28,210 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 07:51:31,191 INFO [train.py:901] (0/2) Epoch 35, batch 2400, loss[loss=0.13, simple_loss=0.2126, pruned_loss=0.02366, over 7263.00 frames. ], tot_loss[loss=0.1333, simple_loss=0.2144, pruned_loss=0.02616, over 1442465.73 frames. ], batch size: 47, lr: 4.70e-03, grad_scale: 8.0 +2023-03-21 07:51:34,763 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 07:51:44,314 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.114e+02 1.798e+02 2.138e+02 2.531e+02 4.727e+02, threshold=4.276e+02, percent-clipped=2.0 +2023-03-21 07:51:44,871 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 07:51:47,327 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 07:51:54,485 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=98462.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:51:57,513 INFO [train.py:901] (0/2) Epoch 35, batch 2450, loss[loss=0.1431, simple_loss=0.2201, pruned_loss=0.03306, over 7314.00 frames. ], tot_loss[loss=0.1333, simple_loss=0.2146, pruned_loss=0.02599, over 1444791.70 frames. ], batch size: 59, lr: 4.69e-03, grad_scale: 8.0 +2023-03-21 07:51:59,639 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=98471.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:52:14,144 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 07:52:22,568 INFO [train.py:901] (0/2) Epoch 35, batch 2500, loss[loss=0.1457, simple_loss=0.2251, pruned_loss=0.03318, over 7323.00 frames. ], tot_loss[loss=0.1329, simple_loss=0.2138, pruned_loss=0.02595, over 1441936.01 frames. ], batch size: 59, lr: 4.69e-03, grad_scale: 8.0 +2023-03-21 07:52:35,898 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.356e+02 1.755e+02 2.080e+02 2.550e+02 4.221e+02, threshold=4.161e+02, percent-clipped=0.0 +2023-03-21 07:52:36,026 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=98543.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:52:36,507 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=98544.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:52:40,600 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 07:52:49,273 INFO [train.py:901] (0/2) Epoch 35, batch 2550, loss[loss=0.1382, simple_loss=0.2202, pruned_loss=0.02809, over 7351.00 frames. ], tot_loss[loss=0.1333, simple_loss=0.2139, pruned_loss=0.02631, over 1444846.01 frames. ], batch size: 63, lr: 4.69e-03, grad_scale: 8.0 +2023-03-21 07:53:01,801 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=98592.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:53:08,361 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=98604.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:53:14,788 INFO [train.py:901] (0/2) Epoch 35, batch 2600, loss[loss=0.1338, simple_loss=0.2135, pruned_loss=0.02703, over 7282.00 frames. ], tot_loss[loss=0.1328, simple_loss=0.2134, pruned_loss=0.02608, over 1443804.31 frames. ], batch size: 57, lr: 4.69e-03, grad_scale: 8.0 +2023-03-21 07:53:27,537 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.227e+02 1.759e+02 1.926e+02 2.182e+02 4.869e+02, threshold=3.852e+02, percent-clipped=2.0 +2023-03-21 07:53:31,200 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2522, 4.4357, 4.2145, 4.4172, 4.1402, 4.3481, 4.6986, 4.7548], + device='cuda:0'), covar=tensor([0.0203, 0.0121, 0.0185, 0.0145, 0.0354, 0.0247, 0.0197, 0.0151], + device='cuda:0'), in_proj_covar=tensor([0.0122, 0.0126, 0.0119, 0.0123, 0.0113, 0.0104, 0.0099, 0.0100], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:53:31,723 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9543, 2.7791, 3.2689, 3.0552, 3.2352, 2.8490, 2.7034, 3.1884], + device='cuda:0'), covar=tensor([0.1527, 0.0853, 0.1113, 0.1293, 0.0632, 0.1337, 0.1805, 0.1357], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0066, 0.0050, 0.0048, 0.0048, 0.0048, 0.0065, 0.0050], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 07:53:39,558 INFO [train.py:901] (0/2) Epoch 35, batch 2650, loss[loss=0.1538, simple_loss=0.2375, pruned_loss=0.03508, over 7341.00 frames. ], tot_loss[loss=0.1327, simple_loss=0.2134, pruned_loss=0.02603, over 1444337.96 frames. ], batch size: 51, lr: 4.69e-03, grad_scale: 8.0 +2023-03-21 07:53:43,637 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2115, 4.7662, 4.7916, 4.7332, 4.5947, 4.2376, 4.8319, 4.5204], + device='cuda:0'), covar=tensor([0.0977, 0.1059, 0.0891, 0.0935, 0.0845, 0.0912, 0.0824, 0.1203], + device='cuda:0'), in_proj_covar=tensor([0.0212, 0.0251, 0.0197, 0.0194, 0.0155, 0.0229, 0.0202, 0.0147], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:54:03,550 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6498, 3.5619, 2.9019, 3.2981, 2.4723, 2.3764, 1.8932, 3.7130], + device='cuda:0'), covar=tensor([0.0050, 0.0052, 0.0187, 0.0084, 0.0225, 0.0539, 0.0637, 0.0048], + device='cuda:0'), in_proj_covar=tensor([0.0110, 0.0090, 0.0111, 0.0095, 0.0125, 0.0133, 0.0129, 0.0102], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 07:54:05,472 INFO [train.py:901] (0/2) Epoch 35, batch 2700, loss[loss=0.1113, simple_loss=0.1721, pruned_loss=0.02526, over 6375.00 frames. ], tot_loss[loss=0.1327, simple_loss=0.2138, pruned_loss=0.02586, over 1443096.12 frames. ], batch size: 27, lr: 4.69e-03, grad_scale: 8.0 +2023-03-21 07:54:18,204 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.314e+02 1.808e+02 2.115e+02 2.423e+02 3.699e+02, threshold=4.229e+02, percent-clipped=0.0 +2023-03-21 07:54:21,240 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6867, 3.6367, 2.7352, 4.2412, 3.4991, 3.7697, 1.9671, 2.6718], + device='cuda:0'), covar=tensor([0.0522, 0.1036, 0.2278, 0.0618, 0.0516, 0.0635, 0.3387, 0.1868], + device='cuda:0'), in_proj_covar=tensor([0.0284, 0.0261, 0.0287, 0.0271, 0.0272, 0.0266, 0.0240, 0.0263], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 07:54:27,535 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=98762.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:54:29,410 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=98766.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:54:29,836 INFO [train.py:901] (0/2) Epoch 35, batch 2750, loss[loss=0.1379, simple_loss=0.2246, pruned_loss=0.02558, over 7306.00 frames. ], tot_loss[loss=0.1332, simple_loss=0.2143, pruned_loss=0.02608, over 1441825.82 frames. ], batch size: 59, lr: 4.69e-03, grad_scale: 8.0 +2023-03-21 07:54:38,645 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.85 vs. limit=2.0 +2023-03-21 07:54:51,380 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=98810.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:54:52,399 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2864, 2.8325, 2.0281, 2.9202, 2.7951, 3.1716, 2.7478, 2.4794], + device='cuda:0'), covar=tensor([0.1966, 0.0879, 0.3602, 0.0534, 0.0238, 0.0194, 0.0253, 0.0284], + device='cuda:0'), in_proj_covar=tensor([0.0254, 0.0231, 0.0250, 0.0254, 0.0194, 0.0193, 0.0213, 0.0223], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 07:54:54,676 INFO [train.py:901] (0/2) Epoch 35, batch 2800, loss[loss=0.1289, simple_loss=0.2159, pruned_loss=0.02096, over 7306.00 frames. ], tot_loss[loss=0.1331, simple_loss=0.2143, pruned_loss=0.02598, over 1442904.40 frames. ], batch size: 49, lr: 4.69e-03, grad_scale: 8.0 +2023-03-21 07:55:05,015 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5517, 1.4927, 1.6207, 1.8558, 1.7217, 1.7879, 1.4730, 1.8988], + device='cuda:0'), covar=tensor([0.1930, 0.3662, 0.1451, 0.1174, 0.2507, 0.1458, 0.2963, 0.1729], + device='cuda:0'), in_proj_covar=tensor([0.0077, 0.0078, 0.0068, 0.0064, 0.0063, 0.0062, 0.0104, 0.0065], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:55:07,377 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-35.pt +2023-03-21 07:55:22,556 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 07:55:22,612 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 07:55:22,639 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 07:55:22,657 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 07:55:22,845 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 07:55:22,862 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 07:55:22,873 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 07:55:23,338 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 07:55:23,423 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 07:55:23,439 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 07:55:23,455 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 07:55:23,628 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 07:55:23,634 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 07:55:23,642 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 07:55:23,789 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 07:55:23,792 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 07:55:23,854 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 07:55:26,025 INFO [train.py:901] (0/2) Epoch 36, batch 0, loss[loss=0.1411, simple_loss=0.2179, pruned_loss=0.03215, over 7269.00 frames. ], tot_loss[loss=0.1411, simple_loss=0.2179, pruned_loss=0.03215, over 7269.00 frames. ], batch size: 57, lr: 4.62e-03, grad_scale: 8.0 +2023-03-21 07:55:26,026 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 07:55:32,403 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.0101, 5.2061, 5.2886, 5.2438, 4.8884, 4.7930, 5.2607, 4.9354], + device='cuda:0'), covar=tensor([0.0257, 0.0316, 0.0309, 0.0389, 0.0346, 0.0285, 0.0267, 0.0488], + device='cuda:0'), in_proj_covar=tensor([0.0212, 0.0251, 0.0196, 0.0195, 0.0154, 0.0229, 0.0203, 0.0148], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:55:39,595 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0259, 3.7713, 3.6657, 3.7509, 3.6944, 3.6227, 3.7406, 3.5172], + device='cuda:0'), covar=tensor([0.0109, 0.0165, 0.0138, 0.0179, 0.0435, 0.0109, 0.0192, 0.0183], + device='cuda:0'), in_proj_covar=tensor([0.0094, 0.0101, 0.0100, 0.0090, 0.0175, 0.0107, 0.0105, 0.0111], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 07:55:51,380 INFO [train.py:935] (0/2) Epoch 36, validation: loss=0.1652, simple_loss=0.2566, pruned_loss=0.03687, over 1622729.00 frames. +2023-03-21 07:55:51,380 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 07:55:52,369 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.162e+02 1.696e+02 1.971e+02 2.312e+02 4.526e+02, threshold=3.942e+02, percent-clipped=1.0 +2023-03-21 07:55:58,409 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 07:56:08,351 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 07:56:09,611 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.50 vs. limit=2.0 +2023-03-21 07:56:16,034 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 07:56:16,653 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=98890.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:56:16,992 INFO [train.py:901] (0/2) Epoch 36, batch 50, loss[loss=0.1501, simple_loss=0.2318, pruned_loss=0.03416, over 6643.00 frames. ], tot_loss[loss=0.134, simple_loss=0.216, pruned_loss=0.02603, over 324234.64 frames. ], batch size: 106, lr: 4.62e-03, grad_scale: 8.0 +2023-03-21 07:56:18,047 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 07:56:19,213 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=98895.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:56:21,203 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 07:56:21,381 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4777, 3.0345, 3.5156, 3.3583, 3.0919, 2.8710, 3.5529, 2.6297], + device='cuda:0'), covar=tensor([0.0429, 0.0455, 0.0606, 0.0600, 0.0701, 0.0996, 0.0695, 0.1944], + device='cuda:0'), in_proj_covar=tensor([0.0335, 0.0339, 0.0275, 0.0356, 0.0294, 0.0293, 0.0347, 0.0256], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:56:21,798 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=98899.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:56:28,588 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.61 vs. limit=2.0 +2023-03-21 07:56:34,061 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-03-21 07:56:39,168 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 07:56:39,181 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 07:56:42,653 INFO [train.py:901] (0/2) Epoch 36, batch 100, loss[loss=0.1357, simple_loss=0.2181, pruned_loss=0.02665, over 7268.00 frames. ], tot_loss[loss=0.134, simple_loss=0.2148, pruned_loss=0.02658, over 572629.43 frames. ], batch size: 64, lr: 4.62e-03, grad_scale: 8.0 +2023-03-21 07:56:43,619 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.155e+02 1.846e+02 2.136e+02 2.564e+02 5.371e+02, threshold=4.271e+02, percent-clipped=2.0 +2023-03-21 07:56:47,827 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=98951.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:56:50,419 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=98956.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:56:57,861 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=98971.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:57:08,777 INFO [train.py:901] (0/2) Epoch 36, batch 150, loss[loss=0.1413, simple_loss=0.2248, pruned_loss=0.02897, over 7318.00 frames. ], tot_loss[loss=0.1327, simple_loss=0.2137, pruned_loss=0.02586, over 767480.29 frames. ], batch size: 44, lr: 4.62e-03, grad_scale: 8.0 +2023-03-21 07:57:10,612 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.22 vs. limit=2.0 +2023-03-21 07:57:29,851 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=99032.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:57:34,197 INFO [train.py:901] (0/2) Epoch 36, batch 200, loss[loss=0.1335, simple_loss=0.2146, pruned_loss=0.02619, over 7266.00 frames. ], tot_loss[loss=0.1322, simple_loss=0.2131, pruned_loss=0.02564, over 916265.34 frames. ], batch size: 70, lr: 4.61e-03, grad_scale: 8.0 +2023-03-21 07:57:35,163 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.253e+02 1.668e+02 1.975e+02 2.351e+02 4.108e+02, threshold=3.950e+02, percent-clipped=0.0 +2023-03-21 07:57:38,702 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 07:57:39,263 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=99051.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:57:42,638 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 07:57:47,332 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=99066.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:57:50,363 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 07:58:00,458 INFO [train.py:901] (0/2) Epoch 36, batch 250, loss[loss=0.137, simple_loss=0.2167, pruned_loss=0.0286, over 7304.00 frames. ], tot_loss[loss=0.1322, simple_loss=0.2134, pruned_loss=0.02544, over 1033272.64 frames. ], batch size: 86, lr: 4.61e-03, grad_scale: 8.0 +2023-03-21 07:58:02,961 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 07:58:10,944 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=99112.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:58:11,839 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=99114.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:58:18,959 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6800, 5.2120, 5.3101, 5.2772, 5.0513, 4.7145, 5.2997, 5.1596], + device='cuda:0'), covar=tensor([0.0345, 0.0355, 0.0334, 0.0379, 0.0346, 0.0372, 0.0333, 0.0407], + device='cuda:0'), in_proj_covar=tensor([0.0214, 0.0254, 0.0200, 0.0197, 0.0157, 0.0230, 0.0207, 0.0150], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 07:58:23,895 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 07:58:25,398 INFO [train.py:901] (0/2) Epoch 36, batch 300, loss[loss=0.1313, simple_loss=0.2127, pruned_loss=0.02499, over 7267.00 frames. ], tot_loss[loss=0.132, simple_loss=0.2131, pruned_loss=0.02543, over 1126083.83 frames. ], batch size: 89, lr: 4.61e-03, grad_scale: 4.0 +2023-03-21 07:58:26,853 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.227e+02 1.659e+02 1.981e+02 2.270e+02 3.464e+02, threshold=3.962e+02, percent-clipped=0.0 +2023-03-21 07:58:33,889 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 07:58:51,381 INFO [train.py:901] (0/2) Epoch 36, batch 350, loss[loss=0.1109, simple_loss=0.1964, pruned_loss=0.01269, over 7339.00 frames. ], tot_loss[loss=0.1324, simple_loss=0.2133, pruned_loss=0.02578, over 1193129.79 frames. ], batch size: 44, lr: 4.61e-03, grad_scale: 4.0 +2023-03-21 07:58:55,460 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=99199.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:59:06,546 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 07:59:10,629 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=99229.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:59:17,815 INFO [train.py:901] (0/2) Epoch 36, batch 400, loss[loss=0.1382, simple_loss=0.2177, pruned_loss=0.02933, over 7348.00 frames. ], tot_loss[loss=0.1331, simple_loss=0.2143, pruned_loss=0.02597, over 1248945.09 frames. ], batch size: 54, lr: 4.61e-03, grad_scale: 8.0 +2023-03-21 07:59:19,284 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.276e+02 1.771e+02 2.180e+02 2.603e+02 7.277e+02, threshold=4.360e+02, percent-clipped=4.0 +2023-03-21 07:59:20,357 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99246.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:59:20,835 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=99247.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:59:22,467 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1540, 3.1700, 2.1684, 3.4467, 2.6095, 3.0373, 1.5690, 2.0727], + device='cuda:0'), covar=tensor([0.0550, 0.0708, 0.3161, 0.0655, 0.0427, 0.0438, 0.3876, 0.2186], + device='cuda:0'), in_proj_covar=tensor([0.0284, 0.0260, 0.0286, 0.0271, 0.0271, 0.0266, 0.0239, 0.0263], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 07:59:22,860 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99251.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:59:31,439 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=99268.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:59:31,693 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.57 vs. limit=2.0 +2023-03-21 07:59:42,549 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=99290.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 07:59:42,899 INFO [train.py:901] (0/2) Epoch 36, batch 450, loss[loss=0.1411, simple_loss=0.2261, pruned_loss=0.02803, over 7309.00 frames. ], tot_loss[loss=0.1329, simple_loss=0.2142, pruned_loss=0.02579, over 1293048.38 frames. ], batch size: 80, lr: 4.61e-03, grad_scale: 4.0 +2023-03-21 07:59:46,974 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 07:59:47,477 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 08:00:01,065 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=99324.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:00:02,482 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99327.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:00:03,568 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=99329.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:00:09,422 INFO [train.py:901] (0/2) Epoch 36, batch 500, loss[loss=0.1358, simple_loss=0.2175, pruned_loss=0.02709, over 7274.00 frames. ], tot_loss[loss=0.1333, simple_loss=0.2145, pruned_loss=0.026, over 1324916.09 frames. ], batch size: 77, lr: 4.61e-03, grad_scale: 4.0 +2023-03-21 08:00:11,399 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.249e+02 1.683e+02 2.019e+02 2.214e+02 6.432e+02, threshold=4.039e+02, percent-clipped=1.0 +2023-03-21 08:00:21,429 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 08:00:22,440 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 08:00:23,453 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 08:00:25,001 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 08:00:30,045 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 08:00:31,641 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=99385.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:00:34,393 INFO [train.py:901] (0/2) Epoch 36, batch 550, loss[loss=0.1151, simple_loss=0.1869, pruned_loss=0.0216, over 6931.00 frames. ], tot_loss[loss=0.1332, simple_loss=0.2143, pruned_loss=0.02602, over 1351092.51 frames. ], batch size: 35, lr: 4.61e-03, grad_scale: 4.0 +2023-03-21 08:00:41,356 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 08:00:43,454 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99407.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:00:49,543 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 08:00:50,642 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=99420.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:00:53,110 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 08:00:59,615 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 08:00:59,963 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.69 vs. limit=2.0 +2023-03-21 08:01:01,120 INFO [train.py:901] (0/2) Epoch 36, batch 600, loss[loss=0.1457, simple_loss=0.2308, pruned_loss=0.03036, over 7218.00 frames. ], tot_loss[loss=0.1334, simple_loss=0.2147, pruned_loss=0.02607, over 1371950.95 frames. ], batch size: 93, lr: 4.61e-03, grad_scale: 4.0 +2023-03-21 08:01:03,076 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.108e+02 1.742e+02 2.120e+02 2.350e+02 3.702e+02, threshold=4.240e+02, percent-clipped=0.0 +2023-03-21 08:01:12,949 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.61 vs. limit=5.0 +2023-03-21 08:01:14,766 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=99468.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 08:01:15,644 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 08:01:21,376 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=99481.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:01:23,790 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 08:01:27,480 INFO [train.py:901] (0/2) Epoch 36, batch 650, loss[loss=0.1007, simple_loss=0.169, pruned_loss=0.0162, over 5786.00 frames. ], tot_loss[loss=0.1322, simple_loss=0.2132, pruned_loss=0.02558, over 1385865.05 frames. ], batch size: 25, lr: 4.60e-03, grad_scale: 4.0 +2023-03-21 08:01:33,549 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0779, 3.7735, 3.7706, 3.9062, 3.7909, 3.7935, 3.9688, 3.5654], + device='cuda:0'), covar=tensor([0.0159, 0.0199, 0.0143, 0.0164, 0.0444, 0.0121, 0.0176, 0.0206], + device='cuda:0'), in_proj_covar=tensor([0.0094, 0.0101, 0.0100, 0.0089, 0.0175, 0.0107, 0.0105, 0.0111], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:01:37,048 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0211, 2.4212, 2.5320, 2.0760, 2.2880, 2.2180, 1.9001, 1.7908], + device='cuda:0'), covar=tensor([0.0537, 0.0393, 0.0188, 0.0330, 0.0410, 0.0472, 0.0513, 0.0282], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0037, 0.0038, 0.0037, 0.0035, 0.0035, 0.0041, 0.0040], + device='cuda:0'), out_proj_covar=tensor([9.8013e-05, 9.6663e-05, 9.6229e-05, 9.3585e-05, 9.2598e-05, 9.1952e-05, + 1.0114e-04, 1.0080e-04], device='cuda:0') +2023-03-21 08:01:41,993 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 08:01:42,153 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0108, 2.8483, 2.0818, 3.2821, 2.6378, 2.8622, 1.3971, 2.0533], + device='cuda:0'), covar=tensor([0.0513, 0.0953, 0.2828, 0.0690, 0.0463, 0.0708, 0.3922, 0.1707], + device='cuda:0'), in_proj_covar=tensor([0.0284, 0.0260, 0.0288, 0.0272, 0.0273, 0.0268, 0.0240, 0.0264], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:01:46,668 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=99529.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 08:01:51,007 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 08:01:52,520 INFO [train.py:901] (0/2) Epoch 36, batch 700, loss[loss=0.1343, simple_loss=0.2142, pruned_loss=0.02717, over 7252.00 frames. ], tot_loss[loss=0.1317, simple_loss=0.2127, pruned_loss=0.02539, over 1398306.62 frames. ], batch size: 47, lr: 4.60e-03, grad_scale: 4.0 +2023-03-21 08:01:54,519 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.217e+02 1.727e+02 1.965e+02 2.188e+02 5.872e+02, threshold=3.931e+02, percent-clipped=1.0 +2023-03-21 08:01:55,065 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=99546.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:01:57,599 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=99551.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:02:06,065 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=99568.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:02:15,466 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5830, 1.7359, 1.5006, 1.6697, 1.7038, 1.6368, 1.6849, 1.3008], + device='cuda:0'), covar=tensor([0.0130, 0.0127, 0.0229, 0.0148, 0.0098, 0.0156, 0.0121, 0.0175], + device='cuda:0'), in_proj_covar=tensor([0.0036, 0.0034, 0.0033, 0.0035, 0.0033, 0.0033, 0.0035, 0.0042], + device='cuda:0'), out_proj_covar=tensor([4.1295e-05, 3.7689e-05, 3.7337e-05, 3.8620e-05, 3.7124e-05, 3.6398e-05, + 3.9895e-05, 4.6812e-05], device='cuda:0') +2023-03-21 08:02:15,873 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99585.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:02:16,360 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 08:02:17,359 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 08:02:17,967 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=99589.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:02:18,839 INFO [train.py:901] (0/2) Epoch 36, batch 750, loss[loss=0.1248, simple_loss=0.2168, pruned_loss=0.01641, over 7128.00 frames. ], tot_loss[loss=0.1313, simple_loss=0.2122, pruned_loss=0.02521, over 1407296.90 frames. ], batch size: 98, lr: 4.60e-03, grad_scale: 4.0 +2023-03-21 08:02:20,361 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=99594.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:02:22,873 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=99599.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:02:31,173 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 08:02:35,732 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 08:02:35,787 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99624.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:02:37,355 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=99627.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:02:38,463 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=99629.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:02:41,822 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 08:02:42,810 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 08:02:44,294 INFO [train.py:901] (0/2) Epoch 36, batch 800, loss[loss=0.1009, simple_loss=0.1749, pruned_loss=0.0134, over 7012.00 frames. ], tot_loss[loss=0.1319, simple_loss=0.2129, pruned_loss=0.02542, over 1414987.53 frames. ], batch size: 35, lr: 4.60e-03, grad_scale: 8.0 +2023-03-21 08:02:46,284 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.255e+02 1.802e+02 2.004e+02 2.443e+02 3.511e+02, threshold=4.007e+02, percent-clipped=0.0 +2023-03-21 08:02:48,983 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=99650.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:02:53,835 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 08:03:02,636 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=99675.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:03:05,146 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99680.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:03:05,186 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=99680.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:03:10,531 INFO [train.py:901] (0/2) Epoch 36, batch 850, loss[loss=0.1309, simple_loss=0.2204, pruned_loss=0.02068, over 7283.00 frames. ], tot_loss[loss=0.1316, simple_loss=0.2127, pruned_loss=0.02529, over 1421323.49 frames. ], batch size: 66, lr: 4.60e-03, grad_scale: 8.0 +2023-03-21 08:03:12,976 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 08:03:13,420 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 08:03:18,365 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 08:03:18,483 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=99707.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:03:18,543 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0035, 2.9483, 2.1802, 3.4940, 2.5590, 2.9907, 1.5345, 2.3059], + device='cuda:0'), covar=tensor([0.0540, 0.0894, 0.2794, 0.0724, 0.0493, 0.0542, 0.4139, 0.1753], + device='cuda:0'), in_proj_covar=tensor([0.0282, 0.0260, 0.0286, 0.0271, 0.0271, 0.0266, 0.0238, 0.0262], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:03:21,892 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 08:03:35,385 INFO [train.py:901] (0/2) Epoch 36, batch 900, loss[loss=0.1384, simple_loss=0.2212, pruned_loss=0.02778, over 7250.00 frames. ], tot_loss[loss=0.1319, simple_loss=0.2128, pruned_loss=0.02546, over 1425251.31 frames. ], batch size: 64, lr: 4.60e-03, grad_scale: 8.0 +2023-03-21 08:03:35,517 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=99741.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:03:37,374 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.129e+02 1.709e+02 1.934e+02 2.363e+02 3.874e+02, threshold=3.867e+02, percent-clipped=0.0 +2023-03-21 08:03:43,192 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=99755.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:03:54,340 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99776.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:03:58,953 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2246, 3.4796, 2.8521, 4.0604, 2.1070, 3.9737, 1.9344, 3.6302], + device='cuda:0'), covar=tensor([0.0193, 0.0820, 0.1420, 0.0181, 0.3885, 0.0237, 0.1197, 0.0410], + device='cuda:0'), in_proj_covar=tensor([0.0203, 0.0252, 0.0265, 0.0208, 0.0251, 0.0213, 0.0231, 0.0230], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:03:59,834 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 08:04:01,798 INFO [train.py:901] (0/2) Epoch 36, batch 950, loss[loss=0.1316, simple_loss=0.2153, pruned_loss=0.02394, over 7280.00 frames. ], tot_loss[loss=0.1315, simple_loss=0.2126, pruned_loss=0.02524, over 1427661.14 frames. ], batch size: 70, lr: 4.60e-03, grad_scale: 8.0 +2023-03-21 08:04:18,755 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99824.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 08:04:24,204 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 08:04:27,787 INFO [train.py:901] (0/2) Epoch 36, batch 1000, loss[loss=0.1101, simple_loss=0.1853, pruned_loss=0.01746, over 6966.00 frames. ], tot_loss[loss=0.1312, simple_loss=0.2123, pruned_loss=0.02507, over 1428801.18 frames. ], batch size: 35, lr: 4.60e-03, grad_scale: 8.0 +2023-03-21 08:04:29,777 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.174e+02 1.667e+02 1.934e+02 2.174e+02 4.994e+02, threshold=3.868e+02, percent-clipped=1.0 +2023-03-21 08:04:45,765 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 08:04:50,295 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=99885.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:04:53,211 INFO [train.py:901] (0/2) Epoch 36, batch 1050, loss[loss=0.1514, simple_loss=0.2362, pruned_loss=0.03333, over 7104.00 frames. ], tot_loss[loss=0.1315, simple_loss=0.2124, pruned_loss=0.02528, over 1429991.00 frames. ], batch size: 98, lr: 4.60e-03, grad_scale: 8.0 +2023-03-21 08:04:59,886 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.27 vs. limit=2.0 +2023-03-21 08:05:06,174 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 08:05:08,464 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.36 vs. limit=5.0 +2023-03-21 08:05:09,838 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99924.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:05:09,894 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=99924.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:05:10,306 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 08:05:11,884 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6799, 3.9507, 3.6966, 4.0080, 3.5751, 3.9258, 4.1322, 4.2089], + device='cuda:0'), covar=tensor([0.0272, 0.0169, 0.0221, 0.0162, 0.0363, 0.0333, 0.0276, 0.0211], + device='cuda:0'), in_proj_covar=tensor([0.0121, 0.0126, 0.0117, 0.0122, 0.0113, 0.0103, 0.0097, 0.0100], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:05:14,981 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=99933.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:05:19,605 INFO [train.py:901] (0/2) Epoch 36, batch 1100, loss[loss=0.1396, simple_loss=0.2259, pruned_loss=0.02663, over 6695.00 frames. ], tot_loss[loss=0.1313, simple_loss=0.2123, pruned_loss=0.02518, over 1431730.88 frames. ], batch size: 107, lr: 4.59e-03, grad_scale: 8.0 +2023-03-21 08:05:21,596 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.278e+02 1.809e+02 2.088e+02 2.567e+02 3.925e+02, threshold=4.176e+02, percent-clipped=1.0 +2023-03-21 08:05:21,672 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=99945.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:05:29,511 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-03-21 08:05:35,333 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=99972.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:05:39,402 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=99980.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:05:39,824 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 08:05:40,327 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 08:05:44,763 INFO [train.py:901] (0/2) Epoch 36, batch 1150, loss[loss=0.1276, simple_loss=0.2085, pruned_loss=0.02334, over 7256.00 frames. ], tot_loss[loss=0.1312, simple_loss=0.2121, pruned_loss=0.0252, over 1433754.98 frames. ], batch size: 89, lr: 4.59e-03, grad_scale: 8.0 +2023-03-21 08:05:49,580 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-100000.pt +2023-03-21 08:05:56,608 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 08:05:56,733 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=100007.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 08:05:57,127 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 08:05:59,001 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.81 vs. limit=5.0 +2023-03-21 08:06:08,472 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=100028.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:06:12,465 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=100036.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:06:13,670 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.21 vs. limit=2.0 +2023-03-21 08:06:14,924 INFO [train.py:901] (0/2) Epoch 36, batch 1200, loss[loss=0.1526, simple_loss=0.2342, pruned_loss=0.03552, over 7130.00 frames. ], tot_loss[loss=0.1316, simple_loss=0.2123, pruned_loss=0.02542, over 1435619.80 frames. ], batch size: 98, lr: 4.59e-03, grad_scale: 8.0 +2023-03-21 08:06:16,953 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.316e+02 1.656e+02 1.985e+02 2.321e+02 4.535e+02, threshold=3.970e+02, percent-clipped=2.0 +2023-03-21 08:06:18,540 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6923, 4.2843, 4.0647, 4.7339, 4.5364, 4.6293, 4.0231, 4.1988], + device='cuda:0'), covar=tensor([0.0957, 0.2636, 0.2367, 0.0982, 0.0925, 0.1375, 0.0904, 0.1132], + device='cuda:0'), in_proj_covar=tensor([0.0256, 0.0383, 0.0293, 0.0301, 0.0226, 0.0360, 0.0222, 0.0265], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:06:19,380 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.65 vs. limit=2.0 +2023-03-21 08:06:28,674 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=100068.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 08:06:29,512 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 08:06:32,620 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=100076.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:06:40,129 INFO [train.py:901] (0/2) Epoch 36, batch 1250, loss[loss=0.1134, simple_loss=0.1918, pruned_loss=0.01752, over 7168.00 frames. ], tot_loss[loss=0.1312, simple_loss=0.2122, pruned_loss=0.02516, over 1436776.92 frames. ], batch size: 39, lr: 4.59e-03, grad_scale: 8.0 +2023-03-21 08:06:52,817 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 08:06:57,357 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 08:06:57,402 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=100124.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:06:57,467 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=100124.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 08:06:58,365 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 08:07:06,032 INFO [train.py:901] (0/2) Epoch 36, batch 1300, loss[loss=0.1444, simple_loss=0.2303, pruned_loss=0.02924, over 7275.00 frames. ], tot_loss[loss=0.1319, simple_loss=0.2131, pruned_loss=0.0254, over 1438358.76 frames. ], batch size: 47, lr: 4.59e-03, grad_scale: 8.0 +2023-03-21 08:07:08,002 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.272e+02 1.807e+02 2.023e+02 2.519e+02 3.983e+02, threshold=4.046e+02, percent-clipped=1.0 +2023-03-21 08:07:21,862 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=100172.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 08:07:22,803 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 08:07:25,260 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 08:07:28,703 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 08:07:31,166 INFO [train.py:901] (0/2) Epoch 36, batch 1350, loss[loss=0.1364, simple_loss=0.221, pruned_loss=0.02587, over 7217.00 frames. ], tot_loss[loss=0.1317, simple_loss=0.2126, pruned_loss=0.02541, over 1437592.69 frames. ], batch size: 93, lr: 4.59e-03, grad_scale: 8.0 +2023-03-21 08:07:39,704 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 08:07:43,344 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4347, 4.2266, 3.7385, 3.9461, 3.4633, 2.7942, 2.0630, 4.4169], + device='cuda:0'), covar=tensor([0.0052, 0.0069, 0.0108, 0.0064, 0.0137, 0.0403, 0.0609, 0.0043], + device='cuda:0'), in_proj_covar=tensor([0.0112, 0.0090, 0.0111, 0.0094, 0.0126, 0.0133, 0.0130, 0.0102], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 08:07:49,296 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=100224.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:07:49,341 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8218, 3.0737, 2.6558, 2.9967, 2.9348, 2.7410, 3.0627, 2.5819], + device='cuda:0'), covar=tensor([0.0639, 0.0463, 0.0902, 0.1027, 0.0922, 0.0605, 0.0787, 0.1536], + device='cuda:0'), in_proj_covar=tensor([0.0056, 0.0058, 0.0066, 0.0058, 0.0055, 0.0060, 0.0056, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:07:56,321 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=100238.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:07:57,706 INFO [train.py:901] (0/2) Epoch 36, batch 1400, loss[loss=0.1282, simple_loss=0.2116, pruned_loss=0.02235, over 7344.00 frames. ], tot_loss[loss=0.1324, simple_loss=0.2132, pruned_loss=0.02578, over 1439679.69 frames. ], batch size: 54, lr: 4.59e-03, grad_scale: 8.0 +2023-03-21 08:07:59,672 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.298e+02 1.761e+02 2.095e+02 2.393e+02 3.680e+02, threshold=4.190e+02, percent-clipped=0.0 +2023-03-21 08:07:59,830 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=100245.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:08:02,638 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.90 vs. limit=2.0 +2023-03-21 08:08:10,726 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 08:08:13,282 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=100272.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:08:23,916 INFO [train.py:901] (0/2) Epoch 36, batch 1450, loss[loss=0.1517, simple_loss=0.2242, pruned_loss=0.03963, over 7236.00 frames. ], tot_loss[loss=0.1319, simple_loss=0.2127, pruned_loss=0.02551, over 1441369.81 frames. ], batch size: 47, lr: 4.59e-03, grad_scale: 8.0 +2023-03-21 08:08:25,022 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=100293.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:08:28,133 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=100299.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:08:34,107 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 08:08:45,877 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2650, 3.4072, 2.4366, 3.7901, 2.8396, 3.3850, 1.6763, 2.3420], + device='cuda:0'), covar=tensor([0.0400, 0.0738, 0.2327, 0.0497, 0.0424, 0.0780, 0.3397, 0.1810], + device='cuda:0'), in_proj_covar=tensor([0.0279, 0.0257, 0.0282, 0.0267, 0.0272, 0.0266, 0.0236, 0.0262], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:08:46,809 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=100336.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:08:49,198 INFO [train.py:901] (0/2) Epoch 36, batch 1500, loss[loss=0.1216, simple_loss=0.2041, pruned_loss=0.01952, over 7314.00 frames. ], tot_loss[loss=0.1319, simple_loss=0.2129, pruned_loss=0.02542, over 1440928.50 frames. ], batch size: 68, lr: 4.59e-03, grad_scale: 8.0 +2023-03-21 08:08:50,183 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 08:08:51,156 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.275e+02 1.764e+02 1.940e+02 2.308e+02 3.586e+02, threshold=3.880e+02, percent-clipped=0.0 +2023-03-21 08:09:00,337 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=100363.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 08:09:12,001 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=100384.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:09:15,489 INFO [train.py:901] (0/2) Epoch 36, batch 1550, loss[loss=0.1282, simple_loss=0.2039, pruned_loss=0.02623, over 7217.00 frames. ], tot_loss[loss=0.1311, simple_loss=0.2121, pruned_loss=0.02508, over 1440831.38 frames. ], batch size: 45, lr: 4.58e-03, grad_scale: 8.0 +2023-03-21 08:09:15,510 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 08:09:21,324 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.69 vs. limit=2.0 +2023-03-21 08:09:21,631 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=100402.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:09:41,389 INFO [train.py:901] (0/2) Epoch 36, batch 1600, loss[loss=0.1416, simple_loss=0.2251, pruned_loss=0.02912, over 7125.00 frames. ], tot_loss[loss=0.1312, simple_loss=0.2123, pruned_loss=0.02503, over 1441385.50 frames. ], batch size: 98, lr: 4.58e-03, grad_scale: 8.0 +2023-03-21 08:09:43,320 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.302e+02 1.792e+02 2.169e+02 2.475e+02 4.833e+02, threshold=4.339e+02, percent-clipped=2.0 +2023-03-21 08:09:46,593 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 08:09:47,627 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 08:09:51,256 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 08:09:53,914 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=100463.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 08:10:00,421 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 08:10:04,413 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 08:10:07,880 INFO [train.py:901] (0/2) Epoch 36, batch 1650, loss[loss=0.1454, simple_loss=0.229, pruned_loss=0.03085, over 7325.00 frames. ], tot_loss[loss=0.132, simple_loss=0.2133, pruned_loss=0.02538, over 1441748.58 frames. ], batch size: 75, lr: 4.58e-03, grad_scale: 8.0 +2023-03-21 08:10:12,403 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 08:10:30,737 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 08:10:33,846 INFO [train.py:901] (0/2) Epoch 36, batch 1700, loss[loss=0.1577, simple_loss=0.2374, pruned_loss=0.03902, over 6654.00 frames. ], tot_loss[loss=0.1318, simple_loss=0.2131, pruned_loss=0.02525, over 1442942.55 frames. ], batch size: 106, lr: 4.58e-03, grad_scale: 8.0 +2023-03-21 08:10:34,854 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 08:10:36,424 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.329e+02 1.808e+02 2.052e+02 2.359e+02 3.497e+02, threshold=4.104e+02, percent-clipped=0.0 +2023-03-21 08:10:45,304 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 08:10:59,257 INFO [train.py:901] (0/2) Epoch 36, batch 1750, loss[loss=0.1539, simple_loss=0.2292, pruned_loss=0.03924, over 7317.00 frames. ], tot_loss[loss=0.1325, simple_loss=0.2136, pruned_loss=0.02565, over 1445375.03 frames. ], batch size: 49, lr: 4.58e-03, grad_scale: 8.0 +2023-03-21 08:11:00,827 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=100594.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:11:09,039 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 08:11:10,128 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 08:11:20,770 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.78 vs. limit=2.0 +2023-03-21 08:11:26,004 INFO [train.py:901] (0/2) Epoch 36, batch 1800, loss[loss=0.1353, simple_loss=0.2142, pruned_loss=0.02822, over 7259.00 frames. ], tot_loss[loss=0.1323, simple_loss=0.2133, pruned_loss=0.0257, over 1443801.24 frames. ], batch size: 64, lr: 4.58e-03, grad_scale: 8.0 +2023-03-21 08:11:26,156 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8598, 3.2350, 2.7196, 3.2338, 3.0776, 2.6566, 3.2060, 2.8183], + device='cuda:0'), covar=tensor([0.1046, 0.0562, 0.1061, 0.0578, 0.0880, 0.0708, 0.0625, 0.1191], + device='cuda:0'), in_proj_covar=tensor([0.0056, 0.0057, 0.0066, 0.0058, 0.0055, 0.0060, 0.0055, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:11:28,012 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.362e+02 1.832e+02 2.095e+02 2.409e+02 4.709e+02, threshold=4.189e+02, percent-clipped=2.0 +2023-03-21 08:11:32,104 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 08:11:35,887 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7960, 2.9720, 2.6398, 3.7753, 1.9366, 3.5719, 1.5859, 3.2656], + device='cuda:0'), covar=tensor([0.0158, 0.1083, 0.1621, 0.0189, 0.3785, 0.0215, 0.1227, 0.0455], + device='cuda:0'), in_proj_covar=tensor([0.0203, 0.0251, 0.0263, 0.0207, 0.0251, 0.0214, 0.0229, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:11:37,361 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=100663.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 08:11:38,375 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1417, 2.4017, 2.5586, 2.0580, 2.3351, 2.2943, 1.9925, 1.9006], + device='cuda:0'), covar=tensor([0.0516, 0.0393, 0.0309, 0.0324, 0.0583, 0.0729, 0.0444, 0.0386], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0038, 0.0039, 0.0037, 0.0035, 0.0036, 0.0041, 0.0040], + device='cuda:0'), out_proj_covar=tensor([9.8030e-05, 9.7789e-05, 9.7388e-05, 9.4600e-05, 9.2820e-05, 9.3702e-05, + 1.0144e-04, 1.0160e-04], device='cuda:0') +2023-03-21 08:11:45,231 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 08:11:51,209 INFO [train.py:901] (0/2) Epoch 36, batch 1850, loss[loss=0.1409, simple_loss=0.2083, pruned_loss=0.03671, over 7233.00 frames. ], tot_loss[loss=0.1324, simple_loss=0.2132, pruned_loss=0.02583, over 1443102.13 frames. ], batch size: 45, lr: 4.58e-03, grad_scale: 8.0 +2023-03-21 08:11:55,146 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.51 vs. limit=2.0 +2023-03-21 08:11:56,368 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 08:12:01,949 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=100711.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 08:12:12,256 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=100730.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 08:12:13,146 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 08:12:17,575 INFO [train.py:901] (0/2) Epoch 36, batch 1900, loss[loss=0.1268, simple_loss=0.2138, pruned_loss=0.0199, over 7317.00 frames. ], tot_loss[loss=0.1325, simple_loss=0.2133, pruned_loss=0.0258, over 1443739.28 frames. ], batch size: 80, lr: 4.58e-03, grad_scale: 8.0 +2023-03-21 08:12:18,651 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1533, 2.2581, 2.4386, 1.9614, 2.2821, 2.2727, 1.8841, 1.8278], + device='cuda:0'), covar=tensor([0.0421, 0.0345, 0.0227, 0.0294, 0.0424, 0.0377, 0.0433, 0.0301], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0038, 0.0039, 0.0038, 0.0036, 0.0037, 0.0041, 0.0041], + device='cuda:0'), out_proj_covar=tensor([9.9027e-05, 9.8436e-05, 9.8324e-05, 9.5786e-05, 9.3927e-05, 9.4491e-05, + 1.0239e-04, 1.0265e-04], device='cuda:0') +2023-03-21 08:12:19,532 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.072e+02 1.719e+02 1.937e+02 2.500e+02 4.543e+02, threshold=3.873e+02, percent-clipped=1.0 +2023-03-21 08:12:26,091 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=100758.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 08:12:36,480 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 08:12:43,114 INFO [train.py:901] (0/2) Epoch 36, batch 1950, loss[loss=0.1576, simple_loss=0.2421, pruned_loss=0.03649, over 6684.00 frames. ], tot_loss[loss=0.1326, simple_loss=0.2137, pruned_loss=0.02576, over 1444492.33 frames. ], batch size: 106, lr: 4.57e-03, grad_scale: 8.0 +2023-03-21 08:12:43,271 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=100791.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 08:12:47,958 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 08:12:53,107 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 08:12:54,068 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 08:13:09,062 INFO [train.py:901] (0/2) Epoch 36, batch 2000, loss[loss=0.1302, simple_loss=0.215, pruned_loss=0.02271, over 7237.00 frames. ], tot_loss[loss=0.1327, simple_loss=0.2139, pruned_loss=0.02574, over 1444151.63 frames. ], batch size: 93, lr: 4.57e-03, grad_scale: 8.0 +2023-03-21 08:13:09,188 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3127, 3.6601, 3.1203, 3.1937, 2.9565, 2.6060, 3.3732, 3.0811], + device='cuda:0'), covar=tensor([0.0565, 0.0557, 0.1011, 0.1114, 0.2195, 0.1212, 0.1104, 0.0729], + device='cuda:0'), in_proj_covar=tensor([0.0056, 0.0057, 0.0066, 0.0058, 0.0055, 0.0060, 0.0055, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:13:10,515 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 08:13:10,969 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.215e+02 1.685e+02 2.083e+02 2.423e+02 3.293e+02, threshold=4.166e+02, percent-clipped=0.0 +2023-03-21 08:13:17,438 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-03-21 08:13:20,591 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 08:13:28,709 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 08:13:34,935 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0512, 3.0381, 2.1472, 3.5497, 2.6195, 3.1857, 1.5020, 2.1211], + device='cuda:0'), covar=tensor([0.0626, 0.0942, 0.3094, 0.0818, 0.0631, 0.0777, 0.4129, 0.2205], + device='cuda:0'), in_proj_covar=tensor([0.0282, 0.0261, 0.0285, 0.0269, 0.0274, 0.0267, 0.0239, 0.0262], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:13:35,221 INFO [train.py:901] (0/2) Epoch 36, batch 2050, loss[loss=0.1452, simple_loss=0.233, pruned_loss=0.0287, over 7321.00 frames. ], tot_loss[loss=0.133, simple_loss=0.2144, pruned_loss=0.02581, over 1445400.00 frames. ], batch size: 80, lr: 4.57e-03, grad_scale: 8.0 +2023-03-21 08:13:36,806 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=100894.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:13:50,972 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0061, 3.2574, 2.7941, 3.1585, 3.2044, 2.8173, 3.1708, 3.0637], + device='cuda:0'), covar=tensor([0.0819, 0.0708, 0.1193, 0.0991, 0.1582, 0.0735, 0.1051, 0.1041], + device='cuda:0'), in_proj_covar=tensor([0.0056, 0.0058, 0.0066, 0.0059, 0.0055, 0.0061, 0.0056, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:14:00,396 INFO [train.py:901] (0/2) Epoch 36, batch 2100, loss[loss=0.1359, simple_loss=0.2122, pruned_loss=0.02982, over 7358.00 frames. ], tot_loss[loss=0.1325, simple_loss=0.2137, pruned_loss=0.02565, over 1443769.84 frames. ], batch size: 51, lr: 4.57e-03, grad_scale: 8.0 +2023-03-21 08:14:00,923 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=100942.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:14:01,393 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 08:14:02,342 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.257e+02 1.697e+02 1.994e+02 2.403e+02 4.710e+02, threshold=3.988e+02, percent-clipped=1.0 +2023-03-21 08:14:03,884 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 08:14:05,542 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0526, 2.2493, 2.2213, 3.2279, 1.5890, 3.0746, 1.4056, 2.9607], + device='cuda:0'), covar=tensor([0.0198, 0.1320, 0.1707, 0.0250, 0.4256, 0.0277, 0.1215, 0.0478], + device='cuda:0'), in_proj_covar=tensor([0.0203, 0.0251, 0.0262, 0.0207, 0.0251, 0.0214, 0.0229, 0.0228], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:14:09,662 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6447, 5.1407, 5.1848, 5.1351, 4.8788, 4.6603, 5.2165, 4.9431], + device='cuda:0'), covar=tensor([0.0465, 0.0391, 0.0405, 0.0510, 0.0394, 0.0411, 0.0337, 0.0511], + device='cuda:0'), in_proj_covar=tensor([0.0212, 0.0251, 0.0194, 0.0195, 0.0153, 0.0227, 0.0203, 0.0148], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 08:14:26,896 INFO [train.py:901] (0/2) Epoch 36, batch 2150, loss[loss=0.1224, simple_loss=0.2073, pruned_loss=0.01875, over 7315.00 frames. ], tot_loss[loss=0.1325, simple_loss=0.2137, pruned_loss=0.02563, over 1442861.62 frames. ], batch size: 80, lr: 4.57e-03, grad_scale: 8.0 +2023-03-21 08:14:42,526 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4055, 2.7097, 2.3580, 2.6807, 2.6411, 2.2518, 2.6794, 2.4388], + device='cuda:0'), covar=tensor([0.0638, 0.0574, 0.0908, 0.0616, 0.0889, 0.0940, 0.0606, 0.1035], + device='cuda:0'), in_proj_covar=tensor([0.0056, 0.0057, 0.0067, 0.0058, 0.0055, 0.0060, 0.0056, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:14:45,573 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3309, 2.7642, 2.0194, 2.9391, 2.7994, 3.0196, 2.8365, 2.5927], + device='cuda:0'), covar=tensor([0.2269, 0.1004, 0.4043, 0.0688, 0.0282, 0.0261, 0.0476, 0.0475], + device='cuda:0'), in_proj_covar=tensor([0.0256, 0.0233, 0.0250, 0.0258, 0.0195, 0.0198, 0.0215, 0.0227], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 08:14:50,431 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 08:14:53,041 INFO [train.py:901] (0/2) Epoch 36, batch 2200, loss[loss=0.1351, simple_loss=0.2205, pruned_loss=0.02485, over 7216.00 frames. ], tot_loss[loss=0.1323, simple_loss=0.2137, pruned_loss=0.02541, over 1443669.63 frames. ], batch size: 93, lr: 4.57e-03, grad_scale: 8.0 +2023-03-21 08:14:55,021 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.164e+02 1.755e+02 1.986e+02 2.437e+02 3.637e+02, threshold=3.971e+02, percent-clipped=0.0 +2023-03-21 08:15:01,656 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=101058.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:15:16,341 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=101086.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 08:15:18,783 INFO [train.py:901] (0/2) Epoch 36, batch 2250, loss[loss=0.1402, simple_loss=0.2266, pruned_loss=0.02693, over 7129.00 frames. ], tot_loss[loss=0.1322, simple_loss=0.2133, pruned_loss=0.02551, over 1442815.58 frames. ], batch size: 98, lr: 4.57e-03, grad_scale: 4.0 +2023-03-21 08:15:26,348 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 08:15:26,360 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 08:15:26,399 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=101106.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:15:26,914 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3429, 4.1633, 4.1193, 4.1978, 3.5861, 4.1219, 4.3866, 3.9096], + device='cuda:0'), covar=tensor([0.0233, 0.0191, 0.0175, 0.0211, 0.0762, 0.0149, 0.0230, 0.0234], + device='cuda:0'), in_proj_covar=tensor([0.0096, 0.0103, 0.0103, 0.0091, 0.0179, 0.0108, 0.0106, 0.0113], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:15:26,952 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=101107.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:15:36,104 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1582, 2.1838, 2.3768, 2.0429, 2.2122, 2.2751, 1.9216, 1.7378], + device='cuda:0'), covar=tensor([0.0345, 0.0471, 0.0328, 0.0274, 0.0471, 0.0367, 0.0359, 0.0349], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0039, 0.0040, 0.0039, 0.0036, 0.0037, 0.0042, 0.0041], + device='cuda:0'), out_proj_covar=tensor([1.0059e-04, 1.0033e-04, 9.9874e-05, 9.7518e-05, 9.5007e-05, 9.5443e-05, + 1.0435e-04, 1.0378e-04], device='cuda:0') +2023-03-21 08:15:38,999 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 08:15:44,461 INFO [train.py:901] (0/2) Epoch 36, batch 2300, loss[loss=0.1275, simple_loss=0.2051, pruned_loss=0.02495, over 7301.00 frames. ], tot_loss[loss=0.1321, simple_loss=0.2133, pruned_loss=0.02548, over 1442884.18 frames. ], batch size: 83, lr: 4.57e-03, grad_scale: 4.0 +2023-03-21 08:15:46,986 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.343e+02 1.779e+02 2.010e+02 2.445e+02 5.360e+02, threshold=4.019e+02, percent-clipped=2.0 +2023-03-21 08:15:56,074 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.95 vs. limit=2.0 +2023-03-21 08:15:58,807 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=101168.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:16:00,842 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5365, 2.8609, 2.4876, 2.7805, 2.7390, 2.4278, 2.8742, 2.6146], + device='cuda:0'), covar=tensor([0.0672, 0.0767, 0.1209, 0.0831, 0.1245, 0.0701, 0.0692, 0.1036], + device='cuda:0'), in_proj_covar=tensor([0.0057, 0.0058, 0.0067, 0.0058, 0.0056, 0.0061, 0.0056, 0.0054], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:16:10,346 INFO [train.py:901] (0/2) Epoch 36, batch 2350, loss[loss=0.1432, simple_loss=0.2251, pruned_loss=0.03068, over 7316.00 frames. ], tot_loss[loss=0.1326, simple_loss=0.2137, pruned_loss=0.0257, over 1444117.61 frames. ], batch size: 83, lr: 4.57e-03, grad_scale: 4.0 +2023-03-21 08:16:25,882 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 08:16:32,836 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 08:16:36,980 INFO [train.py:901] (0/2) Epoch 36, batch 2400, loss[loss=0.1455, simple_loss=0.2251, pruned_loss=0.03298, over 7328.00 frames. ], tot_loss[loss=0.1326, simple_loss=0.2134, pruned_loss=0.02593, over 1441769.35 frames. ], batch size: 75, lr: 4.56e-03, grad_scale: 8.0 +2023-03-21 08:16:39,417 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.253e+02 1.751e+02 2.088e+02 2.465e+02 5.704e+02, threshold=4.177e+02, percent-clipped=6.0 +2023-03-21 08:16:44,526 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 08:16:47,538 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 08:17:02,201 INFO [train.py:901] (0/2) Epoch 36, batch 2450, loss[loss=0.1314, simple_loss=0.2213, pruned_loss=0.02075, over 6718.00 frames. ], tot_loss[loss=0.1324, simple_loss=0.213, pruned_loss=0.02587, over 1439498.47 frames. ], batch size: 106, lr: 4.56e-03, grad_scale: 4.0 +2023-03-21 08:17:13,776 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 08:17:20,251 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-03-21 08:17:28,471 INFO [train.py:901] (0/2) Epoch 36, batch 2500, loss[loss=0.1413, simple_loss=0.2174, pruned_loss=0.0326, over 7347.00 frames. ], tot_loss[loss=0.132, simple_loss=0.2125, pruned_loss=0.02571, over 1440574.16 frames. ], batch size: 54, lr: 4.56e-03, grad_scale: 4.0 +2023-03-21 08:17:31,517 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.249e+02 1.796e+02 2.233e+02 2.673e+02 5.552e+02, threshold=4.467e+02, percent-clipped=2.0 +2023-03-21 08:17:39,180 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 08:17:51,785 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=101386.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 08:17:54,137 INFO [train.py:901] (0/2) Epoch 36, batch 2550, loss[loss=0.09464, simple_loss=0.1601, pruned_loss=0.01456, over 6268.00 frames. ], tot_loss[loss=0.1314, simple_loss=0.212, pruned_loss=0.02538, over 1438203.16 frames. ], batch size: 27, lr: 4.56e-03, grad_scale: 4.0 +2023-03-21 08:18:16,600 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=101434.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 08:18:19,946 INFO [train.py:901] (0/2) Epoch 36, batch 2600, loss[loss=0.1425, simple_loss=0.2202, pruned_loss=0.0324, over 7316.00 frames. ], tot_loss[loss=0.1322, simple_loss=0.2128, pruned_loss=0.02582, over 1438406.42 frames. ], batch size: 86, lr: 4.56e-03, grad_scale: 4.0 +2023-03-21 08:18:22,801 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.157e+02 1.775e+02 2.071e+02 2.453e+02 3.538e+02, threshold=4.142e+02, percent-clipped=0.0 +2023-03-21 08:18:23,814 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9806, 4.1491, 3.9473, 4.0887, 3.9249, 3.8866, 4.2482, 4.2943], + device='cuda:0'), covar=tensor([0.0302, 0.0242, 0.0268, 0.0309, 0.0364, 0.0426, 0.0411, 0.0348], + device='cuda:0'), in_proj_covar=tensor([0.0121, 0.0126, 0.0118, 0.0123, 0.0113, 0.0104, 0.0099, 0.0100], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:18:25,339 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3958, 3.0608, 3.3186, 3.4806, 3.1723, 3.0132, 3.4505, 2.6574], + device='cuda:0'), covar=tensor([0.0319, 0.0511, 0.0601, 0.0569, 0.0693, 0.0914, 0.0654, 0.1997], + device='cuda:0'), in_proj_covar=tensor([0.0333, 0.0339, 0.0275, 0.0357, 0.0293, 0.0291, 0.0348, 0.0253], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 08:18:30,714 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=101463.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:18:34,175 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6275, 3.8224, 3.5957, 3.7910, 3.5239, 3.8406, 4.0381, 4.0781], + device='cuda:0'), covar=tensor([0.0251, 0.0171, 0.0263, 0.0176, 0.0353, 0.0286, 0.0280, 0.0217], + device='cuda:0'), in_proj_covar=tensor([0.0121, 0.0126, 0.0118, 0.0123, 0.0113, 0.0104, 0.0099, 0.0100], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:18:44,273 INFO [train.py:901] (0/2) Epoch 36, batch 2650, loss[loss=0.1223, simple_loss=0.2079, pruned_loss=0.01836, over 7329.00 frames. ], tot_loss[loss=0.1328, simple_loss=0.2136, pruned_loss=0.02598, over 1439173.97 frames. ], batch size: 61, lr: 4.56e-03, grad_scale: 4.0 +2023-03-21 08:19:00,198 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0601, 2.4335, 3.1129, 2.9610, 3.1093, 2.7972, 2.5358, 3.0147], + device='cuda:0'), covar=tensor([0.1203, 0.0974, 0.1164, 0.1535, 0.0707, 0.1193, 0.2137, 0.1348], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0067, 0.0051, 0.0050, 0.0050, 0.0049, 0.0068, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 08:19:08,894 INFO [train.py:901] (0/2) Epoch 36, batch 2700, loss[loss=0.1388, simple_loss=0.2231, pruned_loss=0.02726, over 7328.00 frames. ], tot_loss[loss=0.1324, simple_loss=0.2134, pruned_loss=0.02575, over 1439604.93 frames. ], batch size: 61, lr: 4.56e-03, grad_scale: 4.0 +2023-03-21 08:19:09,475 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0341, 4.1934, 3.9964, 4.2417, 3.7895, 4.2761, 4.5127, 4.5203], + device='cuda:0'), covar=tensor([0.0193, 0.0144, 0.0205, 0.0132, 0.0377, 0.0194, 0.0214, 0.0174], + device='cuda:0'), in_proj_covar=tensor([0.0121, 0.0126, 0.0118, 0.0124, 0.0113, 0.0104, 0.0099, 0.0100], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:19:11,761 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.222e+02 1.662e+02 1.940e+02 2.331e+02 3.530e+02, threshold=3.880e+02, percent-clipped=0.0 +2023-03-21 08:19:32,497 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.93 vs. limit=5.0 +2023-03-21 08:19:34,108 INFO [train.py:901] (0/2) Epoch 36, batch 2750, loss[loss=0.1335, simple_loss=0.2196, pruned_loss=0.02366, over 7303.00 frames. ], tot_loss[loss=0.1321, simple_loss=0.2131, pruned_loss=0.02558, over 1439540.08 frames. ], batch size: 70, lr: 4.56e-03, grad_scale: 4.0 +2023-03-21 08:19:39,525 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.85 vs. limit=2.0 +2023-03-21 08:19:58,924 INFO [train.py:901] (0/2) Epoch 36, batch 2800, loss[loss=0.1297, simple_loss=0.2142, pruned_loss=0.02265, over 7248.00 frames. ], tot_loss[loss=0.1325, simple_loss=0.2134, pruned_loss=0.02576, over 1442134.59 frames. ], batch size: 47, lr: 4.56e-03, grad_scale: 8.0 +2023-03-21 08:20:01,506 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7813, 2.6196, 2.4436, 3.7953, 1.9260, 3.7237, 1.6384, 3.3725], + device='cuda:0'), covar=tensor([0.0195, 0.1304, 0.1810, 0.0218, 0.3821, 0.0251, 0.1198, 0.0404], + device='cuda:0'), in_proj_covar=tensor([0.0202, 0.0246, 0.0258, 0.0205, 0.0248, 0.0212, 0.0227, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:20:01,831 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.167e+02 1.705e+02 1.985e+02 2.296e+02 4.387e+02, threshold=3.971e+02, percent-clipped=1.0 +2023-03-21 08:20:06,371 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8480, 3.4092, 3.9514, 3.9764, 4.1232, 3.9984, 4.0524, 3.8749], + device='cuda:0'), covar=tensor([0.0041, 0.0124, 0.0038, 0.0039, 0.0028, 0.0037, 0.0043, 0.0055], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0068, 0.0056, 0.0055, 0.0054, 0.0060, 0.0048, 0.0075], + device='cuda:0'), out_proj_covar=tensor([8.2439e-05, 1.4048e-04, 1.0358e-04, 9.7488e-05, 9.4255e-05, 1.0596e-04, + 9.2333e-05, 1.4140e-04], device='cuda:0') +2023-03-21 08:20:11,475 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-36.pt +2023-03-21 08:20:27,149 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 08:20:27,204 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 08:20:27,230 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 08:20:27,246 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 08:20:27,421 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 08:20:27,437 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 08:20:27,448 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 08:20:27,546 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 08:20:27,622 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 08:20:27,636 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 08:20:27,654 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 08:20:27,816 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 08:20:28,149 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 08:20:28,157 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 08:20:28,296 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 08:20:28,299 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 08:20:28,357 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 08:20:30,567 INFO [train.py:901] (0/2) Epoch 37, batch 0, loss[loss=0.1195, simple_loss=0.2058, pruned_loss=0.01653, over 7281.00 frames. ], tot_loss[loss=0.1195, simple_loss=0.2058, pruned_loss=0.01653, over 7281.00 frames. ], batch size: 70, lr: 4.49e-03, grad_scale: 8.0 +2023-03-21 08:20:30,569 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 08:20:43,499 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6573, 4.9226, 4.9649, 4.8982, 4.7003, 4.5485, 4.9735, 4.7158], + device='cuda:0'), covar=tensor([0.0471, 0.0365, 0.0351, 0.0478, 0.0314, 0.0345, 0.0321, 0.0465], + device='cuda:0'), in_proj_covar=tensor([0.0212, 0.0251, 0.0194, 0.0196, 0.0153, 0.0227, 0.0202, 0.0148], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 08:20:46,959 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5411, 4.1810, 3.8483, 4.6344, 4.3432, 4.6389, 4.2916, 4.4225], + device='cuda:0'), covar=tensor([0.0797, 0.2111, 0.1992, 0.0992, 0.0838, 0.0963, 0.0572, 0.0839], + device='cuda:0'), in_proj_covar=tensor([0.0255, 0.0383, 0.0291, 0.0300, 0.0226, 0.0357, 0.0221, 0.0266], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:20:57,446 INFO [train.py:935] (0/2) Epoch 37, validation: loss=0.1643, simple_loss=0.2561, pruned_loss=0.03627, over 1622729.00 frames. +2023-03-21 08:20:57,447 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 08:21:04,453 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 08:21:07,078 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=101684.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:21:09,556 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0480, 3.5396, 4.0486, 4.0655, 4.0880, 4.1053, 4.2508, 4.0167], + device='cuda:0'), covar=tensor([0.0031, 0.0101, 0.0027, 0.0029, 0.0027, 0.0030, 0.0030, 0.0044], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0068, 0.0056, 0.0055, 0.0054, 0.0060, 0.0048, 0.0075], + device='cuda:0'), out_proj_covar=tensor([8.2267e-05, 1.4026e-04, 1.0340e-04, 9.7366e-05, 9.4057e-05, 1.0580e-04, + 9.2273e-05, 1.4133e-04], device='cuda:0') +2023-03-21 08:21:15,546 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 08:21:22,118 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 08:21:22,594 INFO [train.py:901] (0/2) Epoch 37, batch 50, loss[loss=0.1343, simple_loss=0.2143, pruned_loss=0.02711, over 7269.00 frames. ], tot_loss[loss=0.1317, simple_loss=0.2125, pruned_loss=0.02542, over 325050.30 frames. ], batch size: 57, lr: 4.49e-03, grad_scale: 8.0 +2023-03-21 08:21:24,123 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 08:21:27,120 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 08:21:37,749 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=101745.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:21:38,567 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.229e+02 1.749e+02 1.978e+02 2.334e+02 3.744e+02, threshold=3.956e+02, percent-clipped=0.0 +2023-03-21 08:21:43,894 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 08:21:44,394 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 08:21:47,956 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=101763.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:21:48,820 INFO [train.py:901] (0/2) Epoch 37, batch 100, loss[loss=0.1271, simple_loss=0.2108, pruned_loss=0.02173, over 7261.00 frames. ], tot_loss[loss=0.1331, simple_loss=0.2149, pruned_loss=0.02568, over 574714.89 frames. ], batch size: 64, lr: 4.49e-03, grad_scale: 8.0 +2023-03-21 08:22:12,331 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=101811.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:22:14,207 INFO [train.py:901] (0/2) Epoch 37, batch 150, loss[loss=0.1274, simple_loss=0.2121, pruned_loss=0.02137, over 7344.00 frames. ], tot_loss[loss=0.1333, simple_loss=0.2147, pruned_loss=0.02596, over 766168.45 frames. ], batch size: 44, lr: 4.49e-03, grad_scale: 8.0 +2023-03-21 08:22:31,469 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.264e+02 1.790e+02 2.048e+02 2.419e+02 6.579e+02, threshold=4.096e+02, percent-clipped=3.0 +2023-03-21 08:22:40,656 INFO [train.py:901] (0/2) Epoch 37, batch 200, loss[loss=0.1015, simple_loss=0.1718, pruned_loss=0.01564, over 7017.00 frames. ], tot_loss[loss=0.1329, simple_loss=0.214, pruned_loss=0.02588, over 917327.59 frames. ], batch size: 35, lr: 4.49e-03, grad_scale: 8.0 +2023-03-21 08:22:46,007 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.83 vs. limit=2.0 +2023-03-21 08:22:46,243 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 08:22:50,731 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 08:22:56,865 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 08:23:05,979 INFO [train.py:901] (0/2) Epoch 37, batch 250, loss[loss=0.1435, simple_loss=0.2223, pruned_loss=0.03239, over 7274.00 frames. ], tot_loss[loss=0.1322, simple_loss=0.2132, pruned_loss=0.02562, over 1033756.66 frames. ], batch size: 70, lr: 4.49e-03, grad_scale: 8.0 +2023-03-21 08:23:08,538 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 08:23:08,642 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=101920.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 08:23:23,073 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.200e+02 1.857e+02 2.138e+02 2.562e+02 4.805e+02, threshold=4.275e+02, percent-clipped=2.0 +2023-03-21 08:23:23,277 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0420, 2.4749, 1.8298, 2.6757, 2.7633, 2.6352, 2.5503, 2.5136], + device='cuda:0'), covar=tensor([0.2463, 0.1060, 0.4202, 0.0856, 0.0306, 0.0355, 0.0352, 0.0467], + device='cuda:0'), in_proj_covar=tensor([0.0252, 0.0228, 0.0247, 0.0255, 0.0194, 0.0196, 0.0214, 0.0224], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 08:23:31,152 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 08:23:32,139 INFO [train.py:901] (0/2) Epoch 37, batch 300, loss[loss=0.1166, simple_loss=0.1939, pruned_loss=0.01966, over 7185.00 frames. ], tot_loss[loss=0.1317, simple_loss=0.2123, pruned_loss=0.02554, over 1121503.59 frames. ], batch size: 39, lr: 4.49e-03, grad_scale: 8.0 +2023-03-21 08:23:40,198 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 08:23:40,304 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=101981.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 08:23:43,340 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=101987.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:23:57,666 INFO [train.py:901] (0/2) Epoch 37, batch 350, loss[loss=0.1487, simple_loss=0.2301, pruned_loss=0.03363, over 7282.00 frames. ], tot_loss[loss=0.1316, simple_loss=0.2123, pruned_loss=0.02551, over 1193472.46 frames. ], batch size: 57, lr: 4.49e-03, grad_scale: 8.0 +2023-03-21 08:24:11,553 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=102040.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:24:14,985 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.224e+02 1.770e+02 2.044e+02 2.367e+02 3.943e+02, threshold=4.088e+02, percent-clipped=0.0 +2023-03-21 08:24:15,490 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 08:24:15,622 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=102048.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:24:22,787 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.80 vs. limit=2.0 +2023-03-21 08:24:23,986 INFO [train.py:901] (0/2) Epoch 37, batch 400, loss[loss=0.1369, simple_loss=0.2123, pruned_loss=0.03071, over 7291.00 frames. ], tot_loss[loss=0.1314, simple_loss=0.2121, pruned_loss=0.02531, over 1247179.14 frames. ], batch size: 47, lr: 4.48e-03, grad_scale: 8.0 +2023-03-21 08:24:28,693 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0485, 2.6406, 3.0151, 3.1196, 2.7136, 2.5713, 3.0298, 2.2590], + device='cuda:0'), covar=tensor([0.0512, 0.0600, 0.0786, 0.0709, 0.0685, 0.1010, 0.0710, 0.2146], + device='cuda:0'), in_proj_covar=tensor([0.0333, 0.0338, 0.0275, 0.0357, 0.0293, 0.0289, 0.0347, 0.0253], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 08:24:31,458 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.57 vs. limit=2.0 +2023-03-21 08:24:42,803 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6150, 4.1003, 4.0312, 4.5551, 4.4127, 4.5100, 3.7349, 4.1846], + device='cuda:0'), covar=tensor([0.0998, 0.3145, 0.2714, 0.1501, 0.1151, 0.1525, 0.1131, 0.1354], + device='cuda:0'), in_proj_covar=tensor([0.0258, 0.0390, 0.0296, 0.0305, 0.0228, 0.0363, 0.0222, 0.0270], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:24:50,341 INFO [train.py:901] (0/2) Epoch 37, batch 450, loss[loss=0.1185, simple_loss=0.1884, pruned_loss=0.02428, over 7194.00 frames. ], tot_loss[loss=0.1313, simple_loss=0.212, pruned_loss=0.02528, over 1290586.03 frames. ], batch size: 39, lr: 4.48e-03, grad_scale: 8.0 +2023-03-21 08:24:56,914 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 08:24:57,394 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 08:25:06,270 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.299e+02 1.769e+02 2.012e+02 2.507e+02 5.212e+02, threshold=4.024e+02, percent-clipped=1.0 +2023-03-21 08:25:15,490 INFO [train.py:901] (0/2) Epoch 37, batch 500, loss[loss=0.1306, simple_loss=0.2178, pruned_loss=0.02166, over 7272.00 frames. ], tot_loss[loss=0.1312, simple_loss=0.212, pruned_loss=0.02526, over 1324432.47 frames. ], batch size: 52, lr: 4.48e-03, grad_scale: 8.0 +2023-03-21 08:25:30,225 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 08:25:31,732 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 08:25:32,217 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 08:25:35,133 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 08:25:39,649 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 08:25:42,202 INFO [train.py:901] (0/2) Epoch 37, batch 550, loss[loss=0.1464, simple_loss=0.2222, pruned_loss=0.03528, over 7333.00 frames. ], tot_loss[loss=0.1318, simple_loss=0.2125, pruned_loss=0.0256, over 1351449.74 frames. ], batch size: 54, lr: 4.48e-03, grad_scale: 8.0 +2023-03-21 08:25:50,748 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 08:25:58,232 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.288e+02 1.789e+02 2.159e+02 2.504e+02 4.340e+02, threshold=4.318e+02, percent-clipped=1.0 +2023-03-21 08:25:58,781 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 08:26:01,846 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 08:26:07,368 INFO [train.py:901] (0/2) Epoch 37, batch 600, loss[loss=0.1383, simple_loss=0.2194, pruned_loss=0.02862, over 7254.00 frames. ], tot_loss[loss=0.1318, simple_loss=0.2128, pruned_loss=0.02543, over 1372615.71 frames. ], batch size: 55, lr: 4.48e-03, grad_scale: 8.0 +2023-03-21 08:26:07,964 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4452, 4.9589, 5.0168, 4.9428, 4.8555, 4.4830, 5.0598, 4.8976], + device='cuda:0'), covar=tensor([0.0477, 0.0386, 0.0398, 0.0505, 0.0321, 0.0419, 0.0320, 0.0494], + device='cuda:0'), in_proj_covar=tensor([0.0214, 0.0253, 0.0198, 0.0198, 0.0155, 0.0230, 0.0203, 0.0150], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 08:26:08,398 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 08:26:13,641 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=102276.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 08:26:25,430 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 08:26:33,501 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 08:26:33,989 INFO [train.py:901] (0/2) Epoch 37, batch 650, loss[loss=0.1457, simple_loss=0.2259, pruned_loss=0.03272, over 7279.00 frames. ], tot_loss[loss=0.131, simple_loss=0.212, pruned_loss=0.02502, over 1386331.96 frames. ], batch size: 77, lr: 4.48e-03, grad_scale: 8.0 +2023-03-21 08:26:46,880 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=102340.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:26:48,405 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=102343.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:26:49,898 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 08:26:50,340 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.201e+02 1.711e+02 1.994e+02 2.328e+02 4.780e+02, threshold=3.988e+02, percent-clipped=1.0 +2023-03-21 08:26:59,331 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 08:27:00,862 INFO [train.py:901] (0/2) Epoch 37, batch 700, loss[loss=0.1291, simple_loss=0.2167, pruned_loss=0.02071, over 7337.00 frames. ], tot_loss[loss=0.1322, simple_loss=0.2134, pruned_loss=0.02551, over 1400265.60 frames. ], batch size: 75, lr: 4.48e-03, grad_scale: 8.0 +2023-03-21 08:27:12,687 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=102388.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:27:12,742 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6587, 4.0898, 4.1923, 4.3185, 4.2090, 4.1385, 4.4558, 3.9138], + device='cuda:0'), covar=tensor([0.0096, 0.0177, 0.0129, 0.0133, 0.0425, 0.0114, 0.0146, 0.0182], + device='cuda:0'), in_proj_covar=tensor([0.0095, 0.0102, 0.0102, 0.0089, 0.0177, 0.0107, 0.0105, 0.0111], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:27:13,872 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9281, 3.2419, 2.7880, 3.0814, 3.0683, 2.6470, 3.1283, 2.8427], + device='cuda:0'), covar=tensor([0.0535, 0.0667, 0.0959, 0.0696, 0.0843, 0.0836, 0.0761, 0.1313], + device='cuda:0'), in_proj_covar=tensor([0.0055, 0.0056, 0.0064, 0.0057, 0.0054, 0.0059, 0.0055, 0.0051], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:27:22,679 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.70 vs. limit=2.0 +2023-03-21 08:27:23,306 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 08:27:24,270 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 08:27:26,748 INFO [train.py:901] (0/2) Epoch 37, batch 750, loss[loss=0.1261, simple_loss=0.21, pruned_loss=0.02115, over 7331.00 frames. ], tot_loss[loss=0.1323, simple_loss=0.213, pruned_loss=0.02581, over 1407327.09 frames. ], batch size: 61, lr: 4.48e-03, grad_scale: 8.0 +2023-03-21 08:27:37,922 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 08:27:42,766 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 08:27:43,784 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.077e+02 1.733e+02 2.069e+02 2.404e+02 6.127e+02, threshold=4.137e+02, percent-clipped=2.0 +2023-03-21 08:27:49,059 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 08:27:50,570 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 08:27:53,670 INFO [train.py:901] (0/2) Epoch 37, batch 800, loss[loss=0.1306, simple_loss=0.2083, pruned_loss=0.02642, over 7285.00 frames. ], tot_loss[loss=0.1319, simple_loss=0.2129, pruned_loss=0.02551, over 1415222.87 frames. ], batch size: 70, lr: 4.48e-03, grad_scale: 8.0 +2023-03-21 08:28:00,661 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 08:28:05,710 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9208, 4.4851, 4.4337, 4.9622, 4.8304, 4.8402, 4.1357, 4.4631], + device='cuda:0'), covar=tensor([0.0971, 0.2696, 0.2243, 0.1159, 0.0832, 0.1431, 0.0922, 0.1316], + device='cuda:0'), in_proj_covar=tensor([0.0260, 0.0386, 0.0294, 0.0304, 0.0227, 0.0365, 0.0222, 0.0269], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:28:09,916 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.85 vs. limit=2.0 +2023-03-21 08:28:18,806 INFO [train.py:901] (0/2) Epoch 37, batch 850, loss[loss=0.1293, simple_loss=0.203, pruned_loss=0.02776, over 7225.00 frames. ], tot_loss[loss=0.1325, simple_loss=0.2136, pruned_loss=0.02568, over 1422559.35 frames. ], batch size: 45, lr: 4.47e-03, grad_scale: 8.0 +2023-03-21 08:28:18,969 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2403, 2.2709, 2.3014, 3.3251, 1.7429, 3.2960, 1.4838, 3.1017], + device='cuda:0'), covar=tensor([0.0216, 0.1285, 0.1851, 0.0232, 0.4062, 0.0264, 0.1308, 0.0402], + device='cuda:0'), in_proj_covar=tensor([0.0201, 0.0246, 0.0258, 0.0205, 0.0248, 0.0212, 0.0229, 0.0225], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:28:19,333 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 08:28:19,766 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 08:28:25,907 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 08:28:29,378 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 08:28:34,009 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.2045, 4.7411, 4.7382, 5.2710, 5.1222, 5.1551, 4.5737, 4.8074], + device='cuda:0'), covar=tensor([0.0794, 0.2357, 0.1860, 0.0860, 0.0710, 0.1081, 0.0610, 0.0903], + device='cuda:0'), in_proj_covar=tensor([0.0260, 0.0385, 0.0292, 0.0303, 0.0227, 0.0364, 0.0222, 0.0270], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:28:35,937 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.128e+02 1.881e+02 2.184e+02 2.686e+02 4.257e+02, threshold=4.368e+02, percent-clipped=1.0 +2023-03-21 08:28:45,135 INFO [train.py:901] (0/2) Epoch 37, batch 900, loss[loss=0.1298, simple_loss=0.2138, pruned_loss=0.02285, over 7360.00 frames. ], tot_loss[loss=0.1318, simple_loss=0.2129, pruned_loss=0.02537, over 1423353.86 frames. ], batch size: 63, lr: 4.47e-03, grad_scale: 8.0 +2023-03-21 08:28:46,598 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.95 vs. limit=2.0 +2023-03-21 08:28:50,849 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=102576.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 08:29:07,840 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 08:29:11,423 INFO [train.py:901] (0/2) Epoch 37, batch 950, loss[loss=0.1402, simple_loss=0.2121, pruned_loss=0.03421, over 7344.00 frames. ], tot_loss[loss=0.132, simple_loss=0.213, pruned_loss=0.02554, over 1427561.38 frames. ], batch size: 51, lr: 4.47e-03, grad_scale: 8.0 +2023-03-21 08:29:16,718 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=102624.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 08:29:24,277 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8576, 4.3525, 4.4175, 4.3635, 4.4033, 3.9777, 4.4387, 4.3817], + device='cuda:0'), covar=tensor([0.0566, 0.0466, 0.0425, 0.0516, 0.0313, 0.0484, 0.0372, 0.0395], + device='cuda:0'), in_proj_covar=tensor([0.0214, 0.0254, 0.0198, 0.0198, 0.0155, 0.0229, 0.0204, 0.0150], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 08:29:26,342 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=102643.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:29:28,170 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.204e+02 1.662e+02 1.972e+02 2.353e+02 5.734e+02, threshold=3.944e+02, percent-clipped=1.0 +2023-03-21 08:29:28,350 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6732, 2.4083, 2.4764, 3.6627, 1.9380, 3.5024, 1.6844, 3.2079], + device='cuda:0'), covar=tensor([0.0161, 0.1376, 0.1719, 0.0187, 0.3718, 0.0215, 0.1115, 0.0423], + device='cuda:0'), in_proj_covar=tensor([0.0204, 0.0249, 0.0262, 0.0207, 0.0251, 0.0213, 0.0231, 0.0227], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:29:31,746 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 08:29:31,859 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6419, 2.2511, 2.9497, 2.6029, 2.8700, 2.7074, 2.3995, 2.8118], + device='cuda:0'), covar=tensor([0.1773, 0.1155, 0.1138, 0.1538, 0.0940, 0.1275, 0.2029, 0.1252], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0068, 0.0051, 0.0050, 0.0050, 0.0049, 0.0068, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 08:29:37,269 INFO [train.py:901] (0/2) Epoch 37, batch 1000, loss[loss=0.1424, simple_loss=0.2279, pruned_loss=0.02841, over 7336.00 frames. ], tot_loss[loss=0.132, simple_loss=0.2133, pruned_loss=0.02538, over 1430825.79 frames. ], batch size: 61, lr: 4.47e-03, grad_scale: 8.0 +2023-03-21 08:29:39,868 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9038, 4.4381, 4.3861, 4.9815, 4.8098, 4.8168, 4.2591, 4.4693], + device='cuda:0'), covar=tensor([0.0898, 0.2452, 0.2355, 0.0974, 0.0865, 0.1179, 0.0800, 0.1295], + device='cuda:0'), in_proj_covar=tensor([0.0261, 0.0388, 0.0297, 0.0307, 0.0228, 0.0367, 0.0225, 0.0272], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:29:50,344 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=102691.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:29:52,274 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 08:30:03,336 INFO [train.py:901] (0/2) Epoch 37, batch 1050, loss[loss=0.1262, simple_loss=0.2062, pruned_loss=0.02311, over 7307.00 frames. ], tot_loss[loss=0.1319, simple_loss=0.213, pruned_loss=0.02542, over 1433478.70 frames. ], batch size: 80, lr: 4.47e-03, grad_scale: 8.0 +2023-03-21 08:30:03,982 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3058, 1.5369, 1.3584, 1.4701, 1.5613, 1.5478, 1.5402, 1.2247], + device='cuda:0'), covar=tensor([0.0169, 0.0174, 0.0211, 0.0132, 0.0141, 0.0130, 0.0105, 0.0176], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0036, 0.0034, 0.0036, 0.0033, 0.0033, 0.0036, 0.0044], + device='cuda:0'), out_proj_covar=tensor([4.2197e-05, 3.9438e-05, 3.8536e-05, 3.9425e-05, 3.7132e-05, 3.6706e-05, + 4.0793e-05, 4.8551e-05], device='cuda:0') +2023-03-21 08:30:14,452 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 08:30:18,499 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 08:30:19,453 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.402e+02 1.793e+02 2.007e+02 2.334e+02 3.523e+02, threshold=4.014e+02, percent-clipped=0.0 +2023-03-21 08:30:28,471 INFO [train.py:901] (0/2) Epoch 37, batch 1100, loss[loss=0.1171, simple_loss=0.2047, pruned_loss=0.01475, over 7282.00 frames. ], tot_loss[loss=0.1314, simple_loss=0.2124, pruned_loss=0.02521, over 1435884.77 frames. ], batch size: 66, lr: 4.47e-03, grad_scale: 8.0 +2023-03-21 08:30:35,809 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.85 vs. limit=2.0 +2023-03-21 08:30:49,093 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 08:30:49,108 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 08:30:55,026 INFO [train.py:901] (0/2) Epoch 37, batch 1150, loss[loss=0.123, simple_loss=0.2131, pruned_loss=0.01642, over 7339.00 frames. ], tot_loss[loss=0.1305, simple_loss=0.2118, pruned_loss=0.02467, over 1438584.17 frames. ], batch size: 73, lr: 4.47e-03, grad_scale: 8.0 +2023-03-21 08:30:58,706 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4017, 2.3733, 2.6824, 2.3425, 2.3993, 2.2186, 2.1319, 2.0498], + device='cuda:0'), covar=tensor([0.0373, 0.0586, 0.0206, 0.0355, 0.0578, 0.0427, 0.0316, 0.0288], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0037, 0.0038, 0.0036, 0.0035, 0.0035, 0.0041, 0.0039], + device='cuda:0'), out_proj_covar=tensor([9.6006e-05, 9.6615e-05, 9.5464e-05, 9.2838e-05, 9.1633e-05, 9.1377e-05, + 9.9988e-05, 9.9446e-05], device='cuda:0') +2023-03-21 08:31:02,022 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 08:31:02,505 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 08:31:10,923 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.364e+02 1.647e+02 1.990e+02 2.353e+02 6.242e+02, threshold=3.979e+02, percent-clipped=4.0 +2023-03-21 08:31:13,095 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=102851.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:31:19,892 INFO [train.py:901] (0/2) Epoch 37, batch 1200, loss[loss=0.1257, simple_loss=0.211, pruned_loss=0.02022, over 7248.00 frames. ], tot_loss[loss=0.1303, simple_loss=0.2116, pruned_loss=0.02454, over 1438970.04 frames. ], batch size: 55, lr: 4.47e-03, grad_scale: 8.0 +2023-03-21 08:31:31,513 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=102887.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:31:34,028 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 08:31:40,246 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5696, 2.9441, 2.6289, 2.7869, 2.8295, 2.4734, 2.8221, 2.7590], + device='cuda:0'), covar=tensor([0.1000, 0.1343, 0.0949, 0.1064, 0.1095, 0.0847, 0.0892, 0.1068], + device='cuda:0'), in_proj_covar=tensor([0.0055, 0.0057, 0.0065, 0.0057, 0.0055, 0.0060, 0.0055, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:31:42,209 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9611, 4.5283, 4.3023, 4.9656, 4.7058, 4.8982, 4.3121, 4.6037], + device='cuda:0'), covar=tensor([0.0840, 0.2702, 0.2340, 0.1162, 0.1018, 0.1233, 0.0764, 0.1104], + device='cuda:0'), in_proj_covar=tensor([0.0257, 0.0384, 0.0294, 0.0303, 0.0227, 0.0365, 0.0223, 0.0270], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:31:42,461 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.67 vs. limit=2.0 +2023-03-21 08:31:44,838 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=102912.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:31:46,203 INFO [train.py:901] (0/2) Epoch 37, batch 1250, loss[loss=0.1153, simple_loss=0.1995, pruned_loss=0.01559, over 7149.00 frames. ], tot_loss[loss=0.131, simple_loss=0.2122, pruned_loss=0.02494, over 1439670.73 frames. ], batch size: 41, lr: 4.47e-03, grad_scale: 8.0 +2023-03-21 08:31:57,759 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 08:32:02,204 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.332e+02 1.760e+02 2.143e+02 2.522e+02 3.461e+02, threshold=4.287e+02, percent-clipped=0.0 +2023-03-21 08:32:02,721 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 08:32:02,870 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=102948.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:32:04,261 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 08:32:11,326 INFO [train.py:901] (0/2) Epoch 37, batch 1300, loss[loss=0.1249, simple_loss=0.2032, pruned_loss=0.02325, over 7269.00 frames. ], tot_loss[loss=0.1306, simple_loss=0.2118, pruned_loss=0.02471, over 1439957.74 frames. ], batch size: 47, lr: 4.46e-03, grad_scale: 8.0 +2023-03-21 08:32:27,557 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 08:32:30,851 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 08:32:32,146 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.00 vs. limit=2.0 +2023-03-21 08:32:34,308 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 08:32:37,938 INFO [train.py:901] (0/2) Epoch 37, batch 1350, loss[loss=0.1282, simple_loss=0.2154, pruned_loss=0.02053, over 7228.00 frames. ], tot_loss[loss=0.1307, simple_loss=0.212, pruned_loss=0.02467, over 1441224.70 frames. ], batch size: 93, lr: 4.46e-03, grad_scale: 8.0 +2023-03-21 08:32:44,082 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 08:32:54,026 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.199e+02 1.761e+02 2.055e+02 2.469e+02 4.318e+02, threshold=4.111e+02, percent-clipped=1.0 +2023-03-21 08:33:01,051 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-03-21 08:33:04,300 INFO [train.py:901] (0/2) Epoch 37, batch 1400, loss[loss=0.1311, simple_loss=0.2133, pruned_loss=0.02448, over 7278.00 frames. ], tot_loss[loss=0.1306, simple_loss=0.2119, pruned_loss=0.02468, over 1441767.46 frames. ], batch size: 57, lr: 4.46e-03, grad_scale: 8.0 +2023-03-21 08:33:16,373 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 08:33:21,120 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=103098.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:33:29,232 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2053, 2.8221, 2.0575, 3.1239, 3.1524, 3.1760, 2.9460, 2.9365], + device='cuda:0'), covar=tensor([0.2344, 0.1024, 0.3798, 0.0691, 0.0286, 0.0243, 0.0434, 0.0496], + device='cuda:0'), in_proj_covar=tensor([0.0251, 0.0229, 0.0247, 0.0256, 0.0196, 0.0195, 0.0214, 0.0225], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 08:33:29,581 INFO [train.py:901] (0/2) Epoch 37, batch 1450, loss[loss=0.1372, simple_loss=0.2179, pruned_loss=0.0283, over 7320.00 frames. ], tot_loss[loss=0.1307, simple_loss=0.2119, pruned_loss=0.02475, over 1441185.48 frames. ], batch size: 59, lr: 4.46e-03, grad_scale: 8.0 +2023-03-21 08:33:38,134 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=103132.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:33:39,619 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 08:33:46,848 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.242e+02 1.709e+02 2.082e+02 2.511e+02 4.451e+02, threshold=4.164e+02, percent-clipped=1.0 +2023-03-21 08:33:50,512 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6992, 2.0560, 2.7457, 2.5809, 2.5192, 2.3134, 2.0940, 2.6531], + device='cuda:0'), covar=tensor([0.1308, 0.1165, 0.0968, 0.1039, 0.1267, 0.1200, 0.2183, 0.1269], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0067, 0.0050, 0.0049, 0.0050, 0.0048, 0.0068, 0.0050], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 08:33:53,020 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=103159.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:33:55,873 INFO [train.py:901] (0/2) Epoch 37, batch 1500, loss[loss=0.1222, simple_loss=0.2074, pruned_loss=0.01846, over 7297.00 frames. ], tot_loss[loss=0.1307, simple_loss=0.2118, pruned_loss=0.02482, over 1440346.75 frames. ], batch size: 70, lr: 4.46e-03, grad_scale: 8.0 +2023-03-21 08:33:56,385 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 08:34:09,571 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8037, 2.8361, 2.1086, 3.1695, 2.3996, 2.7959, 1.4336, 2.0969], + device='cuda:0'), covar=tensor([0.0543, 0.0998, 0.2513, 0.0803, 0.0539, 0.0732, 0.3806, 0.1798], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0257, 0.0280, 0.0268, 0.0272, 0.0265, 0.0235, 0.0260], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:34:10,037 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=103193.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:34:17,271 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=103207.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:34:18,837 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9955, 2.4413, 3.1085, 2.9862, 3.0468, 2.9284, 2.5083, 3.1142], + device='cuda:0'), covar=tensor([0.1536, 0.0966, 0.1078, 0.1230, 0.1144, 0.1112, 0.1887, 0.1077], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0068, 0.0051, 0.0049, 0.0051, 0.0049, 0.0069, 0.0051], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 08:34:19,713 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 08:34:21,184 INFO [train.py:901] (0/2) Epoch 37, batch 1550, loss[loss=0.1294, simple_loss=0.2099, pruned_loss=0.02439, over 7346.00 frames. ], tot_loss[loss=0.1308, simple_loss=0.212, pruned_loss=0.02485, over 1442441.54 frames. ], batch size: 63, lr: 4.46e-03, grad_scale: 8.0 +2023-03-21 08:34:21,774 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9883, 4.5820, 4.3662, 5.0391, 4.7730, 4.9130, 4.4870, 4.5723], + device='cuda:0'), covar=tensor([0.0901, 0.2255, 0.2484, 0.0916, 0.1004, 0.1083, 0.0741, 0.1199], + device='cuda:0'), in_proj_covar=tensor([0.0259, 0.0385, 0.0294, 0.0305, 0.0226, 0.0363, 0.0225, 0.0272], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:34:36,659 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=103243.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:34:38,622 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.258e+02 1.785e+02 2.115e+02 2.545e+02 5.224e+02, threshold=4.230e+02, percent-clipped=3.0 +2023-03-21 08:34:44,229 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3859, 2.5617, 2.3746, 2.6259, 2.6457, 2.3520, 2.6798, 2.4659], + device='cuda:0'), covar=tensor([0.0862, 0.1053, 0.1004, 0.0992, 0.0544, 0.0621, 0.0732, 0.1379], + device='cuda:0'), in_proj_covar=tensor([0.0057, 0.0058, 0.0067, 0.0059, 0.0057, 0.0061, 0.0057, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:34:47,673 INFO [train.py:901] (0/2) Epoch 37, batch 1600, loss[loss=0.1015, simple_loss=0.1629, pruned_loss=0.02004, over 5913.00 frames. ], tot_loss[loss=0.131, simple_loss=0.2123, pruned_loss=0.02487, over 1440250.37 frames. ], batch size: 25, lr: 4.46e-03, grad_scale: 8.0 +2023-03-21 08:34:51,706 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 08:34:52,679 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 08:34:55,096 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 08:35:05,228 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 08:35:07,018 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.96 vs. limit=5.0 +2023-03-21 08:35:08,764 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9459, 2.9036, 2.6275, 3.9547, 1.8970, 3.8695, 1.7050, 3.3247], + device='cuda:0'), covar=tensor([0.0226, 0.1152, 0.1793, 0.0235, 0.4155, 0.0274, 0.1292, 0.0512], + device='cuda:0'), in_proj_covar=tensor([0.0206, 0.0252, 0.0265, 0.0211, 0.0254, 0.0216, 0.0233, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:35:10,140 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 08:35:13,209 INFO [train.py:901] (0/2) Epoch 37, batch 1650, loss[loss=0.1258, simple_loss=0.1988, pruned_loss=0.0264, over 7327.00 frames. ], tot_loss[loss=0.131, simple_loss=0.2119, pruned_loss=0.02499, over 1440090.79 frames. ], batch size: 49, lr: 4.46e-03, grad_scale: 16.0 +2023-03-21 08:35:14,123 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.97 vs. limit=5.0 +2023-03-21 08:35:19,373 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 08:35:30,004 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.290e+02 1.789e+02 2.098e+02 2.571e+02 4.173e+02, threshold=4.196e+02, percent-clipped=0.0 +2023-03-21 08:35:36,025 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 08:35:39,054 INFO [train.py:901] (0/2) Epoch 37, batch 1700, loss[loss=0.135, simple_loss=0.2205, pruned_loss=0.02472, over 7286.00 frames. ], tot_loss[loss=0.1311, simple_loss=0.2121, pruned_loss=0.02509, over 1441732.84 frames. ], batch size: 66, lr: 4.46e-03, grad_scale: 16.0 +2023-03-21 08:35:39,578 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 08:35:39,670 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=103366.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 08:35:44,174 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3253, 4.0643, 3.6437, 3.8874, 3.4882, 2.6394, 2.0645, 4.3986], + device='cuda:0'), covar=tensor([0.0052, 0.0105, 0.0113, 0.0063, 0.0138, 0.0522, 0.0624, 0.0048], + device='cuda:0'), in_proj_covar=tensor([0.0114, 0.0094, 0.0114, 0.0094, 0.0130, 0.0137, 0.0132, 0.0104], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 08:35:50,564 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 08:35:52,248 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3731, 1.7232, 1.3928, 1.4871, 1.7473, 1.6368, 1.6450, 1.2375], + device='cuda:0'), covar=tensor([0.0173, 0.0165, 0.0273, 0.0191, 0.0105, 0.0119, 0.0189, 0.0229], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0035, 0.0034, 0.0035, 0.0033, 0.0033, 0.0036, 0.0044], + device='cuda:0'), out_proj_covar=tensor([4.1771e-05, 3.9122e-05, 3.8212e-05, 3.8900e-05, 3.7187e-05, 3.6796e-05, + 4.0802e-05, 4.8311e-05], device='cuda:0') +2023-03-21 08:36:05,743 INFO [train.py:901] (0/2) Epoch 37, batch 1750, loss[loss=0.1175, simple_loss=0.2022, pruned_loss=0.01636, over 7328.00 frames. ], tot_loss[loss=0.1315, simple_loss=0.2126, pruned_loss=0.02518, over 1440801.10 frames. ], batch size: 44, lr: 4.46e-03, grad_scale: 8.0 +2023-03-21 08:36:07,403 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9815, 3.2783, 2.6135, 3.0497, 3.1096, 2.8281, 3.0573, 2.9319], + device='cuda:0'), covar=tensor([0.0725, 0.0601, 0.1437, 0.0875, 0.0987, 0.0669, 0.1068, 0.1044], + device='cuda:0'), in_proj_covar=tensor([0.0056, 0.0057, 0.0066, 0.0058, 0.0056, 0.0060, 0.0056, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:36:11,960 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=103427.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 08:36:15,899 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 08:36:16,941 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 08:36:22,527 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.197e+02 1.731e+02 1.992e+02 2.394e+02 4.370e+02, threshold=3.985e+02, percent-clipped=1.0 +2023-03-21 08:36:25,696 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=103454.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:36:31,134 INFO [train.py:901] (0/2) Epoch 37, batch 1800, loss[loss=0.1259, simple_loss=0.2116, pruned_loss=0.02007, over 7314.00 frames. ], tot_loss[loss=0.131, simple_loss=0.2125, pruned_loss=0.02479, over 1443613.92 frames. ], batch size: 80, lr: 4.45e-03, grad_scale: 8.0 +2023-03-21 08:36:38,112 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 08:36:43,250 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=103488.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:36:53,007 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 08:36:53,582 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=103507.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:36:57,101 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7870, 1.6269, 2.0021, 2.1585, 2.0175, 2.0540, 2.0466, 2.3119], + device='cuda:0'), covar=tensor([0.2711, 0.3203, 0.1834, 0.1293, 0.2608, 0.3273, 0.2331, 0.1836], + device='cuda:0'), in_proj_covar=tensor([0.0077, 0.0078, 0.0070, 0.0063, 0.0061, 0.0062, 0.0104, 0.0066], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 08:36:57,462 INFO [train.py:901] (0/2) Epoch 37, batch 1850, loss[loss=0.1158, simple_loss=0.197, pruned_loss=0.0173, over 7224.00 frames. ], tot_loss[loss=0.1313, simple_loss=0.2127, pruned_loss=0.02494, over 1444512.34 frames. ], batch size: 50, lr: 4.45e-03, grad_scale: 8.0 +2023-03-21 08:37:02,908 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 08:37:11,685 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=103543.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:37:14,075 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.267e+02 1.720e+02 2.032e+02 2.379e+02 3.358e+02, threshold=4.063e+02, percent-clipped=0.0 +2023-03-21 08:37:17,194 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9295, 4.0236, 3.7895, 3.9866, 3.7902, 3.7297, 4.0981, 4.1734], + device='cuda:0'), covar=tensor([0.0335, 0.0231, 0.0336, 0.0337, 0.0460, 0.0418, 0.0437, 0.0377], + device='cuda:0'), in_proj_covar=tensor([0.0120, 0.0125, 0.0118, 0.0122, 0.0113, 0.0102, 0.0098, 0.0100], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:37:17,659 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=103555.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:37:20,671 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 08:37:22,706 INFO [train.py:901] (0/2) Epoch 37, batch 1900, loss[loss=0.1287, simple_loss=0.1983, pruned_loss=0.02961, over 7242.00 frames. ], tot_loss[loss=0.131, simple_loss=0.2122, pruned_loss=0.02495, over 1444248.41 frames. ], batch size: 45, lr: 4.45e-03, grad_scale: 8.0 +2023-03-21 08:37:37,081 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=103591.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:37:40,527 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.87 vs. limit=2.0 +2023-03-21 08:37:47,918 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 08:37:49,381 INFO [train.py:901] (0/2) Epoch 37, batch 1950, loss[loss=0.1342, simple_loss=0.2134, pruned_loss=0.02746, over 7237.00 frames. ], tot_loss[loss=0.1311, simple_loss=0.2122, pruned_loss=0.02504, over 1444186.56 frames. ], batch size: 47, lr: 4.45e-03, grad_scale: 8.0 +2023-03-21 08:37:49,649 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.74 vs. limit=5.0 +2023-03-21 08:37:49,688 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.01 vs. limit=2.0 +2023-03-21 08:37:58,942 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 08:38:03,634 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 08:38:04,158 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 08:38:06,187 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.135e+02 1.676e+02 1.908e+02 2.226e+02 3.747e+02, threshold=3.816e+02, percent-clipped=0.0 +2023-03-21 08:38:14,817 INFO [train.py:901] (0/2) Epoch 37, batch 2000, loss[loss=0.1389, simple_loss=0.2213, pruned_loss=0.0282, over 7325.00 frames. ], tot_loss[loss=0.1315, simple_loss=0.2127, pruned_loss=0.0252, over 1443598.10 frames. ], batch size: 54, lr: 4.45e-03, grad_scale: 8.0 +2023-03-21 08:38:21,985 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 08:38:32,213 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 08:38:40,171 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 08:38:41,159 INFO [train.py:901] (0/2) Epoch 37, batch 2050, loss[loss=0.121, simple_loss=0.212, pruned_loss=0.01496, over 7304.00 frames. ], tot_loss[loss=0.1319, simple_loss=0.213, pruned_loss=0.02538, over 1445150.05 frames. ], batch size: 80, lr: 4.45e-03, grad_scale: 8.0 +2023-03-21 08:38:44,752 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=103722.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 08:38:53,566 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.65 vs. limit=2.0 +2023-03-21 08:38:53,855 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=103740.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:38:58,273 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.125e+02 1.791e+02 2.137e+02 2.523e+02 4.759e+02, threshold=4.274e+02, percent-clipped=4.0 +2023-03-21 08:39:01,463 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=103754.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:39:06,890 INFO [train.py:901] (0/2) Epoch 37, batch 2100, loss[loss=0.1339, simple_loss=0.2215, pruned_loss=0.02315, over 7300.00 frames. ], tot_loss[loss=0.1319, simple_loss=0.213, pruned_loss=0.02537, over 1444182.30 frames. ], batch size: 86, lr: 4.45e-03, grad_scale: 8.0 +2023-03-21 08:39:07,028 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=103765.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:39:08,988 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2416, 4.7734, 4.7903, 4.7513, 4.7069, 4.2778, 4.8314, 4.6727], + device='cuda:0'), covar=tensor([0.0484, 0.0365, 0.0385, 0.0522, 0.0325, 0.0435, 0.0296, 0.0441], + device='cuda:0'), in_proj_covar=tensor([0.0215, 0.0254, 0.0198, 0.0198, 0.0156, 0.0228, 0.0202, 0.0149], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 08:39:11,891 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-03-21 08:39:14,040 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 08:39:16,651 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 08:39:19,260 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=103788.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:39:26,145 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=103801.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:39:26,576 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=103802.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:39:33,031 INFO [train.py:901] (0/2) Epoch 37, batch 2150, loss[loss=0.1389, simple_loss=0.2218, pruned_loss=0.02803, over 7299.00 frames. ], tot_loss[loss=0.1317, simple_loss=0.213, pruned_loss=0.02521, over 1443077.71 frames. ], batch size: 86, lr: 4.45e-03, grad_scale: 8.0 +2023-03-21 08:39:36,034 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8641, 2.9961, 3.8273, 3.8822, 3.9167, 3.9390, 3.8988, 3.7481], + device='cuda:0'), covar=tensor([0.0029, 0.0131, 0.0036, 0.0035, 0.0034, 0.0031, 0.0039, 0.0052], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0069, 0.0057, 0.0056, 0.0054, 0.0060, 0.0048, 0.0076], + device='cuda:0'), out_proj_covar=tensor([8.1948e-05, 1.4146e-04, 1.0420e-04, 9.7600e-05, 9.3736e-05, 1.0483e-04, + 9.2338e-05, 1.4281e-04], device='cuda:0') +2023-03-21 08:39:38,474 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=103826.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:39:43,359 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=103836.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:39:43,491 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9081, 3.7837, 3.6495, 3.6452, 3.2541, 3.3042, 4.0986, 2.6698], + device='cuda:0'), covar=tensor([0.0431, 0.0754, 0.0579, 0.0688, 0.1078, 0.1271, 0.0766, 0.2777], + device='cuda:0'), in_proj_covar=tensor([0.0333, 0.0342, 0.0274, 0.0358, 0.0295, 0.0290, 0.0351, 0.0252], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 08:39:49,830 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.233e+02 1.798e+02 2.078e+02 2.587e+02 4.516e+02, threshold=4.155e+02, percent-clipped=1.0 +2023-03-21 08:39:58,866 INFO [train.py:901] (0/2) Epoch 37, batch 2200, loss[loss=0.1275, simple_loss=0.2045, pruned_loss=0.02531, over 7356.00 frames. ], tot_loss[loss=0.1326, simple_loss=0.2138, pruned_loss=0.02574, over 1444364.97 frames. ], batch size: 54, lr: 4.45e-03, grad_scale: 8.0 +2023-03-21 08:40:02,448 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 08:40:04,628 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=103876.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:40:20,248 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7139, 2.6092, 2.6497, 3.6243, 2.0587, 3.7280, 1.6809, 3.4059], + device='cuda:0'), covar=tensor([0.0193, 0.1408, 0.1778, 0.0189, 0.4010, 0.0308, 0.1191, 0.0437], + device='cuda:0'), in_proj_covar=tensor([0.0205, 0.0248, 0.0265, 0.0209, 0.0251, 0.0215, 0.0230, 0.0228], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:40:24,113 INFO [train.py:901] (0/2) Epoch 37, batch 2250, loss[loss=0.1459, simple_loss=0.234, pruned_loss=0.02886, over 7248.00 frames. ], tot_loss[loss=0.1322, simple_loss=0.2134, pruned_loss=0.02553, over 1444583.30 frames. ], batch size: 93, lr: 4.44e-03, grad_scale: 8.0 +2023-03-21 08:40:36,075 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=103937.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 08:40:36,483 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 08:40:36,982 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 08:40:42,108 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.085e+02 1.782e+02 2.116e+02 2.591e+02 5.653e+02, threshold=4.232e+02, percent-clipped=4.0 +2023-03-21 08:40:49,141 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 08:40:50,621 INFO [train.py:901] (0/2) Epoch 37, batch 2300, loss[loss=0.1274, simple_loss=0.2126, pruned_loss=0.02112, over 7302.00 frames. ], tot_loss[loss=0.1318, simple_loss=0.213, pruned_loss=0.02533, over 1442990.61 frames. ], batch size: 86, lr: 4.44e-03, grad_scale: 8.0 +2023-03-21 08:41:03,315 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=103990.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:41:04,258 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0122, 4.5063, 4.5581, 4.4950, 4.4993, 4.0674, 4.5687, 4.4253], + device='cuda:0'), covar=tensor([0.0510, 0.0408, 0.0387, 0.0549, 0.0348, 0.0440, 0.0329, 0.0509], + device='cuda:0'), in_proj_covar=tensor([0.0216, 0.0255, 0.0200, 0.0201, 0.0157, 0.0229, 0.0205, 0.0151], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 08:41:06,808 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=103997.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:41:08,481 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-104000.pt +2023-03-21 08:41:20,370 INFO [train.py:901] (0/2) Epoch 37, batch 2350, loss[loss=0.1376, simple_loss=0.2199, pruned_loss=0.02767, over 7255.00 frames. ], tot_loss[loss=0.1317, simple_loss=0.213, pruned_loss=0.0252, over 1442640.14 frames. ], batch size: 64, lr: 4.44e-03, grad_scale: 8.0 +2023-03-21 08:41:24,048 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=104022.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 08:41:27,049 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7493, 1.4894, 1.9002, 2.0845, 1.9877, 2.0806, 1.8625, 2.1816], + device='cuda:0'), covar=tensor([0.2237, 0.4867, 0.2380, 0.1504, 0.0992, 0.1500, 0.2240, 0.2478], + device='cuda:0'), in_proj_covar=tensor([0.0075, 0.0076, 0.0069, 0.0061, 0.0061, 0.0059, 0.0102, 0.0064], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 08:41:37,437 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.348e+02 1.773e+02 2.094e+02 2.536e+02 5.250e+02, threshold=4.187e+02, percent-clipped=1.0 +2023-03-21 08:41:38,085 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=104049.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:41:39,137 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=104051.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:41:39,521 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 08:41:42,784 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=104058.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:41:45,584 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 08:41:46,061 INFO [train.py:901] (0/2) Epoch 37, batch 2400, loss[loss=0.1439, simple_loss=0.23, pruned_loss=0.02887, over 6776.00 frames. ], tot_loss[loss=0.1313, simple_loss=0.213, pruned_loss=0.02477, over 1444307.20 frames. ], batch size: 107, lr: 4.44e-03, grad_scale: 8.0 +2023-03-21 08:41:46,749 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8236, 2.6306, 2.7007, 3.5909, 2.0849, 3.7886, 1.7074, 3.3634], + device='cuda:0'), covar=tensor([0.0193, 0.1229, 0.1676, 0.0236, 0.3756, 0.0295, 0.1240, 0.0447], + device='cuda:0'), in_proj_covar=tensor([0.0205, 0.0246, 0.0263, 0.0207, 0.0250, 0.0213, 0.0228, 0.0227], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:41:48,680 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=104070.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 08:41:51,357 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.77 vs. limit=2.0 +2023-03-21 08:41:55,658 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 08:41:59,379 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 08:42:02,522 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=104096.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:42:09,506 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=104110.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:42:12,513 INFO [train.py:901] (0/2) Epoch 37, batch 2450, loss[loss=0.1418, simple_loss=0.2314, pruned_loss=0.02612, over 7320.00 frames. ], tot_loss[loss=0.1317, simple_loss=0.2133, pruned_loss=0.02506, over 1444403.03 frames. ], batch size: 54, lr: 4.44e-03, grad_scale: 8.0 +2023-03-21 08:42:15,196 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2545, 2.9488, 3.4516, 3.1553, 3.2305, 3.1495, 2.9033, 3.2877], + device='cuda:0'), covar=tensor([0.1150, 0.0656, 0.0779, 0.1523, 0.0971, 0.0910, 0.1436, 0.1164], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0068, 0.0051, 0.0050, 0.0050, 0.0049, 0.0068, 0.0051], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 08:42:15,636 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=104121.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:42:24,623 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 08:42:27,442 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.10 vs. limit=5.0 +2023-03-21 08:42:29,121 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.164e+02 1.742e+02 2.006e+02 2.386e+02 4.524e+02, threshold=4.012e+02, percent-clipped=1.0 +2023-03-21 08:42:37,636 INFO [train.py:901] (0/2) Epoch 37, batch 2500, loss[loss=0.1416, simple_loss=0.2212, pruned_loss=0.03094, over 7134.00 frames. ], tot_loss[loss=0.132, simple_loss=0.2136, pruned_loss=0.02525, over 1444129.47 frames. ], batch size: 98, lr: 4.44e-03, grad_scale: 8.0 +2023-03-21 08:42:49,417 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 08:43:04,292 INFO [train.py:901] (0/2) Epoch 37, batch 2550, loss[loss=0.1064, simple_loss=0.183, pruned_loss=0.01486, over 7177.00 frames. ], tot_loss[loss=0.1318, simple_loss=0.213, pruned_loss=0.02529, over 1442344.68 frames. ], batch size: 39, lr: 4.44e-03, grad_scale: 8.0 +2023-03-21 08:43:04,408 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1666, 3.7823, 4.2599, 4.1960, 4.2062, 4.2709, 4.3006, 4.1858], + device='cuda:0'), covar=tensor([0.0033, 0.0085, 0.0029, 0.0031, 0.0029, 0.0028, 0.0026, 0.0044], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0069, 0.0057, 0.0056, 0.0055, 0.0060, 0.0048, 0.0077], + device='cuda:0'), out_proj_covar=tensor([8.2428e-05, 1.4238e-04, 1.0459e-04, 9.7886e-05, 9.3774e-05, 1.0448e-04, + 9.2980e-05, 1.4358e-04], device='cuda:0') +2023-03-21 08:43:08,003 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4727, 4.0067, 4.0175, 4.1812, 4.1420, 4.0341, 4.3778, 3.6562], + device='cuda:0'), covar=tensor([0.0132, 0.0157, 0.0131, 0.0154, 0.0410, 0.0117, 0.0124, 0.0229], + device='cuda:0'), in_proj_covar=tensor([0.0096, 0.0101, 0.0103, 0.0090, 0.0177, 0.0107, 0.0103, 0.0112], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:43:12,856 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=104232.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 08:43:15,110 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.79 vs. limit=2.0 +2023-03-21 08:43:20,887 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.265e+02 1.707e+02 2.054e+02 2.432e+02 3.967e+02, threshold=4.108e+02, percent-clipped=0.0 +2023-03-21 08:43:29,875 INFO [train.py:901] (0/2) Epoch 37, batch 2600, loss[loss=0.128, simple_loss=0.2123, pruned_loss=0.02187, over 7343.00 frames. ], tot_loss[loss=0.132, simple_loss=0.2135, pruned_loss=0.02529, over 1443247.07 frames. ], batch size: 51, lr: 4.44e-03, grad_scale: 8.0 +2023-03-21 08:43:48,519 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9484, 4.5141, 4.2536, 4.9332, 4.7279, 4.8801, 4.1660, 4.5715], + device='cuda:0'), covar=tensor([0.0762, 0.2337, 0.2181, 0.1010, 0.0884, 0.1147, 0.0806, 0.1046], + device='cuda:0'), in_proj_covar=tensor([0.0256, 0.0383, 0.0293, 0.0307, 0.0224, 0.0365, 0.0224, 0.0272], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:43:54,445 INFO [train.py:901] (0/2) Epoch 37, batch 2650, loss[loss=0.1231, simple_loss=0.2028, pruned_loss=0.02169, over 7344.00 frames. ], tot_loss[loss=0.1312, simple_loss=0.2125, pruned_loss=0.02499, over 1442090.60 frames. ], batch size: 73, lr: 4.44e-03, grad_scale: 8.0 +2023-03-21 08:44:09,763 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=104346.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:44:10,637 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.220e+02 1.763e+02 1.995e+02 2.482e+02 3.985e+02, threshold=3.989e+02, percent-clipped=0.0 +2023-03-21 08:44:12,286 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9864, 3.8644, 3.1531, 3.5481, 2.9065, 2.2067, 1.7465, 4.0692], + device='cuda:0'), covar=tensor([0.0053, 0.0066, 0.0146, 0.0080, 0.0176, 0.0604, 0.0679, 0.0049], + device='cuda:0'), in_proj_covar=tensor([0.0114, 0.0094, 0.0114, 0.0094, 0.0129, 0.0136, 0.0131, 0.0104], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 08:44:13,171 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=104353.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:44:19,674 INFO [train.py:901] (0/2) Epoch 37, batch 2700, loss[loss=0.132, simple_loss=0.2182, pruned_loss=0.02291, over 7310.00 frames. ], tot_loss[loss=0.1312, simple_loss=0.2123, pruned_loss=0.02505, over 1440870.23 frames. ], batch size: 83, lr: 4.43e-03, grad_scale: 8.0 +2023-03-21 08:44:19,807 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5583, 2.9122, 2.5459, 2.8611, 2.8062, 2.5784, 2.9218, 2.7466], + device='cuda:0'), covar=tensor([0.0779, 0.0563, 0.0888, 0.0991, 0.1001, 0.0861, 0.0643, 0.0731], + device='cuda:0'), in_proj_covar=tensor([0.0057, 0.0057, 0.0066, 0.0058, 0.0056, 0.0061, 0.0055, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:44:34,286 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3380, 2.5666, 2.3680, 2.5611, 2.5149, 2.2105, 2.5733, 2.4258], + device='cuda:0'), covar=tensor([0.0798, 0.0737, 0.1017, 0.0834, 0.0851, 0.0862, 0.0839, 0.0890], + device='cuda:0'), in_proj_covar=tensor([0.0057, 0.0057, 0.0066, 0.0058, 0.0056, 0.0061, 0.0055, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:44:35,250 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=104396.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:44:39,934 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=104405.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:44:44,862 INFO [train.py:901] (0/2) Epoch 37, batch 2750, loss[loss=0.1215, simple_loss=0.2092, pruned_loss=0.01688, over 7322.00 frames. ], tot_loss[loss=0.1314, simple_loss=0.2126, pruned_loss=0.02512, over 1442209.09 frames. ], batch size: 61, lr: 4.43e-03, grad_scale: 8.0 +2023-03-21 08:44:47,900 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=104421.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:44:55,339 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3634, 3.0102, 3.2492, 3.3316, 2.8891, 2.9274, 3.1252, 2.4836], + device='cuda:0'), covar=tensor([0.0523, 0.0504, 0.0665, 0.0606, 0.0632, 0.0918, 0.0714, 0.2180], + device='cuda:0'), in_proj_covar=tensor([0.0329, 0.0338, 0.0273, 0.0355, 0.0291, 0.0287, 0.0347, 0.0249], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 08:44:59,164 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=104444.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:45:01,044 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.078e+02 1.660e+02 2.023e+02 2.410e+02 4.270e+02, threshold=4.046e+02, percent-clipped=1.0 +2023-03-21 08:45:09,296 INFO [train.py:901] (0/2) Epoch 37, batch 2800, loss[loss=0.1359, simple_loss=0.2194, pruned_loss=0.02619, over 7320.00 frames. ], tot_loss[loss=0.132, simple_loss=0.2131, pruned_loss=0.02544, over 1443049.26 frames. ], batch size: 80, lr: 4.43e-03, grad_scale: 8.0 +2023-03-21 08:45:11,301 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=104469.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:45:21,798 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-37.pt +2023-03-21 08:45:37,477 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 08:45:37,531 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 08:45:37,556 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 08:45:37,574 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 08:45:37,751 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 08:45:37,767 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 08:45:37,778 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 08:45:37,877 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 08:45:37,953 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 08:45:37,967 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 08:45:37,984 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 08:45:38,497 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 08:45:38,503 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 08:45:38,511 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 08:45:38,652 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 08:45:38,655 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 08:45:38,714 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 08:45:40,897 INFO [train.py:901] (0/2) Epoch 38, batch 0, loss[loss=0.1285, simple_loss=0.2103, pruned_loss=0.02331, over 7345.00 frames. ], tot_loss[loss=0.1285, simple_loss=0.2103, pruned_loss=0.02331, over 7345.00 frames. ], batch size: 51, lr: 4.37e-03, grad_scale: 8.0 +2023-03-21 08:45:40,898 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 08:46:04,454 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7696, 1.5921, 1.9744, 2.0496, 1.9476, 2.1754, 1.8174, 2.1881], + device='cuda:0'), covar=tensor([0.2276, 0.3410, 0.1356, 0.0761, 0.0844, 0.1106, 0.1704, 0.2103], + device='cuda:0'), in_proj_covar=tensor([0.0076, 0.0077, 0.0070, 0.0062, 0.0062, 0.0061, 0.0102, 0.0065], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 08:46:07,745 INFO [train.py:935] (0/2) Epoch 38, validation: loss=0.1635, simple_loss=0.256, pruned_loss=0.03545, over 1622729.00 frames. +2023-03-21 08:46:07,745 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 08:46:14,057 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.41 vs. limit=2.0 +2023-03-21 08:46:15,264 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 08:46:25,935 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 08:46:29,992 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=104532.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 08:46:32,933 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 08:46:33,438 INFO [train.py:901] (0/2) Epoch 38, batch 50, loss[loss=0.1318, simple_loss=0.2101, pruned_loss=0.0267, over 7265.00 frames. ], tot_loss[loss=0.1313, simple_loss=0.2131, pruned_loss=0.02476, over 324542.55 frames. ], batch size: 52, lr: 4.37e-03, grad_scale: 8.0 +2023-03-21 08:46:34,429 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 08:46:36,881 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 08:46:37,842 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.288e+02 1.780e+02 2.077e+02 2.791e+02 4.697e+02, threshold=4.154e+02, percent-clipped=2.0 +2023-03-21 08:46:53,964 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 08:46:54,467 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 08:46:54,513 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=104580.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:46:58,825 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.86 vs. limit=2.0 +2023-03-21 08:46:59,015 INFO [train.py:901] (0/2) Epoch 38, batch 100, loss[loss=0.132, simple_loss=0.2173, pruned_loss=0.02339, over 7269.00 frames. ], tot_loss[loss=0.1314, simple_loss=0.2129, pruned_loss=0.02496, over 573495.57 frames. ], batch size: 70, lr: 4.37e-03, grad_scale: 8.0 +2023-03-21 08:47:07,947 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9138, 2.4570, 3.0521, 2.9057, 3.1331, 2.8578, 2.7060, 2.9868], + device='cuda:0'), covar=tensor([0.1311, 0.0888, 0.1074, 0.1254, 0.0736, 0.1028, 0.1473, 0.1341], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0068, 0.0051, 0.0050, 0.0050, 0.0050, 0.0068, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 08:47:11,914 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6130, 2.9988, 2.6848, 2.8061, 2.9464, 2.6874, 2.9920, 2.9306], + device='cuda:0'), covar=tensor([0.1041, 0.0667, 0.0810, 0.1096, 0.0762, 0.0645, 0.0640, 0.0615], + device='cuda:0'), in_proj_covar=tensor([0.0057, 0.0057, 0.0065, 0.0059, 0.0055, 0.0060, 0.0055, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:47:24,970 INFO [train.py:901] (0/2) Epoch 38, batch 150, loss[loss=0.1493, simple_loss=0.2242, pruned_loss=0.03719, over 7224.00 frames. ], tot_loss[loss=0.1306, simple_loss=0.2122, pruned_loss=0.02451, over 763237.60 frames. ], batch size: 45, lr: 4.37e-03, grad_scale: 8.0 +2023-03-21 08:47:25,602 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2323, 2.7255, 1.9546, 2.9920, 2.8443, 3.1394, 2.7345, 2.9177], + device='cuda:0'), covar=tensor([0.2082, 0.0988, 0.3810, 0.0783, 0.0290, 0.0300, 0.0420, 0.0430], + device='cuda:0'), in_proj_covar=tensor([0.0254, 0.0232, 0.0252, 0.0258, 0.0198, 0.0194, 0.0217, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:47:28,525 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=104646.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:47:29,416 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.300e+02 1.726e+02 2.045e+02 2.303e+02 3.131e+02, threshold=4.089e+02, percent-clipped=0.0 +2023-03-21 08:47:32,764 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=104653.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:47:43,212 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9213, 3.3197, 3.9621, 3.8290, 4.0328, 4.0509, 4.0472, 3.8740], + device='cuda:0'), covar=tensor([0.0031, 0.0109, 0.0029, 0.0033, 0.0030, 0.0025, 0.0033, 0.0045], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0069, 0.0057, 0.0055, 0.0054, 0.0059, 0.0048, 0.0076], + device='cuda:0'), out_proj_covar=tensor([8.1909e-05, 1.4042e-04, 1.0405e-04, 9.6386e-05, 9.2938e-05, 1.0343e-04, + 9.1037e-05, 1.4245e-04], device='cuda:0') +2023-03-21 08:47:50,589 INFO [train.py:901] (0/2) Epoch 38, batch 200, loss[loss=0.1251, simple_loss=0.2117, pruned_loss=0.01926, over 7263.00 frames. ], tot_loss[loss=0.1308, simple_loss=0.2122, pruned_loss=0.02473, over 914253.19 frames. ], batch size: 64, lr: 4.37e-03, grad_scale: 8.0 +2023-03-21 08:47:53,707 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=104694.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:47:56,195 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 08:47:57,247 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=104701.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:47:59,281 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=104705.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:48:00,877 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8957, 3.3316, 3.6583, 3.6517, 3.1824, 3.1390, 3.7839, 2.7321], + device='cuda:0'), covar=tensor([0.0486, 0.0506, 0.0605, 0.0698, 0.0865, 0.1052, 0.0804, 0.2221], + device='cuda:0'), in_proj_covar=tensor([0.0333, 0.0340, 0.0274, 0.0357, 0.0294, 0.0290, 0.0349, 0.0252], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 08:48:01,214 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 08:48:04,385 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2944, 3.6123, 2.4485, 3.7516, 2.8724, 3.3929, 1.7613, 2.3750], + device='cuda:0'), covar=tensor([0.0460, 0.0806, 0.2846, 0.0542, 0.0590, 0.0623, 0.3857, 0.2148], + device='cuda:0'), in_proj_covar=tensor([0.0279, 0.0257, 0.0282, 0.0269, 0.0272, 0.0267, 0.0237, 0.0260], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:48:08,239 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 08:48:16,328 INFO [train.py:901] (0/2) Epoch 38, batch 250, loss[loss=0.1123, simple_loss=0.1898, pruned_loss=0.01738, over 7109.00 frames. ], tot_loss[loss=0.1292, simple_loss=0.21, pruned_loss=0.02421, over 1029181.06 frames. ], batch size: 41, lr: 4.37e-03, grad_scale: 8.0 +2023-03-21 08:48:21,306 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.231e+02 1.611e+02 1.949e+02 2.401e+02 6.980e+02, threshold=3.898e+02, percent-clipped=1.0 +2023-03-21 08:48:21,360 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 08:48:23,898 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=104753.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:48:41,913 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 08:48:42,357 INFO [train.py:901] (0/2) Epoch 38, batch 300, loss[loss=0.129, simple_loss=0.2171, pruned_loss=0.0204, over 7338.00 frames. ], tot_loss[loss=0.1296, simple_loss=0.2106, pruned_loss=0.02424, over 1123899.61 frames. ], batch size: 75, lr: 4.37e-03, grad_scale: 8.0 +2023-03-21 08:48:50,637 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 08:49:07,918 INFO [train.py:901] (0/2) Epoch 38, batch 350, loss[loss=0.1257, simple_loss=0.2042, pruned_loss=0.02366, over 7328.00 frames. ], tot_loss[loss=0.1301, simple_loss=0.2112, pruned_loss=0.02448, over 1195574.27 frames. ], batch size: 44, lr: 4.37e-03, grad_scale: 8.0 +2023-03-21 08:49:12,392 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.194e+02 1.765e+02 1.993e+02 2.346e+02 3.883e+02, threshold=3.986e+02, percent-clipped=0.0 +2023-03-21 08:49:25,514 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 08:49:33,600 INFO [train.py:901] (0/2) Epoch 38, batch 400, loss[loss=0.1378, simple_loss=0.2152, pruned_loss=0.03018, over 7361.00 frames. ], tot_loss[loss=0.1303, simple_loss=0.2116, pruned_loss=0.02448, over 1251588.55 frames. ], batch size: 73, lr: 4.37e-03, grad_scale: 8.0 +2023-03-21 08:49:59,219 INFO [train.py:901] (0/2) Epoch 38, batch 450, loss[loss=0.1149, simple_loss=0.1944, pruned_loss=0.01773, over 7328.00 frames. ], tot_loss[loss=0.1306, simple_loss=0.2118, pruned_loss=0.02469, over 1293887.82 frames. ], batch size: 59, lr: 4.36e-03, grad_scale: 8.0 +2023-03-21 08:50:03,665 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.059e+02 1.744e+02 1.979e+02 2.437e+02 3.979e+02, threshold=3.958e+02, percent-clipped=0.0 +2023-03-21 08:50:03,822 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9333, 2.5452, 3.2518, 2.9442, 3.3019, 2.9569, 2.6400, 3.1792], + device='cuda:0'), covar=tensor([0.1503, 0.0935, 0.0850, 0.1742, 0.0558, 0.0981, 0.1654, 0.1207], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0068, 0.0050, 0.0050, 0.0050, 0.0049, 0.0068, 0.0051], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 08:50:05,679 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 08:50:06,200 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 08:50:24,968 INFO [train.py:901] (0/2) Epoch 38, batch 500, loss[loss=0.1406, simple_loss=0.2198, pruned_loss=0.03066, over 7271.00 frames. ], tot_loss[loss=0.1313, simple_loss=0.2126, pruned_loss=0.02497, over 1327034.40 frames. ], batch size: 77, lr: 4.36e-03, grad_scale: 8.0 +2023-03-21 08:50:38,898 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 08:50:39,883 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 08:50:40,924 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 08:50:42,948 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 08:50:47,486 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 08:50:51,535 INFO [train.py:901] (0/2) Epoch 38, batch 550, loss[loss=0.1449, simple_loss=0.2237, pruned_loss=0.03301, over 7316.00 frames. ], tot_loss[loss=0.1308, simple_loss=0.2122, pruned_loss=0.02466, over 1352755.42 frames. ], batch size: 59, lr: 4.36e-03, grad_scale: 8.0 +2023-03-21 08:50:54,726 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4794, 1.7113, 1.5125, 1.5921, 1.8000, 1.6199, 1.6651, 1.2436], + device='cuda:0'), covar=tensor([0.0177, 0.0157, 0.0330, 0.0166, 0.0144, 0.0148, 0.0142, 0.0169], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0035, 0.0034, 0.0035, 0.0034, 0.0032, 0.0036, 0.0043], + device='cuda:0'), out_proj_covar=tensor([4.1452e-05, 3.8587e-05, 3.8487e-05, 3.8973e-05, 3.7315e-05, 3.6146e-05, + 4.0484e-05, 4.7674e-05], device='cuda:0') +2023-03-21 08:50:56,111 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.231e+02 1.699e+02 2.007e+02 2.448e+02 5.242e+02, threshold=4.013e+02, percent-clipped=3.0 +2023-03-21 08:50:58,206 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6447, 5.2160, 5.3628, 5.2376, 5.0199, 4.6556, 5.3251, 5.0859], + device='cuda:0'), covar=tensor([0.0448, 0.0364, 0.0303, 0.0454, 0.0402, 0.0392, 0.0332, 0.0578], + device='cuda:0'), in_proj_covar=tensor([0.0217, 0.0256, 0.0200, 0.0200, 0.0160, 0.0230, 0.0208, 0.0153], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 08:50:59,107 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 08:51:07,112 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 08:51:10,686 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 08:51:13,200 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7986, 4.3742, 4.2147, 4.7809, 4.6293, 4.7163, 4.1245, 4.3852], + device='cuda:0'), covar=tensor([0.0849, 0.2341, 0.2237, 0.0970, 0.0859, 0.1148, 0.0869, 0.1158], + device='cuda:0'), in_proj_covar=tensor([0.0258, 0.0386, 0.0297, 0.0310, 0.0225, 0.0365, 0.0226, 0.0273], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:51:17,242 INFO [train.py:901] (0/2) Epoch 38, batch 600, loss[loss=0.1418, simple_loss=0.2229, pruned_loss=0.03035, over 7327.00 frames. ], tot_loss[loss=0.1304, simple_loss=0.212, pruned_loss=0.02441, over 1373706.52 frames. ], batch size: 54, lr: 4.36e-03, grad_scale: 8.0 +2023-03-21 08:51:17,732 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 08:51:19,030 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.44 vs. limit=5.0 +2023-03-21 08:51:33,255 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 08:51:34,354 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3020, 2.5631, 2.2547, 2.3157, 2.4876, 2.1425, 2.5342, 2.4314], + device='cuda:0'), covar=tensor([0.0911, 0.0555, 0.0988, 0.1313, 0.0720, 0.0723, 0.0521, 0.0912], + device='cuda:0'), in_proj_covar=tensor([0.0057, 0.0058, 0.0067, 0.0059, 0.0056, 0.0061, 0.0056, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:51:42,768 INFO [train.py:901] (0/2) Epoch 38, batch 650, loss[loss=0.1221, simple_loss=0.1923, pruned_loss=0.02597, over 7129.00 frames. ], tot_loss[loss=0.1304, simple_loss=0.2117, pruned_loss=0.02458, over 1386810.53 frames. ], batch size: 41, lr: 4.36e-03, grad_scale: 8.0 +2023-03-21 08:51:42,788 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 08:51:43,952 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4308, 1.8129, 1.5993, 1.5339, 1.8804, 1.6670, 1.6150, 1.2865], + device='cuda:0'), covar=tensor([0.0204, 0.0143, 0.0309, 0.0210, 0.0129, 0.0112, 0.0195, 0.0167], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0035, 0.0034, 0.0035, 0.0034, 0.0032, 0.0036, 0.0043], + device='cuda:0'), out_proj_covar=tensor([4.1461e-05, 3.8460e-05, 3.8431e-05, 3.9277e-05, 3.7377e-05, 3.6079e-05, + 4.0605e-05, 4.7697e-05], device='cuda:0') +2023-03-21 08:51:47,277 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.200e+02 1.709e+02 2.034e+02 2.305e+02 4.227e+02, threshold=4.069e+02, percent-clipped=1.0 +2023-03-21 08:51:49,493 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5090, 1.7822, 1.5814, 1.5759, 1.8229, 1.6352, 1.5987, 1.2882], + device='cuda:0'), covar=tensor([0.0155, 0.0140, 0.0257, 0.0175, 0.0141, 0.0121, 0.0137, 0.0175], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0035, 0.0034, 0.0035, 0.0033, 0.0032, 0.0036, 0.0043], + device='cuda:0'), out_proj_covar=tensor([4.1306e-05, 3.8341e-05, 3.8247e-05, 3.9179e-05, 3.7266e-05, 3.5985e-05, + 4.0512e-05, 4.7534e-05], device='cuda:0') +2023-03-21 08:51:56,142 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.84 vs. limit=2.0 +2023-03-21 08:51:59,355 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 08:52:08,286 INFO [train.py:901] (0/2) Epoch 38, batch 700, loss[loss=0.1179, simple_loss=0.1973, pruned_loss=0.01921, over 7350.00 frames. ], tot_loss[loss=0.1302, simple_loss=0.2114, pruned_loss=0.02446, over 1399182.56 frames. ], batch size: 51, lr: 4.36e-03, grad_scale: 8.0 +2023-03-21 08:52:08,825 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 08:52:18,247 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3822, 3.4924, 3.4096, 3.4596, 3.2936, 3.3756, 3.7030, 3.7004], + device='cuda:0'), covar=tensor([0.0251, 0.0194, 0.0231, 0.0204, 0.0351, 0.0527, 0.0243, 0.0212], + device='cuda:0'), in_proj_covar=tensor([0.0121, 0.0126, 0.0118, 0.0122, 0.0114, 0.0103, 0.0098, 0.0099], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:52:31,356 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5867, 5.0451, 5.1025, 5.0728, 4.8714, 4.5485, 5.1095, 4.9561], + device='cuda:0'), covar=tensor([0.0518, 0.0396, 0.0393, 0.0509, 0.0372, 0.0411, 0.0382, 0.0453], + device='cuda:0'), in_proj_covar=tensor([0.0216, 0.0256, 0.0200, 0.0200, 0.0159, 0.0230, 0.0208, 0.0152], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 08:52:33,713 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 08:52:33,732 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 08:52:34,228 INFO [train.py:901] (0/2) Epoch 38, batch 750, loss[loss=0.1342, simple_loss=0.2204, pruned_loss=0.02405, over 7292.00 frames. ], tot_loss[loss=0.1307, simple_loss=0.2116, pruned_loss=0.02487, over 1407818.44 frames. ], batch size: 86, lr: 4.36e-03, grad_scale: 8.0 +2023-03-21 08:52:38,657 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.214e+02 1.746e+02 1.996e+02 2.385e+02 4.034e+02, threshold=3.992e+02, percent-clipped=0.0 +2023-03-21 08:52:48,834 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 08:52:52,807 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 08:52:59,306 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 08:52:59,740 INFO [train.py:901] (0/2) Epoch 38, batch 800, loss[loss=0.1569, simple_loss=0.2338, pruned_loss=0.04, over 6705.00 frames. ], tot_loss[loss=0.1306, simple_loss=0.2114, pruned_loss=0.02483, over 1414891.30 frames. ], batch size: 106, lr: 4.36e-03, grad_scale: 8.0 +2023-03-21 08:53:00,751 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 08:53:11,785 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 08:53:13,007 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1951, 2.2081, 2.3394, 3.3808, 1.7506, 3.4257, 1.4205, 3.2040], + device='cuda:0'), covar=tensor([0.0220, 0.1360, 0.1637, 0.0206, 0.3874, 0.0260, 0.1184, 0.0478], + device='cuda:0'), in_proj_covar=tensor([0.0205, 0.0245, 0.0256, 0.0205, 0.0248, 0.0212, 0.0225, 0.0225], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:53:25,513 INFO [train.py:901] (0/2) Epoch 38, batch 850, loss[loss=0.1352, simple_loss=0.213, pruned_loss=0.02872, over 7314.00 frames. ], tot_loss[loss=0.1306, simple_loss=0.2114, pruned_loss=0.0249, over 1421394.40 frames. ], batch size: 49, lr: 4.36e-03, grad_scale: 8.0 +2023-03-21 08:53:30,013 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.181e+02 1.670e+02 2.029e+02 2.408e+02 3.853e+02, threshold=4.058e+02, percent-clipped=0.0 +2023-03-21 08:53:31,041 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 08:53:31,050 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 08:53:36,615 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.66 vs. limit=2.0 +2023-03-21 08:53:37,247 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 08:53:40,780 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 08:53:41,425 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5144, 1.8678, 1.5973, 1.5939, 1.8682, 1.7366, 1.6745, 1.2767], + device='cuda:0'), covar=tensor([0.0139, 0.0178, 0.0248, 0.0266, 0.0130, 0.0148, 0.0242, 0.0173], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0035, 0.0034, 0.0035, 0.0034, 0.0032, 0.0037, 0.0043], + device='cuda:0'), out_proj_covar=tensor([4.1614e-05, 3.8548e-05, 3.8586e-05, 3.9300e-05, 3.7561e-05, 3.6320e-05, + 4.1187e-05, 4.8066e-05], device='cuda:0') +2023-03-21 08:53:50,857 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=105388.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:53:51,203 INFO [train.py:901] (0/2) Epoch 38, batch 900, loss[loss=0.1234, simple_loss=0.2056, pruned_loss=0.0206, over 7316.00 frames. ], tot_loss[loss=0.1305, simple_loss=0.2113, pruned_loss=0.02482, over 1426207.24 frames. ], batch size: 80, lr: 4.35e-03, grad_scale: 8.0 +2023-03-21 08:53:54,160 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.58 vs. limit=2.0 +2023-03-21 08:54:06,459 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-03-21 08:54:17,255 INFO [train.py:901] (0/2) Epoch 38, batch 950, loss[loss=0.1129, simple_loss=0.1885, pruned_loss=0.01865, over 7338.00 frames. ], tot_loss[loss=0.1298, simple_loss=0.2107, pruned_loss=0.02446, over 1429578.93 frames. ], batch size: 44, lr: 4.35e-03, grad_scale: 16.0 +2023-03-21 08:54:18,270 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 08:54:22,332 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.229e+02 1.656e+02 1.886e+02 2.311e+02 4.194e+02, threshold=3.772e+02, percent-clipped=1.0 +2023-03-21 08:54:23,011 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=105449.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 08:54:36,238 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4308, 4.1943, 4.1594, 4.2508, 3.6386, 4.0595, 4.4233, 4.0212], + device='cuda:0'), covar=tensor([0.0323, 0.0186, 0.0161, 0.0227, 0.0869, 0.0195, 0.0224, 0.0209], + device='cuda:0'), in_proj_covar=tensor([0.0095, 0.0102, 0.0103, 0.0089, 0.0177, 0.0107, 0.0104, 0.0112], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 08:54:37,358 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5201, 2.6037, 2.5535, 3.5051, 1.8142, 3.7042, 1.4515, 3.3257], + device='cuda:0'), covar=tensor([0.0223, 0.1331, 0.1646, 0.0205, 0.4187, 0.0294, 0.1321, 0.0365], + device='cuda:0'), in_proj_covar=tensor([0.0206, 0.0246, 0.0257, 0.0205, 0.0250, 0.0214, 0.0227, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:54:38,879 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1002, 2.8945, 2.8816, 3.0717, 2.7115, 2.6881, 3.2212, 2.1781], + device='cuda:0'), covar=tensor([0.0607, 0.0650, 0.0731, 0.0729, 0.0881, 0.1233, 0.0737, 0.2542], + device='cuda:0'), in_proj_covar=tensor([0.0326, 0.0334, 0.0269, 0.0350, 0.0287, 0.0283, 0.0342, 0.0246], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 08:54:41,208 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 08:54:42,296 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=105486.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:54:43,671 INFO [train.py:901] (0/2) Epoch 38, batch 1000, loss[loss=0.1302, simple_loss=0.2123, pruned_loss=0.02402, over 7261.00 frames. ], tot_loss[loss=0.13, simple_loss=0.2113, pruned_loss=0.02432, over 1433030.23 frames. ], batch size: 47, lr: 4.35e-03, grad_scale: 16.0 +2023-03-21 08:54:44,834 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8705, 2.3749, 3.0520, 2.9194, 3.0384, 2.7279, 2.5378, 2.8549], + device='cuda:0'), covar=tensor([0.1489, 0.0911, 0.0863, 0.1569, 0.0792, 0.1532, 0.1844, 0.1728], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0069, 0.0051, 0.0051, 0.0051, 0.0049, 0.0069, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 08:54:50,306 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=105502.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:55:00,874 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 08:55:03,968 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4927, 1.8331, 1.6403, 1.6198, 1.8821, 1.6658, 1.6818, 1.2114], + device='cuda:0'), covar=tensor([0.0167, 0.0121, 0.0177, 0.0177, 0.0110, 0.0192, 0.0149, 0.0178], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0035, 0.0035, 0.0036, 0.0034, 0.0033, 0.0038, 0.0044], + device='cuda:0'), out_proj_covar=tensor([4.2456e-05, 3.9154e-05, 3.9505e-05, 3.9824e-05, 3.8300e-05, 3.7029e-05, + 4.2237e-05, 4.9009e-05], device='cuda:0') +2023-03-21 08:55:09,299 INFO [train.py:901] (0/2) Epoch 38, batch 1050, loss[loss=0.1324, simple_loss=0.2136, pruned_loss=0.02557, over 7305.00 frames. ], tot_loss[loss=0.1299, simple_loss=0.2116, pruned_loss=0.02415, over 1435438.35 frames. ], batch size: 59, lr: 4.35e-03, grad_scale: 16.0 +2023-03-21 08:55:13,467 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=105547.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:55:13,793 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.256e+02 1.762e+02 2.024e+02 2.467e+02 4.563e+02, threshold=4.048e+02, percent-clipped=5.0 +2023-03-21 08:55:20,573 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.89 vs. limit=2.0 +2023-03-21 08:55:21,798 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 08:55:21,911 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=105563.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:55:25,899 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 08:55:34,884 INFO [train.py:901] (0/2) Epoch 38, batch 1100, loss[loss=0.1272, simple_loss=0.2119, pruned_loss=0.0212, over 7266.00 frames. ], tot_loss[loss=0.13, simple_loss=0.2115, pruned_loss=0.02428, over 1437261.79 frames. ], batch size: 64, lr: 4.35e-03, grad_scale: 16.0 +2023-03-21 08:55:40,148 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9071, 3.1498, 3.9647, 3.8171, 4.0211, 3.8923, 3.9934, 3.5095], + device='cuda:0'), covar=tensor([0.0054, 0.0202, 0.0044, 0.0059, 0.0045, 0.0051, 0.0051, 0.0092], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0068, 0.0056, 0.0054, 0.0053, 0.0059, 0.0047, 0.0075], + device='cuda:0'), out_proj_covar=tensor([8.0423e-05, 1.3904e-04, 1.0197e-04, 9.4261e-05, 9.1490e-05, 1.0342e-04, + 8.8974e-05, 1.4011e-04], device='cuda:0') +2023-03-21 08:55:41,161 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-03-21 08:55:56,075 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 08:55:56,091 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 08:55:57,484 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.83 vs. limit=2.0 +2023-03-21 08:56:01,019 INFO [train.py:901] (0/2) Epoch 38, batch 1150, loss[loss=0.1283, simple_loss=0.2098, pruned_loss=0.02341, over 7277.00 frames. ], tot_loss[loss=0.1307, simple_loss=0.2121, pruned_loss=0.0247, over 1440389.29 frames. ], batch size: 57, lr: 4.35e-03, grad_scale: 16.0 +2023-03-21 08:56:05,511 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.238e+02 1.823e+02 2.073e+02 2.521e+02 5.443e+02, threshold=4.146e+02, percent-clipped=2.0 +2023-03-21 08:56:09,161 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 08:56:09,671 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 08:56:26,786 INFO [train.py:901] (0/2) Epoch 38, batch 1200, loss[loss=0.09576, simple_loss=0.1686, pruned_loss=0.01148, over 7035.00 frames. ], tot_loss[loss=0.1302, simple_loss=0.2112, pruned_loss=0.02464, over 1437357.67 frames. ], batch size: 35, lr: 4.35e-03, grad_scale: 16.0 +2023-03-21 08:56:43,643 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 08:56:47,479 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.68 vs. limit=2.0 +2023-03-21 08:56:53,213 INFO [train.py:901] (0/2) Epoch 38, batch 1250, loss[loss=0.1475, simple_loss=0.2328, pruned_loss=0.03107, over 7276.00 frames. ], tot_loss[loss=0.131, simple_loss=0.2121, pruned_loss=0.02496, over 1438593.67 frames. ], batch size: 70, lr: 4.35e-03, grad_scale: 8.0 +2023-03-21 08:56:55,783 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=105744.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 08:56:58,136 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.196e+02 1.762e+02 2.046e+02 2.416e+02 3.370e+02, threshold=4.092e+02, percent-clipped=0.0 +2023-03-21 08:57:06,238 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5608, 2.7836, 3.5356, 3.5722, 3.5923, 3.6576, 3.4594, 3.4426], + device='cuda:0'), covar=tensor([0.0027, 0.0141, 0.0035, 0.0032, 0.0032, 0.0028, 0.0065, 0.0057], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0069, 0.0057, 0.0055, 0.0054, 0.0060, 0.0048, 0.0076], + device='cuda:0'), out_proj_covar=tensor([8.1612e-05, 1.4069e-04, 1.0366e-04, 9.5777e-05, 9.3427e-05, 1.0471e-04, + 9.0654e-05, 1.4261e-04], device='cuda:0') +2023-03-21 08:57:06,636 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 08:57:11,136 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 08:57:12,639 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 08:57:18,884 INFO [train.py:901] (0/2) Epoch 38, batch 1300, loss[loss=0.1182, simple_loss=0.2039, pruned_loss=0.0162, over 7155.00 frames. ], tot_loss[loss=0.1308, simple_loss=0.2119, pruned_loss=0.02487, over 1437943.22 frames. ], batch size: 41, lr: 4.35e-03, grad_scale: 8.0 +2023-03-21 08:57:20,478 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=105792.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:57:24,009 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=105799.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:57:26,881 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9652, 2.6380, 2.8546, 3.1678, 2.5463, 2.6506, 3.0186, 2.2700], + device='cuda:0'), covar=tensor([0.0560, 0.0553, 0.0780, 0.0665, 0.0727, 0.0937, 0.0709, 0.2336], + device='cuda:0'), in_proj_covar=tensor([0.0328, 0.0335, 0.0271, 0.0353, 0.0290, 0.0284, 0.0346, 0.0248], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 08:57:36,940 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 08:57:38,919 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 08:57:43,018 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 08:57:44,742 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.34 vs. limit=5.0 +2023-03-21 08:57:44,965 INFO [train.py:901] (0/2) Epoch 38, batch 1350, loss[loss=0.1227, simple_loss=0.2016, pruned_loss=0.02187, over 7328.00 frames. ], tot_loss[loss=0.1304, simple_loss=0.2115, pruned_loss=0.02469, over 1438486.13 frames. ], batch size: 42, lr: 4.35e-03, grad_scale: 8.0 +2023-03-21 08:57:46,550 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=105842.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:57:49,967 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.223e+02 1.666e+02 2.004e+02 2.449e+02 3.831e+02, threshold=4.008e+02, percent-clipped=0.0 +2023-03-21 08:57:52,077 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=105853.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:57:52,974 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 08:57:54,532 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=105858.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:57:55,639 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=105860.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:58:00,656 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2592, 2.4480, 2.5307, 2.1796, 2.3106, 2.3661, 2.0273, 1.8528], + device='cuda:0'), covar=tensor([0.0561, 0.0372, 0.0271, 0.0271, 0.0520, 0.0297, 0.0321, 0.0334], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0038, 0.0038, 0.0038, 0.0036, 0.0037, 0.0040, 0.0040], + device='cuda:0'), out_proj_covar=tensor([9.7845e-05, 9.8272e-05, 9.6316e-05, 9.6045e-05, 9.4063e-05, 9.4081e-05, + 1.0067e-04, 1.0141e-04], device='cuda:0') +2023-03-21 08:58:02,670 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1164, 2.4309, 2.4073, 3.3622, 1.8115, 3.5005, 1.4476, 3.2236], + device='cuda:0'), covar=tensor([0.0231, 0.1421, 0.1797, 0.0212, 0.4195, 0.0381, 0.1312, 0.0425], + device='cuda:0'), in_proj_covar=tensor([0.0206, 0.0247, 0.0259, 0.0206, 0.0249, 0.0216, 0.0227, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:58:10,537 INFO [train.py:901] (0/2) Epoch 38, batch 1400, loss[loss=0.1468, simple_loss=0.2246, pruned_loss=0.03449, over 7271.00 frames. ], tot_loss[loss=0.1308, simple_loss=0.212, pruned_loss=0.02484, over 1441665.41 frames. ], batch size: 57, lr: 4.34e-03, grad_scale: 8.0 +2023-03-21 08:58:21,035 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.55 vs. limit=2.0 +2023-03-21 08:58:25,192 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 08:58:35,974 INFO [train.py:901] (0/2) Epoch 38, batch 1450, loss[loss=0.1327, simple_loss=0.2162, pruned_loss=0.02456, over 7286.00 frames. ], tot_loss[loss=0.1321, simple_loss=0.213, pruned_loss=0.02559, over 1440419.55 frames. ], batch size: 77, lr: 4.34e-03, grad_scale: 8.0 +2023-03-21 08:58:41,016 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.294e+02 1.750e+02 2.032e+02 2.346e+02 3.706e+02, threshold=4.064e+02, percent-clipped=0.0 +2023-03-21 08:58:43,789 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.44 vs. limit=5.0 +2023-03-21 08:58:49,292 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 08:59:01,720 INFO [train.py:901] (0/2) Epoch 38, batch 1500, loss[loss=0.1102, simple_loss=0.1899, pruned_loss=0.01525, over 7156.00 frames. ], tot_loss[loss=0.1317, simple_loss=0.2126, pruned_loss=0.02539, over 1440626.67 frames. ], batch size: 39, lr: 4.34e-03, grad_scale: 8.0 +2023-03-21 08:59:06,389 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 08:59:14,363 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3222, 2.7939, 2.0352, 3.4859, 3.3384, 3.3577, 3.1726, 3.0866], + device='cuda:0'), covar=tensor([0.2341, 0.1201, 0.4228, 0.0706, 0.0267, 0.0367, 0.0440, 0.0517], + device='cuda:0'), in_proj_covar=tensor([0.0257, 0.0233, 0.0252, 0.0260, 0.0201, 0.0196, 0.0218, 0.0227], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 08:59:27,814 INFO [train.py:901] (0/2) Epoch 38, batch 1550, loss[loss=0.1282, simple_loss=0.2111, pruned_loss=0.02267, over 7286.00 frames. ], tot_loss[loss=0.1319, simple_loss=0.2132, pruned_loss=0.02524, over 1443806.77 frames. ], batch size: 66, lr: 4.34e-03, grad_scale: 8.0 +2023-03-21 08:59:29,317 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 08:59:30,470 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=106044.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:59:33,489 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.162e+02 1.752e+02 2.054e+02 2.320e+02 6.446e+02, threshold=4.109e+02, percent-clipped=1.0 +2023-03-21 08:59:36,381 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.86 vs. limit=2.0 +2023-03-21 08:59:47,819 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=106077.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 08:59:54,275 INFO [train.py:901] (0/2) Epoch 38, batch 1600, loss[loss=0.1426, simple_loss=0.2256, pruned_loss=0.02986, over 7329.00 frames. ], tot_loss[loss=0.1318, simple_loss=0.2134, pruned_loss=0.02509, over 1445753.35 frames. ], batch size: 49, lr: 4.34e-03, grad_scale: 8.0 +2023-03-21 08:59:55,846 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=106092.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:00:02,454 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 09:00:02,964 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 09:00:05,482 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 09:00:13,896 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.77 vs. limit=2.0 +2023-03-21 09:00:15,608 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 09:00:19,943 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=106138.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:00:20,300 INFO [train.py:901] (0/2) Epoch 38, batch 1650, loss[loss=0.1377, simple_loss=0.2171, pruned_loss=0.02915, over 7302.00 frames. ], tot_loss[loss=0.1313, simple_loss=0.2126, pruned_loss=0.02494, over 1446408.29 frames. ], batch size: 83, lr: 4.34e-03, grad_scale: 8.0 +2023-03-21 09:00:20,323 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 09:00:21,906 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=106142.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:00:24,937 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=106148.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:00:25,299 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.144e+02 1.757e+02 2.005e+02 2.321e+02 4.774e+02, threshold=4.009e+02, percent-clipped=1.0 +2023-03-21 09:00:27,862 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 09:00:28,415 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=106155.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:00:30,001 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=106158.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:00:44,077 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 09:00:45,992 INFO [train.py:901] (0/2) Epoch 38, batch 1700, loss[loss=0.1226, simple_loss=0.2021, pruned_loss=0.02154, over 7302.00 frames. ], tot_loss[loss=0.1314, simple_loss=0.213, pruned_loss=0.02492, over 1446420.35 frames. ], batch size: 68, lr: 4.34e-03, grad_scale: 8.0 +2023-03-21 09:00:46,551 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=106190.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:00:48,567 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 09:00:54,914 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=106206.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:00:59,376 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 09:01:11,910 INFO [train.py:901] (0/2) Epoch 38, batch 1750, loss[loss=0.1459, simple_loss=0.2302, pruned_loss=0.03086, over 7361.00 frames. ], tot_loss[loss=0.1311, simple_loss=0.2124, pruned_loss=0.02496, over 1444485.57 frames. ], batch size: 63, lr: 4.34e-03, grad_scale: 8.0 +2023-03-21 09:01:16,876 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.200e+02 1.849e+02 2.238e+02 2.695e+02 5.615e+02, threshold=4.475e+02, percent-clipped=4.0 +2023-03-21 09:01:24,443 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 09:01:25,954 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 09:01:27,041 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=106268.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:01:35,428 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9161, 4.3517, 4.3148, 4.3571, 4.3513, 3.9649, 4.4011, 4.1934], + device='cuda:0'), covar=tensor([0.0986, 0.1128, 0.1033, 0.0935, 0.0757, 0.1006, 0.0902, 0.1117], + device='cuda:0'), in_proj_covar=tensor([0.0217, 0.0255, 0.0200, 0.0201, 0.0159, 0.0231, 0.0209, 0.0152], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:01:37,319 INFO [train.py:901] (0/2) Epoch 38, batch 1800, loss[loss=0.1224, simple_loss=0.2124, pruned_loss=0.01625, over 7314.00 frames. ], tot_loss[loss=0.1309, simple_loss=0.212, pruned_loss=0.02488, over 1442586.45 frames. ], batch size: 83, lr: 4.34e-03, grad_scale: 8.0 +2023-03-21 09:01:46,129 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 09:01:54,411 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3664, 4.9425, 5.0120, 4.9631, 4.8490, 4.4592, 5.0278, 4.8336], + device='cuda:0'), covar=tensor([0.0497, 0.0381, 0.0373, 0.0478, 0.0351, 0.0444, 0.0366, 0.0480], + device='cuda:0'), in_proj_covar=tensor([0.0218, 0.0255, 0.0200, 0.0201, 0.0159, 0.0230, 0.0209, 0.0152], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:01:57,994 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=106329.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:01:59,900 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 09:02:02,074 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=106336.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:02:03,437 INFO [train.py:901] (0/2) Epoch 38, batch 1850, loss[loss=0.1275, simple_loss=0.2181, pruned_loss=0.01844, over 7284.00 frames. ], tot_loss[loss=0.1301, simple_loss=0.2111, pruned_loss=0.0245, over 1440434.50 frames. ], batch size: 77, lr: 4.34e-03, grad_scale: 8.0 +2023-03-21 09:02:08,335 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.034e+02 1.796e+02 2.169e+02 2.438e+02 6.598e+02, threshold=4.338e+02, percent-clipped=1.0 +2023-03-21 09:02:09,904 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 09:02:26,992 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 09:02:28,475 INFO [train.py:901] (0/2) Epoch 38, batch 1900, loss[loss=0.1355, simple_loss=0.2179, pruned_loss=0.0266, over 7240.00 frames. ], tot_loss[loss=0.1301, simple_loss=0.2111, pruned_loss=0.0245, over 1440880.55 frames. ], batch size: 45, lr: 4.33e-03, grad_scale: 8.0 +2023-03-21 09:02:32,618 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=106397.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:02:52,416 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=106433.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:02:52,855 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 09:02:55,360 INFO [train.py:901] (0/2) Epoch 38, batch 1950, loss[loss=0.1468, simple_loss=0.2281, pruned_loss=0.03274, over 7280.00 frames. ], tot_loss[loss=0.1305, simple_loss=0.2115, pruned_loss=0.02471, over 1440909.78 frames. ], batch size: 66, lr: 4.33e-03, grad_scale: 8.0 +2023-03-21 09:03:00,017 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=106448.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:03:00,385 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.321e+02 1.800e+02 2.066e+02 2.441e+02 4.062e+02, threshold=4.133e+02, percent-clipped=0.0 +2023-03-21 09:03:03,094 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=106454.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:03:03,467 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 09:03:03,543 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=106455.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:03:08,527 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 09:03:09,539 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 09:03:11,212 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4895, 2.4196, 2.4627, 3.6140, 1.8950, 3.8245, 1.6181, 3.2912], + device='cuda:0'), covar=tensor([0.0153, 0.1350, 0.1744, 0.0186, 0.3966, 0.0233, 0.1313, 0.0405], + device='cuda:0'), in_proj_covar=tensor([0.0207, 0.0247, 0.0262, 0.0208, 0.0249, 0.0216, 0.0227, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 09:03:20,990 INFO [train.py:901] (0/2) Epoch 38, batch 2000, loss[loss=0.1338, simple_loss=0.2231, pruned_loss=0.02224, over 7122.00 frames. ], tot_loss[loss=0.1308, simple_loss=0.212, pruned_loss=0.02482, over 1441571.70 frames. ], batch size: 98, lr: 4.33e-03, grad_scale: 8.0 +2023-03-21 09:03:22,107 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5571, 4.0837, 3.9778, 4.5641, 4.4067, 4.4243, 3.8061, 4.1309], + device='cuda:0'), covar=tensor([0.0976, 0.2580, 0.2346, 0.1076, 0.0889, 0.1429, 0.1092, 0.1394], + device='cuda:0'), in_proj_covar=tensor([0.0261, 0.0386, 0.0296, 0.0310, 0.0224, 0.0367, 0.0227, 0.0270], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:03:24,586 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=106496.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:03:26,062 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 09:03:28,087 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=106503.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:03:34,762 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=106515.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:03:38,136 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 09:03:39,939 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-03-21 09:03:44,980 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 09:03:45,287 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.50 vs. limit=5.0 +2023-03-21 09:03:46,436 INFO [train.py:901] (0/2) Epoch 38, batch 2050, loss[loss=0.1337, simple_loss=0.2177, pruned_loss=0.02483, over 7282.00 frames. ], tot_loss[loss=0.1315, simple_loss=0.2129, pruned_loss=0.02504, over 1443953.53 frames. ], batch size: 77, lr: 4.33e-03, grad_scale: 8.0 +2023-03-21 09:03:51,401 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.246e+02 1.730e+02 1.940e+02 2.264e+02 3.831e+02, threshold=3.879e+02, percent-clipped=0.0 +2023-03-21 09:03:58,561 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2248, 4.5128, 4.3264, 4.5565, 4.0698, 4.4930, 4.7370, 4.8371], + device='cuda:0'), covar=tensor([0.0217, 0.0136, 0.0168, 0.0127, 0.0308, 0.0252, 0.0214, 0.0162], + device='cuda:0'), in_proj_covar=tensor([0.0121, 0.0125, 0.0119, 0.0122, 0.0114, 0.0103, 0.0098, 0.0100], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:04:12,062 INFO [train.py:901] (0/2) Epoch 38, batch 2100, loss[loss=0.1439, simple_loss=0.2306, pruned_loss=0.02859, over 7263.00 frames. ], tot_loss[loss=0.1313, simple_loss=0.2127, pruned_loss=0.02493, over 1445050.38 frames. ], batch size: 57, lr: 4.33e-03, grad_scale: 8.0 +2023-03-21 09:04:16,847 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.87 vs. limit=2.0 +2023-03-21 09:04:20,456 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 09:04:22,918 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 09:04:30,485 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=106624.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:04:32,067 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4883, 1.8825, 1.5171, 1.7490, 1.9311, 1.7704, 1.7969, 1.4051], + device='cuda:0'), covar=tensor([0.0204, 0.0137, 0.0260, 0.0165, 0.0114, 0.0137, 0.0147, 0.0173], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0036, 0.0035, 0.0036, 0.0035, 0.0033, 0.0037, 0.0044], + device='cuda:0'), out_proj_covar=tensor([4.2235e-05, 3.9360e-05, 3.9493e-05, 4.0281e-05, 3.8599e-05, 3.7109e-05, + 4.1501e-05, 4.9208e-05], device='cuda:0') +2023-03-21 09:04:37,921 INFO [train.py:901] (0/2) Epoch 38, batch 2150, loss[loss=0.1348, simple_loss=0.2205, pruned_loss=0.02453, over 7346.00 frames. ], tot_loss[loss=0.131, simple_loss=0.2121, pruned_loss=0.02492, over 1443288.75 frames. ], batch size: 61, lr: 4.33e-03, grad_scale: 8.0 +2023-03-21 09:04:42,848 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.352e+02 1.839e+02 2.093e+02 2.461e+02 3.515e+02, threshold=4.185e+02, percent-clipped=0.0 +2023-03-21 09:04:46,098 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0147, 2.7566, 3.0893, 3.0784, 2.8403, 2.6408, 3.0966, 2.2740], + device='cuda:0'), covar=tensor([0.0549, 0.0534, 0.0675, 0.0742, 0.0718, 0.0945, 0.0688, 0.2408], + device='cuda:0'), in_proj_covar=tensor([0.0332, 0.0338, 0.0274, 0.0362, 0.0294, 0.0289, 0.0350, 0.0252], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:05:04,245 INFO [train.py:901] (0/2) Epoch 38, batch 2200, loss[loss=0.1001, simple_loss=0.1794, pruned_loss=0.01035, over 7164.00 frames. ], tot_loss[loss=0.1305, simple_loss=0.2116, pruned_loss=0.02465, over 1442587.23 frames. ], batch size: 41, lr: 4.33e-03, grad_scale: 8.0 +2023-03-21 09:05:05,586 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.50 vs. limit=2.0 +2023-03-21 09:05:05,836 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=106692.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:05:07,270 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 09:05:08,612 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.81 vs. limit=2.0 +2023-03-21 09:05:26,384 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=106733.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:05:29,247 INFO [train.py:901] (0/2) Epoch 38, batch 2250, loss[loss=0.1005, simple_loss=0.1804, pruned_loss=0.01028, over 7120.00 frames. ], tot_loss[loss=0.1303, simple_loss=0.2116, pruned_loss=0.02452, over 1442780.59 frames. ], batch size: 39, lr: 4.33e-03, grad_scale: 8.0 +2023-03-21 09:05:34,241 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.029e+02 1.807e+02 2.184e+02 2.460e+02 5.015e+02, threshold=4.369e+02, percent-clipped=2.0 +2023-03-21 09:05:42,568 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 09:05:43,058 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 09:05:51,693 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=106781.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:05:55,686 INFO [train.py:901] (0/2) Epoch 38, batch 2300, loss[loss=0.1326, simple_loss=0.2134, pruned_loss=0.02593, over 7250.00 frames. ], tot_loss[loss=0.1299, simple_loss=0.2108, pruned_loss=0.02454, over 1439907.22 frames. ], batch size: 55, lr: 4.33e-03, grad_scale: 8.0 +2023-03-21 09:05:56,705 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 09:06:06,575 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=106810.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:06:10,149 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7292, 1.4710, 1.9035, 2.1273, 1.9410, 2.0605, 1.7017, 2.0748], + device='cuda:0'), covar=tensor([0.1956, 0.5212, 0.1104, 0.0782, 0.1582, 0.1725, 0.1909, 0.1695], + device='cuda:0'), in_proj_covar=tensor([0.0077, 0.0078, 0.0070, 0.0062, 0.0061, 0.0059, 0.0101, 0.0066], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:06:21,176 INFO [train.py:901] (0/2) Epoch 38, batch 2350, loss[loss=0.1238, simple_loss=0.2053, pruned_loss=0.0211, over 7346.00 frames. ], tot_loss[loss=0.1306, simple_loss=0.2115, pruned_loss=0.02479, over 1440522.06 frames. ], batch size: 44, lr: 4.33e-03, grad_scale: 8.0 +2023-03-21 09:06:21,326 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2236, 2.5348, 2.6497, 2.3811, 2.3198, 2.5195, 2.1308, 1.8785], + device='cuda:0'), covar=tensor([0.0636, 0.0467, 0.0183, 0.0255, 0.0661, 0.0318, 0.0548, 0.0378], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0039, 0.0039, 0.0039, 0.0037, 0.0037, 0.0042, 0.0041], + device='cuda:0'), out_proj_covar=tensor([1.0146e-04, 1.0082e-04, 9.9440e-05, 9.8669e-05, 9.6755e-05, 9.6656e-05, + 1.0456e-04, 1.0479e-04], device='cuda:0') +2023-03-21 09:06:26,820 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 9.316e+01 1.748e+02 2.140e+02 2.533e+02 7.476e+02, threshold=4.280e+02, percent-clipped=2.0 +2023-03-21 09:06:34,160 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7849, 1.4790, 1.9828, 2.2148, 2.0113, 2.1869, 1.7880, 2.2291], + device='cuda:0'), covar=tensor([0.2625, 0.5068, 0.2144, 0.0992, 0.2077, 0.1956, 0.2243, 0.2196], + device='cuda:0'), in_proj_covar=tensor([0.0077, 0.0078, 0.0070, 0.0062, 0.0061, 0.0060, 0.0101, 0.0066], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:06:42,036 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 09:06:47,467 INFO [train.py:901] (0/2) Epoch 38, batch 2400, loss[loss=0.137, simple_loss=0.2184, pruned_loss=0.02781, over 7319.00 frames. ], tot_loss[loss=0.1313, simple_loss=0.2122, pruned_loss=0.0252, over 1438101.63 frames. ], batch size: 75, lr: 4.32e-03, grad_scale: 8.0 +2023-03-21 09:06:48,489 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 09:06:58,539 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 09:07:01,605 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 09:07:05,197 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=106924.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:07:08,025 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-03-21 09:07:13,217 INFO [train.py:901] (0/2) Epoch 38, batch 2450, loss[loss=0.1106, simple_loss=0.183, pruned_loss=0.0191, over 7219.00 frames. ], tot_loss[loss=0.1305, simple_loss=0.2115, pruned_loss=0.02478, over 1438107.68 frames. ], batch size: 39, lr: 4.32e-03, grad_scale: 8.0 +2023-03-21 09:07:18,719 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.280e+02 1.784e+02 2.053e+02 2.551e+02 4.549e+02, threshold=4.105e+02, percent-clipped=1.0 +2023-03-21 09:07:27,303 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8834, 5.4760, 5.5502, 5.4835, 5.2123, 4.9696, 5.5285, 5.3040], + device='cuda:0'), covar=tensor([0.0502, 0.0307, 0.0299, 0.0381, 0.0342, 0.0417, 0.0310, 0.0434], + device='cuda:0'), in_proj_covar=tensor([0.0216, 0.0255, 0.0200, 0.0199, 0.0158, 0.0229, 0.0209, 0.0152], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:07:28,790 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 09:07:30,372 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=106972.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:07:38,877 INFO [train.py:901] (0/2) Epoch 38, batch 2500, loss[loss=0.1335, simple_loss=0.2203, pruned_loss=0.02337, over 7238.00 frames. ], tot_loss[loss=0.1306, simple_loss=0.2117, pruned_loss=0.02473, over 1440004.51 frames. ], batch size: 93, lr: 4.32e-03, grad_scale: 8.0 +2023-03-21 09:07:40,507 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=106992.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:07:48,827 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.83 vs. limit=2.0 +2023-03-21 09:07:54,710 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 09:08:05,293 INFO [train.py:901] (0/2) Epoch 38, batch 2550, loss[loss=0.1111, simple_loss=0.1817, pruned_loss=0.02029, over 7026.00 frames. ], tot_loss[loss=0.1308, simple_loss=0.2117, pruned_loss=0.02496, over 1439677.57 frames. ], batch size: 35, lr: 4.32e-03, grad_scale: 8.0 +2023-03-21 09:08:05,439 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7506, 1.4072, 1.9405, 2.1337, 1.8962, 2.1029, 1.6952, 2.1233], + device='cuda:0'), covar=tensor([0.1719, 0.4121, 0.1483, 0.1087, 0.1701, 0.1948, 0.1936, 0.2708], + device='cuda:0'), in_proj_covar=tensor([0.0077, 0.0078, 0.0070, 0.0062, 0.0061, 0.0060, 0.0101, 0.0066], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:08:05,857 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=107040.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:08:10,295 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.329e+02 1.774e+02 2.109e+02 2.411e+02 4.197e+02, threshold=4.218e+02, percent-clipped=1.0 +2023-03-21 09:08:10,507 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3235, 2.9668, 3.2506, 3.3289, 3.0557, 2.8161, 3.3074, 2.5278], + device='cuda:0'), covar=tensor([0.0500, 0.0467, 0.0764, 0.0615, 0.0662, 0.0882, 0.0610, 0.2166], + device='cuda:0'), in_proj_covar=tensor([0.0327, 0.0334, 0.0270, 0.0356, 0.0291, 0.0285, 0.0346, 0.0248], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:08:15,938 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7738, 2.9703, 3.6082, 3.6634, 3.6469, 3.6744, 3.6434, 3.5634], + device='cuda:0'), covar=tensor([0.0023, 0.0121, 0.0034, 0.0034, 0.0034, 0.0035, 0.0055, 0.0053], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0069, 0.0057, 0.0056, 0.0055, 0.0060, 0.0048, 0.0077], + device='cuda:0'), out_proj_covar=tensor([8.1791e-05, 1.3992e-04, 1.0368e-04, 9.6679e-05, 9.3640e-05, 1.0527e-04, + 9.1801e-05, 1.4347e-04], device='cuda:0') +2023-03-21 09:08:30,033 INFO [train.py:901] (0/2) Epoch 38, batch 2600, loss[loss=0.1299, simple_loss=0.2207, pruned_loss=0.01958, over 7278.00 frames. ], tot_loss[loss=0.1304, simple_loss=0.2111, pruned_loss=0.02482, over 1438275.57 frames. ], batch size: 77, lr: 4.32e-03, grad_scale: 8.0 +2023-03-21 09:08:40,512 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=107110.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:08:54,450 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4626, 3.0699, 3.4202, 3.3756, 3.2632, 3.0097, 3.6307, 2.7262], + device='cuda:0'), covar=tensor([0.0413, 0.0464, 0.0677, 0.0601, 0.0664, 0.0933, 0.0663, 0.2088], + device='cuda:0'), in_proj_covar=tensor([0.0326, 0.0332, 0.0269, 0.0355, 0.0289, 0.0284, 0.0345, 0.0247], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:08:55,266 INFO [train.py:901] (0/2) Epoch 38, batch 2650, loss[loss=0.1255, simple_loss=0.2147, pruned_loss=0.01818, over 7204.00 frames. ], tot_loss[loss=0.1306, simple_loss=0.2117, pruned_loss=0.02474, over 1438291.09 frames. ], batch size: 93, lr: 4.32e-03, grad_scale: 8.0 +2023-03-21 09:08:55,886 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=107140.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 09:08:57,545 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.95 vs. limit=2.0 +2023-03-21 09:09:00,237 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.213e+02 1.805e+02 2.120e+02 2.557e+02 4.024e+02, threshold=4.240e+02, percent-clipped=0.0 +2023-03-21 09:09:01,782 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=107152.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:09:04,643 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=107158.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:09:20,368 INFO [train.py:901] (0/2) Epoch 38, batch 2700, loss[loss=0.1354, simple_loss=0.2183, pruned_loss=0.02623, over 7272.00 frames. ], tot_loss[loss=0.1303, simple_loss=0.2116, pruned_loss=0.0245, over 1438793.94 frames. ], batch size: 47, lr: 4.32e-03, grad_scale: 8.0 +2023-03-21 09:09:26,657 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=107201.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 09:09:32,438 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=107213.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 09:09:44,939 INFO [train.py:901] (0/2) Epoch 38, batch 2750, loss[loss=0.1435, simple_loss=0.2246, pruned_loss=0.03127, over 7264.00 frames. ], tot_loss[loss=0.1313, simple_loss=0.2122, pruned_loss=0.02514, over 1439286.90 frames. ], batch size: 55, lr: 4.32e-03, grad_scale: 8.0 +2023-03-21 09:09:49,724 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.156e+02 1.753e+02 2.094e+02 2.542e+02 3.915e+02, threshold=4.187e+02, percent-clipped=0.0 +2023-03-21 09:10:09,090 INFO [train.py:901] (0/2) Epoch 38, batch 2800, loss[loss=0.1393, simple_loss=0.2157, pruned_loss=0.03147, over 7308.00 frames. ], tot_loss[loss=0.1314, simple_loss=0.2124, pruned_loss=0.02521, over 1441214.70 frames. ], batch size: 49, lr: 4.32e-03, grad_scale: 8.0 +2023-03-21 09:10:21,637 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-38.pt +2023-03-21 09:10:37,066 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 09:10:37,120 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 09:10:37,146 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 09:10:37,164 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 09:10:37,338 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 09:10:37,353 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 09:10:37,364 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 09:10:37,780 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 09:10:37,854 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 09:10:37,870 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 09:10:37,886 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 09:10:38,048 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 09:10:38,053 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 09:10:38,061 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 09:10:38,200 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 09:10:38,202 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 09:10:38,261 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 09:10:40,359 INFO [train.py:901] (0/2) Epoch 39, batch 0, loss[loss=0.1287, simple_loss=0.2098, pruned_loss=0.02383, over 7270.00 frames. ], tot_loss[loss=0.1287, simple_loss=0.2098, pruned_loss=0.02383, over 7270.00 frames. ], batch size: 52, lr: 4.26e-03, grad_scale: 8.0 +2023-03-21 09:10:40,360 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 09:10:49,843 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3216, 3.3306, 2.6393, 3.8360, 2.7675, 3.1709, 1.8487, 2.8386], + device='cuda:0'), covar=tensor([0.0376, 0.0630, 0.2267, 0.0450, 0.0342, 0.0713, 0.3836, 0.1695], + device='cuda:0'), in_proj_covar=tensor([0.0278, 0.0257, 0.0282, 0.0270, 0.0272, 0.0266, 0.0235, 0.0260], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 09:10:54,136 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.1619, 5.4987, 5.5438, 5.5033, 5.1007, 5.0203, 5.5312, 5.1925], + device='cuda:0'), covar=tensor([0.0333, 0.0265, 0.0262, 0.0331, 0.0346, 0.0288, 0.0257, 0.0435], + device='cuda:0'), in_proj_covar=tensor([0.0217, 0.0257, 0.0201, 0.0200, 0.0159, 0.0229, 0.0209, 0.0151], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:11:04,242 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7601, 1.5532, 1.6547, 2.0768, 1.8287, 1.9795, 1.4207, 1.9987], + device='cuda:0'), covar=tensor([0.2246, 0.4741, 0.1712, 0.1435, 0.2329, 0.2184, 0.2474, 0.1391], + device='cuda:0'), in_proj_covar=tensor([0.0076, 0.0077, 0.0070, 0.0062, 0.0061, 0.0060, 0.0101, 0.0066], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:11:05,882 INFO [train.py:935] (0/2) Epoch 39, validation: loss=0.1649, simple_loss=0.257, pruned_loss=0.03634, over 1622729.00 frames. +2023-03-21 09:11:05,883 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 09:11:12,872 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 09:11:17,202 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.82 vs. limit=5.0 +2023-03-21 09:11:22,898 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 09:11:23,860 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.279e+02 1.717e+02 1.963e+02 2.298e+02 3.449e+02, threshold=3.926e+02, percent-clipped=0.0 +2023-03-21 09:11:30,539 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 09:11:31,502 INFO [train.py:901] (0/2) Epoch 39, batch 50, loss[loss=0.1389, simple_loss=0.2252, pruned_loss=0.0263, over 7240.00 frames. ], tot_loss[loss=0.133, simple_loss=0.215, pruned_loss=0.02553, over 327754.12 frames. ], batch size: 93, lr: 4.26e-03, grad_scale: 8.0 +2023-03-21 09:11:33,081 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 09:11:35,474 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 09:11:51,328 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.74 vs. limit=2.0 +2023-03-21 09:11:52,921 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 09:11:53,354 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 09:11:57,313 INFO [train.py:901] (0/2) Epoch 39, batch 100, loss[loss=0.1327, simple_loss=0.2155, pruned_loss=0.0249, over 7347.00 frames. ], tot_loss[loss=0.1331, simple_loss=0.2143, pruned_loss=0.02598, over 574587.77 frames. ], batch size: 54, lr: 4.26e-03, grad_scale: 8.0 +2023-03-21 09:12:15,950 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.043e+02 1.717e+02 1.993e+02 2.406e+02 7.451e+02, threshold=3.985e+02, percent-clipped=2.0 +2023-03-21 09:12:23,011 INFO [train.py:901] (0/2) Epoch 39, batch 150, loss[loss=0.1526, simple_loss=0.2359, pruned_loss=0.0346, over 6698.00 frames. ], tot_loss[loss=0.1309, simple_loss=0.2125, pruned_loss=0.02468, over 766214.24 frames. ], batch size: 106, lr: 4.26e-03, grad_scale: 8.0 +2023-03-21 09:12:40,180 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=107496.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 09:12:46,183 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=107508.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 09:12:47,245 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=107510.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:12:48,620 INFO [train.py:901] (0/2) Epoch 39, batch 200, loss[loss=0.1329, simple_loss=0.2197, pruned_loss=0.02303, over 7244.00 frames. ], tot_loss[loss=0.1309, simple_loss=0.2122, pruned_loss=0.02479, over 916106.09 frames. ], batch size: 55, lr: 4.26e-03, grad_scale: 8.0 +2023-03-21 09:12:52,631 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 09:12:57,188 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 09:13:03,286 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 09:13:07,779 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.289e+02 1.722e+02 1.965e+02 2.423e+02 4.640e+02, threshold=3.930e+02, percent-clipped=2.0 +2023-03-21 09:13:15,074 INFO [train.py:901] (0/2) Epoch 39, batch 250, loss[loss=0.1438, simple_loss=0.2317, pruned_loss=0.02796, over 6733.00 frames. ], tot_loss[loss=0.1311, simple_loss=0.2128, pruned_loss=0.02468, over 1033092.01 frames. ], batch size: 106, lr: 4.25e-03, grad_scale: 8.0 +2023-03-21 09:13:16,535 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 09:13:19,140 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=107571.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:13:24,676 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0300, 2.6862, 2.9910, 3.0625, 2.8737, 2.6488, 2.8854, 2.2941], + device='cuda:0'), covar=tensor([0.0455, 0.0592, 0.0684, 0.0748, 0.0793, 0.1017, 0.0691, 0.2227], + device='cuda:0'), in_proj_covar=tensor([0.0325, 0.0333, 0.0270, 0.0355, 0.0290, 0.0286, 0.0347, 0.0248], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:13:37,333 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 09:13:40,234 INFO [train.py:901] (0/2) Epoch 39, batch 300, loss[loss=0.1449, simple_loss=0.2229, pruned_loss=0.03341, over 7205.00 frames. ], tot_loss[loss=0.1308, simple_loss=0.2122, pruned_loss=0.02466, over 1122512.16 frames. ], batch size: 50, lr: 4.25e-03, grad_scale: 8.0 +2023-03-21 09:13:45,483 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9608, 4.2570, 4.0301, 4.1917, 3.8950, 3.9413, 4.2054, 4.3143], + device='cuda:0'), covar=tensor([0.0336, 0.0192, 0.0281, 0.0244, 0.0388, 0.0397, 0.0410, 0.0310], + device='cuda:0'), in_proj_covar=tensor([0.0120, 0.0123, 0.0117, 0.0121, 0.0111, 0.0101, 0.0096, 0.0098], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:13:46,372 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 09:13:59,550 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.007e+02 1.763e+02 1.989e+02 2.300e+02 3.698e+02, threshold=3.979e+02, percent-clipped=0.0 +2023-03-21 09:14:01,646 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9339, 5.5411, 5.5574, 5.4999, 5.2414, 5.0527, 5.5765, 5.4212], + device='cuda:0'), covar=tensor([0.0454, 0.0317, 0.0350, 0.0469, 0.0344, 0.0393, 0.0291, 0.0396], + device='cuda:0'), in_proj_covar=tensor([0.0217, 0.0256, 0.0201, 0.0202, 0.0160, 0.0229, 0.0209, 0.0150], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:14:04,732 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=107659.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:14:06,580 INFO [train.py:901] (0/2) Epoch 39, batch 350, loss[loss=0.1203, simple_loss=0.1961, pruned_loss=0.0222, over 7278.00 frames. ], tot_loss[loss=0.13, simple_loss=0.211, pruned_loss=0.02447, over 1190347.47 frames. ], batch size: 52, lr: 4.25e-03, grad_scale: 8.0 +2023-03-21 09:14:13,104 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6177, 3.0991, 2.6320, 2.7806, 2.8069, 2.6393, 2.9463, 2.7962], + device='cuda:0'), covar=tensor([0.0791, 0.0620, 0.1012, 0.1047, 0.0940, 0.0625, 0.0786, 0.0881], + device='cuda:0'), in_proj_covar=tensor([0.0058, 0.0058, 0.0066, 0.0059, 0.0056, 0.0061, 0.0056, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:14:22,322 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 09:14:31,998 INFO [train.py:901] (0/2) Epoch 39, batch 400, loss[loss=0.1198, simple_loss=0.1872, pruned_loss=0.0262, over 7028.00 frames. ], tot_loss[loss=0.13, simple_loss=0.2114, pruned_loss=0.02429, over 1248498.86 frames. ], batch size: 35, lr: 4.25e-03, grad_scale: 16.0 +2023-03-21 09:14:35,594 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=107720.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:14:38,104 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5270, 1.7786, 1.4783, 1.5564, 1.7683, 1.6701, 1.6657, 1.4159], + device='cuda:0'), covar=tensor([0.0165, 0.0178, 0.0470, 0.0228, 0.0163, 0.0168, 0.0219, 0.0165], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0037, 0.0036, 0.0037, 0.0036, 0.0034, 0.0038, 0.0045], + device='cuda:0'), out_proj_covar=tensor([4.2950e-05, 4.1099e-05, 4.0492e-05, 4.1117e-05, 3.9688e-05, 3.7669e-05, + 4.2721e-05, 5.0020e-05], device='cuda:0') +2023-03-21 09:14:43,762 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2500, 2.4010, 2.4198, 2.1938, 2.3700, 2.3120, 1.9072, 1.7151], + device='cuda:0'), covar=tensor([0.0356, 0.0364, 0.0358, 0.0210, 0.0362, 0.0472, 0.0483, 0.0416], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0038, 0.0039, 0.0038, 0.0036, 0.0037, 0.0042, 0.0041], + device='cuda:0'), out_proj_covar=tensor([1.0031e-04, 9.9455e-05, 9.8534e-05, 9.7228e-05, 9.5466e-05, 9.4830e-05, + 1.0341e-04, 1.0436e-04], device='cuda:0') +2023-03-21 09:14:44,777 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=107737.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:14:50,591 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.205e+02 1.824e+02 2.104e+02 2.384e+02 4.420e+02, threshold=4.207e+02, percent-clipped=2.0 +2023-03-21 09:14:57,695 INFO [train.py:901] (0/2) Epoch 39, batch 450, loss[loss=0.129, simple_loss=0.2093, pruned_loss=0.02436, over 7219.00 frames. ], tot_loss[loss=0.1297, simple_loss=0.2111, pruned_loss=0.02416, over 1291175.12 frames. ], batch size: 45, lr: 4.25e-03, grad_scale: 16.0 +2023-03-21 09:15:03,160 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 09:15:03,173 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 09:15:14,972 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=107796.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 09:15:15,996 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=107798.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 09:15:21,323 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=107808.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:15:23,760 INFO [train.py:901] (0/2) Epoch 39, batch 500, loss[loss=0.1303, simple_loss=0.2111, pruned_loss=0.02474, over 7270.00 frames. ], tot_loss[loss=0.1288, simple_loss=0.21, pruned_loss=0.02384, over 1322174.51 frames. ], batch size: 47, lr: 4.25e-03, grad_scale: 16.0 +2023-03-21 09:15:37,394 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 09:15:38,924 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 09:15:39,378 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 09:15:39,932 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=107844.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 09:15:41,929 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 09:15:42,401 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.285e+02 1.745e+02 2.038e+02 2.318e+02 3.534e+02, threshold=4.076e+02, percent-clipped=0.0 +2023-03-21 09:15:45,875 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=107856.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:15:46,367 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 09:15:49,359 INFO [train.py:901] (0/2) Epoch 39, batch 550, loss[loss=0.1614, simple_loss=0.2443, pruned_loss=0.0392, over 6710.00 frames. ], tot_loss[loss=0.1285, simple_loss=0.2094, pruned_loss=0.02378, over 1347413.61 frames. ], batch size: 106, lr: 4.25e-03, grad_scale: 8.0 +2023-03-21 09:15:50,906 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=107866.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:15:57,429 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3091, 3.5777, 2.4756, 3.9383, 3.0155, 3.2700, 1.7274, 2.5602], + device='cuda:0'), covar=tensor([0.0432, 0.0798, 0.2646, 0.0566, 0.0426, 0.0609, 0.4016, 0.1791], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0255, 0.0281, 0.0270, 0.0270, 0.0264, 0.0235, 0.0258], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 09:15:57,760 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 09:16:06,939 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 09:16:10,486 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 09:16:15,527 INFO [train.py:901] (0/2) Epoch 39, batch 600, loss[loss=0.107, simple_loss=0.1627, pruned_loss=0.02569, over 5745.00 frames. ], tot_loss[loss=0.1295, simple_loss=0.2102, pruned_loss=0.02434, over 1368575.12 frames. ], batch size: 25, lr: 4.25e-03, grad_scale: 8.0 +2023-03-21 09:16:17,518 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 09:16:24,173 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3867, 3.1209, 3.2889, 3.2584, 3.0103, 2.8708, 3.3243, 2.4694], + device='cuda:0'), covar=tensor([0.0555, 0.0577, 0.0690, 0.0605, 0.0705, 0.0985, 0.0985, 0.2269], + device='cuda:0'), in_proj_covar=tensor([0.0327, 0.0335, 0.0271, 0.0355, 0.0289, 0.0285, 0.0349, 0.0247], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:16:26,635 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6876, 2.2076, 2.7062, 2.5993, 2.6227, 2.5871, 2.2781, 2.7133], + device='cuda:0'), covar=tensor([0.1250, 0.1060, 0.0812, 0.0950, 0.0729, 0.0890, 0.1748, 0.0910], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0067, 0.0051, 0.0050, 0.0050, 0.0048, 0.0067, 0.0051], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 09:16:33,502 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 09:16:33,982 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.115e+02 1.697e+02 2.000e+02 2.452e+02 4.115e+02, threshold=4.000e+02, percent-clipped=1.0 +2023-03-21 09:16:37,143 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.1222, 5.6397, 5.6509, 5.5847, 5.4493, 5.3057, 5.7112, 5.6044], + device='cuda:0'), covar=tensor([0.0442, 0.0329, 0.0383, 0.0455, 0.0300, 0.0313, 0.0284, 0.0343], + device='cuda:0'), in_proj_covar=tensor([0.0216, 0.0254, 0.0200, 0.0199, 0.0159, 0.0226, 0.0206, 0.0148], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:16:40,507 INFO [train.py:901] (0/2) Epoch 39, batch 650, loss[loss=0.1286, simple_loss=0.2112, pruned_loss=0.02299, over 7292.00 frames. ], tot_loss[loss=0.1297, simple_loss=0.2103, pruned_loss=0.02453, over 1384010.30 frames. ], batch size: 68, lr: 4.25e-03, grad_scale: 8.0 +2023-03-21 09:16:41,991 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 09:16:55,258 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1425, 2.3161, 2.4816, 2.3089, 2.2400, 2.1431, 2.0926, 1.9258], + device='cuda:0'), covar=tensor([0.0735, 0.0638, 0.0504, 0.0309, 0.0817, 0.0888, 0.0399, 0.0427], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0039, 0.0039, 0.0038, 0.0036, 0.0037, 0.0041, 0.0041], + device='cuda:0'), out_proj_covar=tensor([1.0036e-04, 1.0021e-04, 9.8406e-05, 9.7139e-05, 9.5766e-05, 9.5651e-05, + 1.0288e-04, 1.0422e-04], device='cuda:0') +2023-03-21 09:16:59,568 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 09:17:00,421 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-108000.pt +2023-03-21 09:17:10,693 INFO [train.py:901] (0/2) Epoch 39, batch 700, loss[loss=0.1307, simple_loss=0.2137, pruned_loss=0.02389, over 7321.00 frames. ], tot_loss[loss=0.13, simple_loss=0.2106, pruned_loss=0.02469, over 1396530.64 frames. ], batch size: 59, lr: 4.25e-03, grad_scale: 8.0 +2023-03-21 09:17:11,757 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=108015.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:17:13,234 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 09:17:29,090 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.255e+02 1.760e+02 2.052e+02 2.429e+02 4.118e+02, threshold=4.103e+02, percent-clipped=1.0 +2023-03-21 09:17:36,201 INFO [train.py:901] (0/2) Epoch 39, batch 750, loss[loss=0.1343, simple_loss=0.2149, pruned_loss=0.02678, over 7330.00 frames. ], tot_loss[loss=0.1298, simple_loss=0.2107, pruned_loss=0.0244, over 1406445.49 frames. ], batch size: 59, lr: 4.25e-03, grad_scale: 8.0 +2023-03-21 09:17:38,247 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 09:17:38,768 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 09:17:44,366 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6535, 3.8806, 3.6834, 3.9042, 3.4825, 3.8241, 4.1025, 4.1319], + device='cuda:0'), covar=tensor([0.0234, 0.0160, 0.0253, 0.0176, 0.0391, 0.0313, 0.0242, 0.0201], + device='cuda:0'), in_proj_covar=tensor([0.0121, 0.0125, 0.0119, 0.0123, 0.0113, 0.0102, 0.0098, 0.0100], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:17:51,865 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=108093.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 09:17:52,743 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 09:17:56,677 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 09:18:01,657 INFO [train.py:901] (0/2) Epoch 39, batch 800, loss[loss=0.1387, simple_loss=0.2227, pruned_loss=0.02738, over 7322.00 frames. ], tot_loss[loss=0.13, simple_loss=0.2111, pruned_loss=0.02446, over 1415515.83 frames. ], batch size: 59, lr: 4.24e-03, grad_scale: 8.0 +2023-03-21 09:18:02,671 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 09:18:03,685 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 09:18:08,334 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7238, 4.4474, 4.4389, 4.2642, 4.1516, 3.0875, 2.4299, 4.7768], + device='cuda:0'), covar=tensor([0.0055, 0.0081, 0.0075, 0.0073, 0.0108, 0.0484, 0.0544, 0.0054], + device='cuda:0'), in_proj_covar=tensor([0.0115, 0.0094, 0.0114, 0.0095, 0.0131, 0.0136, 0.0131, 0.0105], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 09:18:13,790 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 09:18:20,774 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.260e+02 1.831e+02 2.034e+02 2.384e+02 3.789e+02, threshold=4.068e+02, percent-clipped=0.0 +2023-03-21 09:18:27,267 INFO [train.py:901] (0/2) Epoch 39, batch 850, loss[loss=0.1254, simple_loss=0.2163, pruned_loss=0.01726, over 7292.00 frames. ], tot_loss[loss=0.131, simple_loss=0.2121, pruned_loss=0.02494, over 1420035.03 frames. ], batch size: 68, lr: 4.24e-03, grad_scale: 8.0 +2023-03-21 09:18:27,514 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.17 vs. limit=5.0 +2023-03-21 09:18:29,476 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=108166.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:18:30,051 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0734, 2.5495, 1.8932, 2.7643, 2.9104, 2.6526, 2.7235, 2.6064], + device='cuda:0'), covar=tensor([0.1985, 0.1043, 0.3745, 0.0739, 0.0310, 0.0267, 0.0436, 0.0414], + device='cuda:0'), in_proj_covar=tensor([0.0251, 0.0229, 0.0248, 0.0257, 0.0198, 0.0195, 0.0216, 0.0224], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 09:18:33,925 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 09:18:34,410 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 09:18:39,452 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 09:18:43,620 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 09:18:53,338 INFO [train.py:901] (0/2) Epoch 39, batch 900, loss[loss=0.1449, simple_loss=0.227, pruned_loss=0.03145, over 7356.00 frames. ], tot_loss[loss=0.1311, simple_loss=0.2122, pruned_loss=0.02505, over 1423722.78 frames. ], batch size: 51, lr: 4.24e-03, grad_scale: 8.0 +2023-03-21 09:18:53,862 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=108214.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:19:12,502 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.190e+02 1.645e+02 1.970e+02 2.398e+02 5.896e+02, threshold=3.941e+02, percent-clipped=2.0 +2023-03-21 09:19:19,546 INFO [train.py:901] (0/2) Epoch 39, batch 950, loss[loss=0.1549, simple_loss=0.2274, pruned_loss=0.04117, over 7350.00 frames. ], tot_loss[loss=0.1307, simple_loss=0.2121, pruned_loss=0.02467, over 1430192.79 frames. ], batch size: 54, lr: 4.24e-03, grad_scale: 8.0 +2023-03-21 09:19:20,045 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 09:19:22,190 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4379, 1.6615, 1.4675, 1.4751, 1.6935, 1.6294, 1.5688, 1.1536], + device='cuda:0'), covar=tensor([0.0186, 0.0210, 0.0249, 0.0223, 0.0162, 0.0153, 0.0190, 0.0210], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0036, 0.0035, 0.0037, 0.0035, 0.0033, 0.0037, 0.0045], + device='cuda:0'), out_proj_covar=tensor([4.2342e-05, 4.0496e-05, 3.9759e-05, 4.0722e-05, 3.8931e-05, 3.7232e-05, + 4.2108e-05, 4.9924e-05], device='cuda:0') +2023-03-21 09:19:43,306 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 09:19:44,853 INFO [train.py:901] (0/2) Epoch 39, batch 1000, loss[loss=0.1463, simple_loss=0.2254, pruned_loss=0.03359, over 7314.00 frames. ], tot_loss[loss=0.1313, simple_loss=0.2128, pruned_loss=0.0249, over 1434635.89 frames. ], batch size: 80, lr: 4.24e-03, grad_scale: 8.0 +2023-03-21 09:19:45,916 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=108315.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:19:51,495 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=108325.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:20:04,451 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.251e+02 1.792e+02 2.080e+02 2.512e+02 3.266e+02, threshold=4.160e+02, percent-clipped=0.0 +2023-03-21 09:20:04,982 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 09:20:10,964 INFO [train.py:901] (0/2) Epoch 39, batch 1050, loss[loss=0.1218, simple_loss=0.2074, pruned_loss=0.01811, over 7343.00 frames. ], tot_loss[loss=0.1306, simple_loss=0.212, pruned_loss=0.02459, over 1437686.84 frames. ], batch size: 44, lr: 4.24e-03, grad_scale: 8.0 +2023-03-21 09:20:11,046 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=108363.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:20:22,712 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=108386.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:20:24,177 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7925, 3.2655, 2.6139, 2.7772, 2.7560, 2.7223, 2.7877, 2.8094], + device='cuda:0'), covar=tensor([0.0969, 0.0462, 0.1320, 0.1989, 0.1541, 0.0743, 0.1306, 0.1293], + device='cuda:0'), in_proj_covar=tensor([0.0058, 0.0058, 0.0066, 0.0059, 0.0057, 0.0061, 0.0056, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:20:26,084 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=108393.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:20:26,945 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 09:20:31,775 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 09:20:36,759 INFO [train.py:901] (0/2) Epoch 39, batch 1100, loss[loss=0.1289, simple_loss=0.2137, pruned_loss=0.02209, over 7279.00 frames. ], tot_loss[loss=0.1303, simple_loss=0.2119, pruned_loss=0.02439, over 1440525.31 frames. ], batch size: 68, lr: 4.24e-03, grad_scale: 8.0 +2023-03-21 09:20:39,002 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1704, 3.3253, 2.3547, 3.7827, 2.7625, 3.0414, 1.6139, 2.4909], + device='cuda:0'), covar=tensor([0.0478, 0.0799, 0.2631, 0.0632, 0.0529, 0.0608, 0.3866, 0.1851], + device='cuda:0'), in_proj_covar=tensor([0.0278, 0.0258, 0.0283, 0.0272, 0.0272, 0.0265, 0.0235, 0.0260], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 09:20:46,388 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=108432.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:20:51,401 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=108441.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:20:55,870 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.198e+02 1.659e+02 1.991e+02 2.465e+02 4.884e+02, threshold=3.983e+02, percent-clipped=1.0 +2023-03-21 09:21:00,984 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 09:21:01,484 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 09:21:02,527 INFO [train.py:901] (0/2) Epoch 39, batch 1150, loss[loss=0.1385, simple_loss=0.2209, pruned_loss=0.02806, over 7314.00 frames. ], tot_loss[loss=0.1307, simple_loss=0.2118, pruned_loss=0.02482, over 1439051.55 frames. ], batch size: 80, lr: 4.24e-03, grad_scale: 8.0 +2023-03-21 09:21:15,066 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 09:21:15,563 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 09:21:18,359 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=108493.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 09:21:24,305 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6051, 1.8189, 1.5283, 1.6424, 1.8654, 1.8161, 1.7230, 1.2781], + device='cuda:0'), covar=tensor([0.0167, 0.0202, 0.0250, 0.0225, 0.0133, 0.0132, 0.0164, 0.0219], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0036, 0.0034, 0.0036, 0.0035, 0.0033, 0.0037, 0.0045], + device='cuda:0'), out_proj_covar=tensor([4.1708e-05, 4.0343e-05, 3.9070e-05, 4.0316e-05, 3.8475e-05, 3.7045e-05, + 4.1462e-05, 4.9577e-05], device='cuda:0') +2023-03-21 09:21:28,210 INFO [train.py:901] (0/2) Epoch 39, batch 1200, loss[loss=0.1097, simple_loss=0.1851, pruned_loss=0.01713, over 7025.00 frames. ], tot_loss[loss=0.1313, simple_loss=0.2124, pruned_loss=0.02505, over 1439051.72 frames. ], batch size: 35, lr: 4.24e-03, grad_scale: 8.0 +2023-03-21 09:21:47,344 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.212e+02 1.679e+02 2.002e+02 2.337e+02 3.953e+02, threshold=4.003e+02, percent-clipped=0.0 +2023-03-21 09:21:47,899 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 09:21:53,858 INFO [train.py:901] (0/2) Epoch 39, batch 1250, loss[loss=0.1137, simple_loss=0.1815, pruned_loss=0.02294, over 6269.00 frames. ], tot_loss[loss=0.1312, simple_loss=0.2122, pruned_loss=0.0251, over 1437591.51 frames. ], batch size: 27, lr: 4.24e-03, grad_scale: 8.0 +2023-03-21 09:22:01,040 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=108577.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:22:03,204 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1343, 2.5712, 2.0708, 2.8926, 3.0102, 2.8823, 2.7510, 2.8018], + device='cuda:0'), covar=tensor([0.2203, 0.1025, 0.3474, 0.0767, 0.0286, 0.0252, 0.0365, 0.0466], + device='cuda:0'), in_proj_covar=tensor([0.0250, 0.0227, 0.0246, 0.0256, 0.0198, 0.0194, 0.0215, 0.0224], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 09:22:10,995 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 09:22:14,893 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 09:22:15,949 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 09:22:20,616 INFO [train.py:901] (0/2) Epoch 39, batch 1300, loss[loss=0.1379, simple_loss=0.2182, pruned_loss=0.02887, over 7250.00 frames. ], tot_loss[loss=0.1321, simple_loss=0.213, pruned_loss=0.0256, over 1438134.96 frames. ], batch size: 55, lr: 4.23e-03, grad_scale: 8.0 +2023-03-21 09:22:33,455 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=108638.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:22:39,211 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.296e+02 1.751e+02 2.074e+02 2.340e+02 4.099e+02, threshold=4.148e+02, percent-clipped=1.0 +2023-03-21 09:22:40,244 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 09:22:42,264 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 09:22:46,456 INFO [train.py:901] (0/2) Epoch 39, batch 1350, loss[loss=0.112, simple_loss=0.183, pruned_loss=0.0205, over 7034.00 frames. ], tot_loss[loss=0.1318, simple_loss=0.2127, pruned_loss=0.02548, over 1438384.07 frames. ], batch size: 35, lr: 4.23e-03, grad_scale: 8.0 +2023-03-21 09:22:46,973 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 09:22:55,612 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=108681.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:22:56,573 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 09:23:12,145 INFO [train.py:901] (0/2) Epoch 39, batch 1400, loss[loss=0.1247, simple_loss=0.2092, pruned_loss=0.02004, over 7290.00 frames. ], tot_loss[loss=0.1313, simple_loss=0.2125, pruned_loss=0.02505, over 1439133.93 frames. ], batch size: 77, lr: 4.23e-03, grad_scale: 8.0 +2023-03-21 09:23:27,757 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 09:23:31,340 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.124e+02 1.707e+02 2.003e+02 2.347e+02 5.252e+02, threshold=4.006e+02, percent-clipped=1.0 +2023-03-21 09:23:35,915 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8601, 2.9718, 2.1704, 3.2854, 2.4168, 2.7939, 1.4014, 2.1353], + device='cuda:0'), covar=tensor([0.0617, 0.1087, 0.3267, 0.0851, 0.0674, 0.0794, 0.4287, 0.2193], + device='cuda:0'), in_proj_covar=tensor([0.0278, 0.0257, 0.0281, 0.0270, 0.0271, 0.0265, 0.0234, 0.0260], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 09:23:37,765 INFO [train.py:901] (0/2) Epoch 39, batch 1450, loss[loss=0.1307, simple_loss=0.2169, pruned_loss=0.02228, over 7271.00 frames. ], tot_loss[loss=0.1313, simple_loss=0.2127, pruned_loss=0.025, over 1440841.16 frames. ], batch size: 77, lr: 4.23e-03, grad_scale: 8.0 +2023-03-21 09:23:51,095 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=108788.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 09:23:52,603 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=108791.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:23:53,016 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 09:24:03,877 INFO [train.py:901] (0/2) Epoch 39, batch 1500, loss[loss=0.1201, simple_loss=0.2051, pruned_loss=0.01756, over 7310.00 frames. ], tot_loss[loss=0.1309, simple_loss=0.2123, pruned_loss=0.02474, over 1440427.60 frames. ], batch size: 86, lr: 4.23e-03, grad_scale: 8.0 +2023-03-21 09:24:09,365 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 09:24:09,727 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.65 vs. limit=5.0 +2023-03-21 09:24:15,009 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3561, 3.4777, 2.5457, 4.0038, 3.0582, 3.3102, 1.8728, 2.6289], + device='cuda:0'), covar=tensor([0.0405, 0.0792, 0.2413, 0.0524, 0.0426, 0.0633, 0.3828, 0.1651], + device='cuda:0'), in_proj_covar=tensor([0.0277, 0.0256, 0.0279, 0.0269, 0.0270, 0.0265, 0.0234, 0.0259], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 09:24:22,946 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.141e+02 1.726e+02 1.983e+02 2.350e+02 3.872e+02, threshold=3.966e+02, percent-clipped=0.0 +2023-03-21 09:24:24,106 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=108852.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:24:29,442 INFO [train.py:901] (0/2) Epoch 39, batch 1550, loss[loss=0.1359, simple_loss=0.2172, pruned_loss=0.02729, over 7326.00 frames. ], tot_loss[loss=0.1308, simple_loss=0.2118, pruned_loss=0.02492, over 1439035.49 frames. ], batch size: 54, lr: 4.23e-03, grad_scale: 8.0 +2023-03-21 09:24:34,000 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 09:24:41,837 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7976, 2.3657, 2.3036, 3.8389, 1.7942, 3.5033, 1.2950, 3.3873], + device='cuda:0'), covar=tensor([0.0194, 0.1471, 0.1968, 0.0280, 0.4469, 0.0322, 0.1580, 0.0437], + device='cuda:0'), in_proj_covar=tensor([0.0204, 0.0244, 0.0259, 0.0205, 0.0248, 0.0215, 0.0227, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 09:24:55,275 INFO [train.py:901] (0/2) Epoch 39, batch 1600, loss[loss=0.09407, simple_loss=0.154, pruned_loss=0.0171, over 5992.00 frames. ], tot_loss[loss=0.1305, simple_loss=0.2114, pruned_loss=0.02478, over 1436912.59 frames. ], batch size: 25, lr: 4.23e-03, grad_scale: 8.0 +2023-03-21 09:25:04,880 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 09:25:05,428 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 09:25:05,989 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=108933.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:25:08,379 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 09:25:14,454 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.272e+02 1.678e+02 2.010e+02 2.437e+02 3.858e+02, threshold=4.020e+02, percent-clipped=0.0 +2023-03-21 09:25:18,606 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 09:25:21,634 INFO [train.py:901] (0/2) Epoch 39, batch 1650, loss[loss=0.1378, simple_loss=0.218, pruned_loss=0.02882, over 7232.00 frames. ], tot_loss[loss=0.1299, simple_loss=0.2109, pruned_loss=0.02442, over 1436115.76 frames. ], batch size: 50, lr: 4.23e-03, grad_scale: 8.0 +2023-03-21 09:25:22,632 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 09:25:30,140 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 09:25:30,683 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=108981.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:25:45,856 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2388, 4.4514, 4.2219, 4.4750, 3.9536, 4.4055, 4.7460, 4.8040], + device='cuda:0'), covar=tensor([0.0192, 0.0127, 0.0204, 0.0138, 0.0353, 0.0235, 0.0181, 0.0151], + device='cuda:0'), in_proj_covar=tensor([0.0122, 0.0126, 0.0120, 0.0124, 0.0114, 0.0102, 0.0098, 0.0101], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:25:47,316 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 09:25:47,795 INFO [train.py:901] (0/2) Epoch 39, batch 1700, loss[loss=0.134, simple_loss=0.2272, pruned_loss=0.02037, over 6699.00 frames. ], tot_loss[loss=0.1302, simple_loss=0.2114, pruned_loss=0.02448, over 1434303.35 frames. ], batch size: 107, lr: 4.23e-03, grad_scale: 8.0 +2023-03-21 09:25:51,831 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 09:25:56,057 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=109029.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:26:00,154 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=109037.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 09:26:03,038 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 09:26:07,087 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.192e+02 1.706e+02 2.028e+02 2.336e+02 4.447e+02, threshold=4.056e+02, percent-clipped=2.0 +2023-03-21 09:26:13,719 INFO [train.py:901] (0/2) Epoch 39, batch 1750, loss[loss=0.1244, simple_loss=0.2139, pruned_loss=0.01745, over 7352.00 frames. ], tot_loss[loss=0.13, simple_loss=0.2109, pruned_loss=0.0245, over 1435348.96 frames. ], batch size: 73, lr: 4.23e-03, grad_scale: 8.0 +2023-03-21 09:26:26,350 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=109088.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:26:27,782 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 09:26:28,274 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 09:26:32,007 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=109098.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 09:26:32,498 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5605, 2.6380, 3.5845, 3.6743, 3.7143, 3.7139, 3.3907, 3.5781], + device='cuda:0'), covar=tensor([0.0036, 0.0166, 0.0038, 0.0030, 0.0030, 0.0029, 0.0064, 0.0053], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0070, 0.0058, 0.0056, 0.0055, 0.0061, 0.0049, 0.0079], + device='cuda:0'), out_proj_covar=tensor([8.2204e-05, 1.4155e-04, 1.0482e-04, 9.7715e-05, 9.4315e-05, 1.0734e-04, + 9.3987e-05, 1.4610e-04], device='cuda:0') +2023-03-21 09:26:39,419 INFO [train.py:901] (0/2) Epoch 39, batch 1800, loss[loss=0.1378, simple_loss=0.2224, pruned_loss=0.02658, over 6757.00 frames. ], tot_loss[loss=0.13, simple_loss=0.2108, pruned_loss=0.02454, over 1436134.65 frames. ], batch size: 107, lr: 4.22e-03, grad_scale: 8.0 +2023-03-21 09:26:51,057 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 09:26:51,640 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=109136.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:26:52,390 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.44 vs. limit=2.0 +2023-03-21 09:26:53,727 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1285, 2.4229, 1.9357, 3.1477, 2.7333, 3.1734, 2.6213, 2.7592], + device='cuda:0'), covar=tensor([0.2252, 0.1320, 0.3990, 0.0706, 0.0272, 0.0304, 0.0387, 0.0407], + device='cuda:0'), in_proj_covar=tensor([0.0250, 0.0229, 0.0247, 0.0258, 0.0198, 0.0195, 0.0217, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 09:26:57,055 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=109147.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:26:58,455 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.276e+02 1.788e+02 2.056e+02 2.387e+02 1.604e+03, threshold=4.112e+02, percent-clipped=1.0 +2023-03-21 09:27:04,217 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 09:27:05,205 INFO [train.py:901] (0/2) Epoch 39, batch 1850, loss[loss=0.09832, simple_loss=0.1633, pruned_loss=0.01666, over 6362.00 frames. ], tot_loss[loss=0.1301, simple_loss=0.2111, pruned_loss=0.02452, over 1437810.79 frames. ], batch size: 28, lr: 4.22e-03, grad_scale: 8.0 +2023-03-21 09:27:11,084 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.19 vs. limit=5.0 +2023-03-21 09:27:14,713 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 09:27:21,532 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=109194.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:27:31,333 INFO [train.py:901] (0/2) Epoch 39, batch 1900, loss[loss=0.1337, simple_loss=0.2208, pruned_loss=0.02326, over 7324.00 frames. ], tot_loss[loss=0.1299, simple_loss=0.211, pruned_loss=0.02437, over 1439224.53 frames. ], batch size: 75, lr: 4.22e-03, grad_scale: 8.0 +2023-03-21 09:27:32,338 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 09:27:42,177 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=109233.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:27:50,591 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.101e+02 1.796e+02 1.987e+02 2.391e+02 7.670e+02, threshold=3.974e+02, percent-clipped=1.0 +2023-03-21 09:27:53,400 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=109255.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:27:57,257 INFO [train.py:901] (0/2) Epoch 39, batch 1950, loss[loss=0.1417, simple_loss=0.2165, pruned_loss=0.03341, over 7326.00 frames. ], tot_loss[loss=0.1301, simple_loss=0.2115, pruned_loss=0.02437, over 1441400.94 frames. ], batch size: 75, lr: 4.22e-03, grad_scale: 8.0 +2023-03-21 09:27:57,773 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 09:28:03,027 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8115, 2.9446, 2.7783, 2.9744, 2.9008, 2.6058, 2.9690, 2.7525], + device='cuda:0'), covar=tensor([0.0840, 0.0585, 0.0889, 0.0949, 0.0782, 0.0939, 0.0464, 0.0887], + device='cuda:0'), in_proj_covar=tensor([0.0059, 0.0060, 0.0068, 0.0060, 0.0058, 0.0063, 0.0057, 0.0054], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:28:06,979 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=109281.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:28:08,945 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 09:28:13,383 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 09:28:13,526 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=109294.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:28:13,892 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 09:28:21,662 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6091, 3.9149, 4.1502, 4.2215, 4.1572, 3.9515, 4.3941, 3.8262], + device='cuda:0'), covar=tensor([0.0159, 0.0210, 0.0132, 0.0172, 0.0454, 0.0164, 0.0153, 0.0214], + device='cuda:0'), in_proj_covar=tensor([0.0096, 0.0103, 0.0103, 0.0089, 0.0179, 0.0109, 0.0106, 0.0115], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:28:23,530 INFO [train.py:901] (0/2) Epoch 39, batch 2000, loss[loss=0.1252, simple_loss=0.2127, pruned_loss=0.01888, over 7341.00 frames. ], tot_loss[loss=0.1306, simple_loss=0.2121, pruned_loss=0.02458, over 1442234.70 frames. ], batch size: 54, lr: 4.22e-03, grad_scale: 8.0 +2023-03-21 09:28:26,179 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=109318.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:28:30,051 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 09:28:40,073 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 09:28:41,142 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8877, 4.3982, 4.3026, 4.8342, 4.6927, 4.7577, 4.3008, 4.4192], + device='cuda:0'), covar=tensor([0.0732, 0.2185, 0.2025, 0.0853, 0.0799, 0.1064, 0.0710, 0.1079], + device='cuda:0'), in_proj_covar=tensor([0.0258, 0.0387, 0.0294, 0.0308, 0.0224, 0.0363, 0.0227, 0.0272], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:28:42,038 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.094e+02 1.673e+02 2.043e+02 2.333e+02 6.605e+02, threshold=4.086e+02, percent-clipped=1.0 +2023-03-21 09:28:45,163 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.85 vs. limit=2.0 +2023-03-21 09:28:45,387 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=109355.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:28:48,733 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 09:28:49,200 INFO [train.py:901] (0/2) Epoch 39, batch 2050, loss[loss=0.1271, simple_loss=0.1922, pruned_loss=0.03104, over 7004.00 frames. ], tot_loss[loss=0.1306, simple_loss=0.2121, pruned_loss=0.02459, over 1441409.39 frames. ], batch size: 35, lr: 4.22e-03, grad_scale: 8.0 +2023-03-21 09:28:57,399 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=109379.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 09:29:04,847 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=109393.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 09:29:11,461 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.45 vs. limit=2.0 +2023-03-21 09:29:15,136 INFO [train.py:901] (0/2) Epoch 39, batch 2100, loss[loss=0.147, simple_loss=0.2337, pruned_loss=0.03013, over 6692.00 frames. ], tot_loss[loss=0.1302, simple_loss=0.2113, pruned_loss=0.0245, over 1439184.45 frames. ], batch size: 106, lr: 4.22e-03, grad_scale: 8.0 +2023-03-21 09:29:22,150 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 09:29:24,723 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5950, 3.9657, 4.1244, 4.2379, 4.1649, 4.0118, 4.3770, 3.8317], + device='cuda:0'), covar=tensor([0.0159, 0.0225, 0.0156, 0.0150, 0.0433, 0.0138, 0.0145, 0.0209], + device='cuda:0'), in_proj_covar=tensor([0.0096, 0.0103, 0.0104, 0.0090, 0.0179, 0.0109, 0.0106, 0.0115], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:29:25,596 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 09:29:32,965 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=109447.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:29:34,358 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.266e+02 1.657e+02 1.969e+02 2.192e+02 3.456e+02, threshold=3.938e+02, percent-clipped=0.0 +2023-03-21 09:29:35,541 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3631, 4.5254, 4.3281, 4.5563, 4.1812, 4.5495, 4.8150, 4.8558], + device='cuda:0'), covar=tensor([0.0195, 0.0158, 0.0186, 0.0138, 0.0306, 0.0251, 0.0202, 0.0163], + device='cuda:0'), in_proj_covar=tensor([0.0123, 0.0128, 0.0122, 0.0126, 0.0115, 0.0103, 0.0100, 0.0102], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:29:41,188 INFO [train.py:901] (0/2) Epoch 39, batch 2150, loss[loss=0.1371, simple_loss=0.2267, pruned_loss=0.02375, over 6702.00 frames. ], tot_loss[loss=0.1304, simple_loss=0.2115, pruned_loss=0.02464, over 1440367.61 frames. ], batch size: 106, lr: 4.22e-03, grad_scale: 8.0 +2023-03-21 09:29:58,028 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=109495.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:29:58,555 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1503, 3.7939, 3.7429, 3.8406, 3.7382, 3.5956, 3.9528, 3.5914], + device='cuda:0'), covar=tensor([0.0125, 0.0175, 0.0138, 0.0172, 0.0450, 0.0139, 0.0167, 0.0188], + device='cuda:0'), in_proj_covar=tensor([0.0097, 0.0103, 0.0104, 0.0091, 0.0181, 0.0110, 0.0106, 0.0116], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:30:00,118 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5870, 2.5366, 2.5741, 3.8633, 1.9686, 3.5912, 1.6371, 3.4563], + device='cuda:0'), covar=tensor([0.0173, 0.1374, 0.1630, 0.0219, 0.3610, 0.0319, 0.1153, 0.0441], + device='cuda:0'), in_proj_covar=tensor([0.0204, 0.0243, 0.0259, 0.0206, 0.0249, 0.0215, 0.0224, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 09:30:07,008 INFO [train.py:901] (0/2) Epoch 39, batch 2200, loss[loss=0.1252, simple_loss=0.2031, pruned_loss=0.02368, over 7265.00 frames. ], tot_loss[loss=0.1304, simple_loss=0.2116, pruned_loss=0.02454, over 1441185.62 frames. ], batch size: 47, lr: 4.22e-03, grad_scale: 8.0 +2023-03-21 09:30:11,544 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 09:30:26,278 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.154e+02 1.848e+02 2.118e+02 2.463e+02 5.500e+02, threshold=4.236e+02, percent-clipped=2.0 +2023-03-21 09:30:26,382 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=109550.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:30:33,270 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.60 vs. limit=2.0 +2023-03-21 09:30:33,365 INFO [train.py:901] (0/2) Epoch 39, batch 2250, loss[loss=0.1212, simple_loss=0.2088, pruned_loss=0.01681, over 7296.00 frames. ], tot_loss[loss=0.1309, simple_loss=0.2124, pruned_loss=0.02472, over 1443778.30 frames. ], batch size: 80, lr: 4.22e-03, grad_scale: 8.0 +2023-03-21 09:30:41,444 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=109579.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:30:42,442 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2240, 2.2227, 2.4567, 2.1124, 2.4744, 2.3483, 1.8908, 1.7333], + device='cuda:0'), covar=tensor([0.0444, 0.0547, 0.0286, 0.0282, 0.0333, 0.0503, 0.0419, 0.0330], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0040, 0.0040, 0.0039, 0.0036, 0.0037, 0.0042, 0.0042], + device='cuda:0'), out_proj_covar=tensor([1.0065e-04, 1.0197e-04, 1.0056e-04, 9.8433e-05, 9.6244e-05, 9.6449e-05, + 1.0485e-04, 1.0535e-04], device='cuda:0') +2023-03-21 09:30:46,650 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 09:30:46,662 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 09:30:56,942 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.76 vs. limit=5.0 +2023-03-21 09:30:58,620 INFO [train.py:901] (0/2) Epoch 39, batch 2300, loss[loss=0.1452, simple_loss=0.2241, pruned_loss=0.03317, over 7317.00 frames. ], tot_loss[loss=0.1312, simple_loss=0.2123, pruned_loss=0.02505, over 1442080.61 frames. ], batch size: 49, lr: 4.22e-03, grad_scale: 8.0 +2023-03-21 09:30:58,644 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 09:31:07,364 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4239, 4.9325, 4.9970, 4.9803, 4.8670, 4.3996, 5.0461, 4.8852], + device='cuda:0'), covar=tensor([0.0488, 0.0427, 0.0422, 0.0455, 0.0346, 0.0482, 0.0344, 0.0451], + device='cuda:0'), in_proj_covar=tensor([0.0220, 0.0261, 0.0203, 0.0201, 0.0162, 0.0231, 0.0211, 0.0149], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:31:13,047 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=109640.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:31:18,544 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.188e+02 1.786e+02 2.016e+02 2.497e+02 7.057e+02, threshold=4.032e+02, percent-clipped=1.0 +2023-03-21 09:31:18,642 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=109650.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:31:23,135 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=109659.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:31:25,034 INFO [train.py:901] (0/2) Epoch 39, batch 2350, loss[loss=0.1299, simple_loss=0.2091, pruned_loss=0.02536, over 7314.00 frames. ], tot_loss[loss=0.1307, simple_loss=0.212, pruned_loss=0.0247, over 1441117.37 frames. ], batch size: 49, lr: 4.21e-03, grad_scale: 8.0 +2023-03-21 09:31:25,689 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5164, 1.7697, 1.6140, 1.6603, 1.8388, 1.6939, 1.7114, 1.3225], + device='cuda:0'), covar=tensor([0.0162, 0.0173, 0.0216, 0.0186, 0.0123, 0.0133, 0.0126, 0.0217], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0036, 0.0035, 0.0036, 0.0035, 0.0033, 0.0037, 0.0045], + device='cuda:0'), out_proj_covar=tensor([4.1862e-05, 4.0267e-05, 3.9613e-05, 4.0135e-05, 3.9127e-05, 3.7249e-05, + 4.1622e-05, 4.9559e-05], device='cuda:0') +2023-03-21 09:31:30,659 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=109674.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 09:31:40,231 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=109693.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 09:31:47,030 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 09:31:50,626 INFO [train.py:901] (0/2) Epoch 39, batch 2400, loss[loss=0.1461, simple_loss=0.2344, pruned_loss=0.02895, over 7363.00 frames. ], tot_loss[loss=0.1308, simple_loss=0.2121, pruned_loss=0.02477, over 1444250.73 frames. ], batch size: 73, lr: 4.21e-03, grad_scale: 8.0 +2023-03-21 09:31:51,030 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.62 vs. limit=2.0 +2023-03-21 09:31:53,140 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 09:31:54,225 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=109720.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:31:59,957 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2163, 4.7085, 4.7351, 4.7118, 4.6814, 4.2838, 4.8048, 4.7060], + device='cuda:0'), covar=tensor([0.0482, 0.0361, 0.0405, 0.0461, 0.0279, 0.0380, 0.0312, 0.0344], + device='cuda:0'), in_proj_covar=tensor([0.0218, 0.0258, 0.0201, 0.0199, 0.0161, 0.0229, 0.0209, 0.0148], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:32:03,832 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 09:32:05,352 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=109741.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 09:32:06,298 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 09:32:09,799 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.181e+02 1.635e+02 2.013e+02 2.308e+02 3.868e+02, threshold=4.027e+02, percent-clipped=0.0 +2023-03-21 09:32:16,318 INFO [train.py:901] (0/2) Epoch 39, batch 2450, loss[loss=0.1436, simple_loss=0.2227, pruned_loss=0.03224, over 7288.00 frames. ], tot_loss[loss=0.1309, simple_loss=0.2122, pruned_loss=0.02483, over 1442889.96 frames. ], batch size: 68, lr: 4.21e-03, grad_scale: 8.0 +2023-03-21 09:32:32,791 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 09:32:42,311 INFO [train.py:901] (0/2) Epoch 39, batch 2500, loss[loss=0.1312, simple_loss=0.2093, pruned_loss=0.02655, over 7262.00 frames. ], tot_loss[loss=0.1312, simple_loss=0.2127, pruned_loss=0.02489, over 1442619.46 frames. ], batch size: 52, lr: 4.21e-03, grad_scale: 8.0 +2023-03-21 09:32:58,338 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 09:32:59,889 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7586, 2.8615, 3.7842, 3.6860, 3.8010, 3.7371, 3.8225, 3.5043], + device='cuda:0'), covar=tensor([0.0047, 0.0189, 0.0048, 0.0048, 0.0046, 0.0055, 0.0046, 0.0079], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0068, 0.0056, 0.0055, 0.0054, 0.0060, 0.0047, 0.0076], + device='cuda:0'), out_proj_covar=tensor([7.9888e-05, 1.3678e-04, 1.0223e-04, 9.5135e-05, 9.1680e-05, 1.0483e-04, + 8.9361e-05, 1.4135e-04], device='cuda:0') +2023-03-21 09:33:01,308 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.139e+02 1.761e+02 2.069e+02 2.509e+02 3.938e+02, threshold=4.137e+02, percent-clipped=0.0 +2023-03-21 09:33:01,443 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=109850.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:33:07,858 INFO [train.py:901] (0/2) Epoch 39, batch 2550, loss[loss=0.1382, simple_loss=0.231, pruned_loss=0.02272, over 6689.00 frames. ], tot_loss[loss=0.1314, simple_loss=0.213, pruned_loss=0.02485, over 1444023.86 frames. ], batch size: 106, lr: 4.21e-03, grad_scale: 16.0 +2023-03-21 09:33:18,705 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0814, 2.7318, 3.3549, 3.1816, 3.2576, 3.1148, 2.7309, 3.0872], + device='cuda:0'), covar=tensor([0.1234, 0.0776, 0.0753, 0.0916, 0.0696, 0.0882, 0.1467, 0.1264], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0068, 0.0051, 0.0050, 0.0049, 0.0048, 0.0067, 0.0051], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 09:33:26,172 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=109898.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:33:34,177 INFO [train.py:901] (0/2) Epoch 39, batch 2600, loss[loss=0.1127, simple_loss=0.1939, pruned_loss=0.01571, over 7149.00 frames. ], tot_loss[loss=0.1317, simple_loss=0.2133, pruned_loss=0.02499, over 1444790.92 frames. ], batch size: 41, lr: 4.21e-03, grad_scale: 8.0 +2023-03-21 09:33:39,177 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2190, 4.6516, 4.6821, 4.6414, 4.6327, 4.2287, 4.7365, 4.6000], + device='cuda:0'), covar=tensor([0.0497, 0.0375, 0.0350, 0.0492, 0.0288, 0.0425, 0.0284, 0.0369], + device='cuda:0'), in_proj_covar=tensor([0.0218, 0.0259, 0.0201, 0.0200, 0.0161, 0.0231, 0.0210, 0.0149], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:33:45,221 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=109935.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:33:52,583 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=109950.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:33:52,979 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.178e+02 1.771e+02 2.022e+02 2.327e+02 3.789e+02, threshold=4.044e+02, percent-clipped=0.0 +2023-03-21 09:33:58,881 INFO [train.py:901] (0/2) Epoch 39, batch 2650, loss[loss=0.1447, simple_loss=0.2193, pruned_loss=0.03504, over 7234.00 frames. ], tot_loss[loss=0.1316, simple_loss=0.2133, pruned_loss=0.02489, over 1446271.22 frames. ], batch size: 45, lr: 4.21e-03, grad_scale: 8.0 +2023-03-21 09:34:04,485 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=109974.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:34:09,965 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5638, 1.9809, 1.8431, 1.9249, 2.0228, 1.7850, 1.7007, 1.5217], + device='cuda:0'), covar=tensor([0.0213, 0.0212, 0.0210, 0.0198, 0.0126, 0.0160, 0.0180, 0.0207], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0037, 0.0035, 0.0037, 0.0036, 0.0034, 0.0038, 0.0045], + device='cuda:0'), out_proj_covar=tensor([4.2297e-05, 4.0714e-05, 3.9861e-05, 4.0787e-05, 3.9949e-05, 3.7767e-05, + 4.2239e-05, 4.9999e-05], device='cuda:0') +2023-03-21 09:34:16,345 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=109998.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:34:23,964 INFO [train.py:901] (0/2) Epoch 39, batch 2700, loss[loss=0.1284, simple_loss=0.2165, pruned_loss=0.02015, over 7322.00 frames. ], tot_loss[loss=0.1307, simple_loss=0.2125, pruned_loss=0.02449, over 1445397.26 frames. ], batch size: 80, lr: 4.21e-03, grad_scale: 8.0 +2023-03-21 09:34:25,032 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110015.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:34:26,098 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=110017.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:34:28,519 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=110022.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:34:39,330 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=110044.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:34:42,664 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.192e+02 1.798e+02 2.201e+02 2.521e+02 4.025e+02, threshold=4.402e+02, percent-clipped=0.0 +2023-03-21 09:34:48,451 INFO [train.py:901] (0/2) Epoch 39, batch 2750, loss[loss=0.1339, simple_loss=0.2206, pruned_loss=0.02362, over 7256.00 frames. ], tot_loss[loss=0.1307, simple_loss=0.2124, pruned_loss=0.02446, over 1443029.56 frames. ], batch size: 52, lr: 4.21e-03, grad_scale: 8.0 +2023-03-21 09:34:53,714 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.66 vs. limit=2.0 +2023-03-21 09:34:55,969 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=110078.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:35:09,615 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=110105.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:35:11,122 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7655, 1.5999, 1.8711, 2.2742, 1.9881, 2.0074, 1.6137, 2.0630], + device='cuda:0'), covar=tensor([0.2338, 0.3681, 0.2654, 0.1566, 0.3739, 0.1189, 0.1786, 0.2726], + device='cuda:0'), in_proj_covar=tensor([0.0079, 0.0080, 0.0072, 0.0064, 0.0065, 0.0064, 0.0107, 0.0068], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:35:12,056 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=110110.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:35:13,409 INFO [train.py:901] (0/2) Epoch 39, batch 2800, loss[loss=0.1222, simple_loss=0.2074, pruned_loss=0.01855, over 7264.00 frames. ], tot_loss[loss=0.1305, simple_loss=0.2124, pruned_loss=0.0243, over 1444467.06 frames. ], batch size: 52, lr: 4.21e-03, grad_scale: 8.0 +2023-03-21 09:35:14,494 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=110115.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:35:15,871 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3983, 2.5803, 2.7015, 2.2480, 2.5623, 2.4788, 2.2462, 2.0893], + device='cuda:0'), covar=tensor([0.0613, 0.0636, 0.0389, 0.0347, 0.0615, 0.0597, 0.0337, 0.0356], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0038, 0.0039, 0.0038, 0.0036, 0.0037, 0.0041, 0.0041], + device='cuda:0'), out_proj_covar=tensor([9.9419e-05, 9.9545e-05, 9.8888e-05, 9.6762e-05, 9.4919e-05, 9.5218e-05, + 1.0255e-04, 1.0418e-04], device='cuda:0') +2023-03-21 09:35:25,851 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-39.pt +2023-03-21 09:35:41,175 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 09:35:41,236 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 09:35:41,262 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 09:35:41,279 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 09:35:41,459 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 09:35:41,475 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 09:35:41,486 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 09:35:41,585 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 09:35:41,666 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 09:35:41,681 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 09:35:41,698 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 09:35:42,232 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 09:35:42,239 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 09:35:42,246 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 09:35:42,421 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 09:35:42,425 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 09:35:42,510 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 09:35:45,302 INFO [train.py:901] (0/2) Epoch 40, batch 0, loss[loss=0.1312, simple_loss=0.2205, pruned_loss=0.02097, over 7320.00 frames. ], tot_loss[loss=0.1312, simple_loss=0.2205, pruned_loss=0.02097, over 7320.00 frames. ], batch size: 80, lr: 4.15e-03, grad_scale: 8.0 +2023-03-21 09:35:45,304 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 09:35:53,416 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5432, 3.0599, 3.4706, 3.3672, 3.1070, 3.0380, 3.4535, 2.4980], + device='cuda:0'), covar=tensor([0.0405, 0.0398, 0.0601, 0.0616, 0.0632, 0.0884, 0.0497, 0.2472], + device='cuda:0'), in_proj_covar=tensor([0.0332, 0.0336, 0.0273, 0.0357, 0.0289, 0.0288, 0.0348, 0.0248], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:35:54,169 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0390, 3.0937, 2.3471, 3.5231, 2.3855, 2.8797, 1.6098, 2.5260], + device='cuda:0'), covar=tensor([0.0504, 0.0779, 0.2873, 0.0676, 0.0488, 0.0653, 0.4176, 0.1737], + device='cuda:0'), in_proj_covar=tensor([0.0279, 0.0258, 0.0283, 0.0270, 0.0271, 0.0264, 0.0234, 0.0260], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 09:36:11,283 INFO [train.py:935] (0/2) Epoch 40, validation: loss=0.1653, simple_loss=0.2572, pruned_loss=0.03673, over 1622729.00 frames. +2023-03-21 09:36:11,284 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 09:36:18,218 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 09:36:18,537 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.38 vs. limit=5.0 +2023-03-21 09:36:18,681 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.279e+02 1.747e+02 2.118e+02 2.577e+02 4.675e+02, threshold=4.235e+02, percent-clipped=1.0 +2023-03-21 09:36:25,091 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.78 vs. limit=2.0 +2023-03-21 09:36:28,754 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 09:36:28,881 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8657, 3.0002, 2.7906, 3.0350, 3.0297, 2.6779, 3.1093, 2.8558], + device='cuda:0'), covar=tensor([0.0766, 0.0710, 0.0786, 0.0988, 0.0840, 0.0580, 0.0739, 0.1024], + device='cuda:0'), in_proj_covar=tensor([0.0059, 0.0060, 0.0068, 0.0060, 0.0058, 0.0062, 0.0057, 0.0054], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:36:28,901 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=110171.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:36:31,381 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=110176.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:36:36,119 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 09:36:36,614 INFO [train.py:901] (0/2) Epoch 40, batch 50, loss[loss=0.1319, simple_loss=0.2119, pruned_loss=0.02598, over 7346.00 frames. ], tot_loss[loss=0.131, simple_loss=0.2125, pruned_loss=0.02474, over 326229.25 frames. ], batch size: 51, lr: 4.15e-03, grad_scale: 8.0 +2023-03-21 09:36:36,726 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3656, 3.5118, 3.4377, 3.4172, 3.3241, 3.3049, 3.7191, 3.7194], + device='cuda:0'), covar=tensor([0.0261, 0.0196, 0.0231, 0.0235, 0.0334, 0.0690, 0.0255, 0.0213], + device='cuda:0'), in_proj_covar=tensor([0.0125, 0.0129, 0.0123, 0.0127, 0.0117, 0.0104, 0.0101, 0.0104], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:36:38,647 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 09:36:41,624 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 09:36:52,668 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.68 vs. limit=2.0 +2023-03-21 09:36:58,140 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9618, 3.1675, 3.9807, 4.0113, 3.9320, 3.9272, 4.1061, 3.9077], + device='cuda:0'), covar=tensor([0.0033, 0.0139, 0.0033, 0.0029, 0.0033, 0.0037, 0.0035, 0.0053], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0069, 0.0057, 0.0056, 0.0055, 0.0061, 0.0048, 0.0078], + device='cuda:0'), out_proj_covar=tensor([8.0803e-05, 1.3966e-04, 1.0405e-04, 9.6395e-05, 9.3415e-05, 1.0637e-04, + 9.1408e-05, 1.4399e-04], device='cuda:0') +2023-03-21 09:36:59,063 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 09:36:59,510 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 09:37:02,085 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=110235.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:37:02,972 INFO [train.py:901] (0/2) Epoch 40, batch 100, loss[loss=0.1444, simple_loss=0.2274, pruned_loss=0.0307, over 7304.00 frames. ], tot_loss[loss=0.1302, simple_loss=0.211, pruned_loss=0.02469, over 574717.68 frames. ], batch size: 80, lr: 4.15e-03, grad_scale: 8.0 +2023-03-21 09:37:10,702 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.125e+02 1.749e+02 2.051e+02 2.351e+02 3.808e+02, threshold=4.103e+02, percent-clipped=0.0 +2023-03-21 09:37:14,064 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.60 vs. limit=2.0 +2023-03-21 09:37:26,982 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=110283.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:37:27,046 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=110283.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:37:28,933 INFO [train.py:901] (0/2) Epoch 40, batch 150, loss[loss=0.1141, simple_loss=0.1931, pruned_loss=0.0176, over 6939.00 frames. ], tot_loss[loss=0.1302, simple_loss=0.2113, pruned_loss=0.02454, over 766502.47 frames. ], batch size: 35, lr: 4.15e-03, grad_scale: 8.0 +2023-03-21 09:37:31,057 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9898, 3.2115, 3.9684, 3.9984, 3.9868, 4.0960, 4.1371, 3.9524], + device='cuda:0'), covar=tensor([0.0033, 0.0127, 0.0031, 0.0034, 0.0032, 0.0029, 0.0036, 0.0051], + device='cuda:0'), in_proj_covar=tensor([0.0047, 0.0069, 0.0057, 0.0056, 0.0054, 0.0060, 0.0048, 0.0077], + device='cuda:0'), out_proj_covar=tensor([8.0078e-05, 1.3848e-04, 1.0317e-04, 9.5666e-05, 9.2732e-05, 1.0519e-04, + 9.0438e-05, 1.4262e-04], device='cuda:0') +2023-03-21 09:37:37,090 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=110303.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:37:43,742 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=110315.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:37:45,330 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=110318.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:37:55,290 INFO [train.py:901] (0/2) Epoch 40, batch 200, loss[loss=0.141, simple_loss=0.2159, pruned_loss=0.03308, over 7316.00 frames. ], tot_loss[loss=0.1308, simple_loss=0.2118, pruned_loss=0.02487, over 915731.09 frames. ], batch size: 59, lr: 4.15e-03, grad_scale: 8.0 +2023-03-21 09:37:59,002 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=110344.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:38:00,931 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 09:38:02,479 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.033e+02 1.699e+02 2.011e+02 2.332e+02 3.842e+02, threshold=4.021e+02, percent-clipped=0.0 +2023-03-21 09:38:05,531 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 09:38:08,553 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=110363.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:38:09,130 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=110364.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:38:10,979 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 09:38:13,523 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110373.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:38:15,110 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0071, 3.7012, 3.5950, 3.6963, 3.6286, 3.4806, 3.7930, 3.4008], + device='cuda:0'), covar=tensor([0.0121, 0.0174, 0.0132, 0.0202, 0.0413, 0.0131, 0.0159, 0.0200], + device='cuda:0'), in_proj_covar=tensor([0.0098, 0.0104, 0.0105, 0.0092, 0.0182, 0.0111, 0.0107, 0.0116], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:38:16,679 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=110379.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:38:18,884 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.74 vs. limit=5.0 +2023-03-21 09:38:20,501 INFO [train.py:901] (0/2) Epoch 40, batch 250, loss[loss=0.1229, simple_loss=0.1971, pruned_loss=0.02434, over 6976.00 frames. ], tot_loss[loss=0.1304, simple_loss=0.2114, pruned_loss=0.02469, over 1032994.07 frames. ], batch size: 35, lr: 4.15e-03, grad_scale: 8.0 +2023-03-21 09:38:23,008 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 09:38:28,096 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110400.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:38:42,259 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.81 vs. limit=2.0 +2023-03-21 09:38:45,448 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 09:38:46,959 INFO [train.py:901] (0/2) Epoch 40, batch 300, loss[loss=0.1349, simple_loss=0.2237, pruned_loss=0.02303, over 7300.00 frames. ], tot_loss[loss=0.1298, simple_loss=0.2105, pruned_loss=0.02452, over 1121999.12 frames. ], batch size: 59, lr: 4.15e-03, grad_scale: 8.0 +2023-03-21 09:38:49,978 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9606, 2.4386, 2.6900, 4.0035, 1.8981, 3.6922, 1.4774, 3.5047], + device='cuda:0'), covar=tensor([0.0175, 0.1501, 0.1620, 0.0189, 0.4343, 0.0267, 0.1513, 0.0467], + device='cuda:0'), in_proj_covar=tensor([0.0205, 0.0244, 0.0261, 0.0206, 0.0250, 0.0215, 0.0225, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 09:38:53,832 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.290e+02 1.777e+02 2.048e+02 2.389e+02 4.929e+02, threshold=4.095e+02, percent-clipped=1.0 +2023-03-21 09:38:53,873 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 09:39:01,345 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110466.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:39:03,916 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110471.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:39:06,759 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.87 vs. limit=2.0 +2023-03-21 09:39:12,427 INFO [train.py:901] (0/2) Epoch 40, batch 350, loss[loss=0.1321, simple_loss=0.2233, pruned_loss=0.02047, over 6803.00 frames. ], tot_loss[loss=0.1297, simple_loss=0.2104, pruned_loss=0.02453, over 1192182.01 frames. ], batch size: 106, lr: 4.15e-03, grad_scale: 8.0 +2023-03-21 09:39:28,528 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 09:39:37,958 INFO [train.py:901] (0/2) Epoch 40, batch 400, loss[loss=0.1265, simple_loss=0.2115, pruned_loss=0.02081, over 7303.00 frames. ], tot_loss[loss=0.1296, simple_loss=0.2105, pruned_loss=0.02436, over 1247652.10 frames. ], batch size: 80, lr: 4.14e-03, grad_scale: 8.0 +2023-03-21 09:39:41,164 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4775, 1.8340, 1.4800, 1.5738, 1.7654, 1.6311, 1.7309, 1.4534], + device='cuda:0'), covar=tensor([0.0203, 0.0176, 0.0383, 0.0194, 0.0149, 0.0167, 0.0131, 0.0188], + device='cuda:0'), in_proj_covar=tensor([0.0037, 0.0036, 0.0035, 0.0037, 0.0036, 0.0034, 0.0037, 0.0045], + device='cuda:0'), out_proj_covar=tensor([4.2201e-05, 4.0432e-05, 3.9648e-05, 4.1298e-05, 3.9770e-05, 3.7508e-05, + 4.2127e-05, 5.0122e-05], device='cuda:0') +2023-03-21 09:39:45,024 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.224e+02 1.773e+02 2.034e+02 2.439e+02 3.776e+02, threshold=4.069e+02, percent-clipped=0.0 +2023-03-21 09:39:58,728 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6759, 2.8663, 3.6172, 3.7104, 3.6530, 3.7529, 3.6297, 3.5546], + device='cuda:0'), covar=tensor([0.0031, 0.0131, 0.0035, 0.0028, 0.0035, 0.0029, 0.0043, 0.0056], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0069, 0.0057, 0.0056, 0.0055, 0.0060, 0.0048, 0.0077], + device='cuda:0'), out_proj_covar=tensor([8.0253e-05, 1.3945e-04, 1.0341e-04, 9.5569e-05, 9.2956e-05, 1.0499e-04, + 8.9807e-05, 1.4228e-04], device='cuda:0') +2023-03-21 09:40:04,296 INFO [train.py:901] (0/2) Epoch 40, batch 450, loss[loss=0.1149, simple_loss=0.1944, pruned_loss=0.0177, over 7332.00 frames. ], tot_loss[loss=0.1294, simple_loss=0.2105, pruned_loss=0.02422, over 1290858.07 frames. ], batch size: 44, lr: 4.14e-03, grad_scale: 8.0 +2023-03-21 09:40:08,868 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 09:40:09,894 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 09:40:16,870 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=110611.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:40:29,884 INFO [train.py:901] (0/2) Epoch 40, batch 500, loss[loss=0.1085, simple_loss=0.1975, pruned_loss=0.009733, over 7304.00 frames. ], tot_loss[loss=0.1291, simple_loss=0.2102, pruned_loss=0.02399, over 1322044.52 frames. ], batch size: 86, lr: 4.14e-03, grad_scale: 8.0 +2023-03-21 09:40:30,977 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110639.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:40:36,901 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.326e+02 1.687e+02 1.978e+02 2.331e+02 4.016e+02, threshold=3.957e+02, percent-clipped=0.0 +2023-03-21 09:40:41,558 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110659.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:40:43,079 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 09:40:44,614 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 09:40:45,142 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 09:40:47,140 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 09:40:48,778 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=110672.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:40:49,245 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=110673.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:40:49,716 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110674.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:40:51,695 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 09:40:56,162 INFO [train.py:901] (0/2) Epoch 40, batch 550, loss[loss=0.14, simple_loss=0.2207, pruned_loss=0.02965, over 7343.00 frames. ], tot_loss[loss=0.1292, simple_loss=0.2104, pruned_loss=0.02401, over 1348935.80 frames. ], batch size: 61, lr: 4.14e-03, grad_scale: 8.0 +2023-03-21 09:41:02,694 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 09:41:02,763 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=110700.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:41:10,793 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8122, 2.4456, 2.9788, 2.8618, 2.8558, 2.6306, 2.5204, 2.7753], + device='cuda:0'), covar=tensor([0.1426, 0.0684, 0.0966, 0.1132, 0.1023, 0.1381, 0.1777, 0.1215], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0068, 0.0051, 0.0051, 0.0050, 0.0049, 0.0068, 0.0051], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 09:41:11,165 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 09:41:13,193 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=110721.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:41:14,142 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 09:41:14,721 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1783, 4.4335, 4.1351, 4.4316, 4.0170, 4.4060, 4.7190, 4.7360], + device='cuda:0'), covar=tensor([0.0183, 0.0131, 0.0231, 0.0143, 0.0374, 0.0259, 0.0199, 0.0168], + device='cuda:0'), in_proj_covar=tensor([0.0124, 0.0129, 0.0123, 0.0126, 0.0117, 0.0104, 0.0101, 0.0103], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:41:20,857 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=110736.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:41:21,224 INFO [train.py:901] (0/2) Epoch 40, batch 600, loss[loss=0.1189, simple_loss=0.2064, pruned_loss=0.01566, over 7271.00 frames. ], tot_loss[loss=0.1293, simple_loss=0.2105, pruned_loss=0.02407, over 1370050.94 frames. ], batch size: 70, lr: 4.14e-03, grad_scale: 8.0 +2023-03-21 09:41:21,251 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 09:41:27,437 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=110748.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:41:29,435 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.347e+02 1.717e+02 1.997e+02 2.356e+02 3.346e+02, threshold=3.993e+02, percent-clipped=0.0 +2023-03-21 09:41:37,222 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=110766.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:41:38,147 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 09:41:39,727 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=110771.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:41:46,651 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 09:41:47,610 INFO [train.py:901] (0/2) Epoch 40, batch 650, loss[loss=0.1408, simple_loss=0.2189, pruned_loss=0.0314, over 7211.00 frames. ], tot_loss[loss=0.1297, simple_loss=0.211, pruned_loss=0.02423, over 1385903.70 frames. ], batch size: 50, lr: 4.14e-03, grad_scale: 8.0 +2023-03-21 09:41:52,688 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=110797.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 09:42:01,429 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=110814.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:42:02,980 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6216, 4.1271, 3.9095, 4.6581, 4.3876, 4.5009, 4.0647, 4.1064], + device='cuda:0'), covar=tensor([0.0894, 0.2561, 0.2582, 0.0980, 0.1005, 0.1257, 0.0875, 0.1457], + device='cuda:0'), in_proj_covar=tensor([0.0261, 0.0395, 0.0299, 0.0306, 0.0230, 0.0368, 0.0229, 0.0276], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:42:03,967 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=110819.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:42:04,404 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 09:42:09,211 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3446, 2.5563, 2.3436, 2.5797, 2.4527, 2.1857, 2.6740, 2.4470], + device='cuda:0'), covar=tensor([0.0777, 0.0530, 0.0886, 0.0690, 0.0719, 0.0800, 0.0666, 0.0933], + device='cuda:0'), in_proj_covar=tensor([0.0058, 0.0059, 0.0067, 0.0059, 0.0057, 0.0062, 0.0056, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:42:13,064 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 09:42:14,125 INFO [train.py:901] (0/2) Epoch 40, batch 700, loss[loss=0.1289, simple_loss=0.2099, pruned_loss=0.02388, over 7301.00 frames. ], tot_loss[loss=0.1295, simple_loss=0.2109, pruned_loss=0.0241, over 1400244.18 frames. ], batch size: 83, lr: 4.14e-03, grad_scale: 8.0 +2023-03-21 09:42:21,053 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.283e+02 1.664e+02 1.895e+02 2.126e+02 3.944e+02, threshold=3.791e+02, percent-clipped=0.0 +2023-03-21 09:42:29,770 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=110868.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:42:35,289 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2099, 2.3057, 2.4875, 2.1637, 2.4164, 2.2139, 2.0111, 1.7886], + device='cuda:0'), covar=tensor([0.0533, 0.0407, 0.0272, 0.0282, 0.0636, 0.0428, 0.0364, 0.0378], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0038, 0.0040, 0.0038, 0.0036, 0.0037, 0.0041, 0.0041], + device='cuda:0'), out_proj_covar=tensor([9.9619e-05, 9.9698e-05, 9.9318e-05, 9.7210e-05, 9.5400e-05, 9.5274e-05, + 1.0329e-04, 1.0432e-04], device='cuda:0') +2023-03-21 09:42:38,148 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 09:42:38,660 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 09:42:39,152 INFO [train.py:901] (0/2) Epoch 40, batch 750, loss[loss=0.1312, simple_loss=0.2107, pruned_loss=0.02587, over 7232.00 frames. ], tot_loss[loss=0.1294, simple_loss=0.2105, pruned_loss=0.02414, over 1407495.99 frames. ], batch size: 45, lr: 4.14e-03, grad_scale: 8.0 +2023-03-21 09:42:45,234 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0503, 2.7623, 3.1336, 3.0481, 2.8937, 2.7744, 3.0714, 2.3114], + device='cuda:0'), covar=tensor([0.0505, 0.0518, 0.0684, 0.0638, 0.0610, 0.0920, 0.0483, 0.2401], + device='cuda:0'), in_proj_covar=tensor([0.0326, 0.0330, 0.0268, 0.0349, 0.0284, 0.0281, 0.0341, 0.0242], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:42:51,520 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 09:42:56,676 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 09:42:58,543 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.03 vs. limit=5.0 +2023-03-21 09:43:01,386 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=110929.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:43:02,784 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 09:43:04,278 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 09:43:05,223 INFO [train.py:901] (0/2) Epoch 40, batch 800, loss[loss=0.1408, simple_loss=0.2251, pruned_loss=0.02825, over 7256.00 frames. ], tot_loss[loss=0.1299, simple_loss=0.2112, pruned_loss=0.02435, over 1412917.25 frames. ], batch size: 64, lr: 4.14e-03, grad_scale: 8.0 +2023-03-21 09:43:06,302 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=110939.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:43:12,148 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.148e+02 1.823e+02 2.168e+02 2.507e+02 3.727e+02, threshold=4.336e+02, percent-clipped=0.0 +2023-03-21 09:43:14,243 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 09:43:14,908 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2635, 2.2449, 2.2743, 3.4466, 1.8574, 3.1743, 1.3916, 3.1469], + device='cuda:0'), covar=tensor([0.0266, 0.1584, 0.2170, 0.0287, 0.4494, 0.0394, 0.1478, 0.0544], + device='cuda:0'), in_proj_covar=tensor([0.0205, 0.0244, 0.0259, 0.0205, 0.0249, 0.0215, 0.0223, 0.0225], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 09:43:15,918 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=110958.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:43:16,370 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=110959.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:43:20,327 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=110967.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:43:23,952 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=110974.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:43:30,315 INFO [train.py:901] (0/2) Epoch 40, batch 850, loss[loss=0.1144, simple_loss=0.198, pruned_loss=0.01542, over 7329.00 frames. ], tot_loss[loss=0.1298, simple_loss=0.2113, pruned_loss=0.02421, over 1420639.91 frames. ], batch size: 61, lr: 4.14e-03, grad_scale: 8.0 +2023-03-21 09:43:30,363 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=110987.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:43:33,294 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 09:43:33,299 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 09:43:39,726 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 09:43:41,248 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=111007.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:43:43,255 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 09:43:47,947 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=111019.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:43:49,376 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=111022.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:43:56,762 INFO [train.py:901] (0/2) Epoch 40, batch 900, loss[loss=0.1427, simple_loss=0.2205, pruned_loss=0.03242, over 7318.00 frames. ], tot_loss[loss=0.1304, simple_loss=0.2118, pruned_loss=0.02451, over 1426652.26 frames. ], batch size: 61, lr: 4.14e-03, grad_scale: 8.0 +2023-03-21 09:44:03,880 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.196e+02 1.783e+02 2.026e+02 2.389e+02 5.093e+02, threshold=4.052e+02, percent-clipped=1.0 +2023-03-21 09:44:17,775 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7836, 3.1306, 2.7855, 3.1183, 3.0876, 2.6191, 3.1571, 2.9672], + device='cuda:0'), covar=tensor([0.0951, 0.0905, 0.0987, 0.1074, 0.1153, 0.1091, 0.0818, 0.1195], + device='cuda:0'), in_proj_covar=tensor([0.0058, 0.0059, 0.0067, 0.0059, 0.0058, 0.0062, 0.0057, 0.0054], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:44:20,207 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 09:44:22,199 INFO [train.py:901] (0/2) Epoch 40, batch 950, loss[loss=0.129, simple_loss=0.2154, pruned_loss=0.02127, over 7275.00 frames. ], tot_loss[loss=0.1299, simple_loss=0.2115, pruned_loss=0.02416, over 1432057.01 frames. ], batch size: 77, lr: 4.13e-03, grad_scale: 8.0 +2023-03-21 09:44:25,416 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=111092.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 09:44:43,305 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6639, 2.3711, 2.4781, 3.7391, 1.9663, 3.5292, 1.4000, 3.3623], + device='cuda:0'), covar=tensor([0.0204, 0.1535, 0.1804, 0.0183, 0.3990, 0.0265, 0.1283, 0.0449], + device='cuda:0'), in_proj_covar=tensor([0.0205, 0.0244, 0.0259, 0.0204, 0.0248, 0.0215, 0.0223, 0.0225], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 09:44:45,207 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 09:44:48,699 INFO [train.py:901] (0/2) Epoch 40, batch 1000, loss[loss=0.1349, simple_loss=0.2228, pruned_loss=0.02356, over 7311.00 frames. ], tot_loss[loss=0.1291, simple_loss=0.2107, pruned_loss=0.02381, over 1435563.83 frames. ], batch size: 59, lr: 4.13e-03, grad_scale: 8.0 +2023-03-21 09:44:55,752 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.230e+02 1.691e+02 1.890e+02 2.255e+02 3.563e+02, threshold=3.780e+02, percent-clipped=0.0 +2023-03-21 09:45:05,715 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 09:45:14,411 INFO [train.py:901] (0/2) Epoch 40, batch 1050, loss[loss=0.1348, simple_loss=0.2192, pruned_loss=0.02517, over 7276.00 frames. ], tot_loss[loss=0.1293, simple_loss=0.2108, pruned_loss=0.02393, over 1438078.01 frames. ], batch size: 68, lr: 4.13e-03, grad_scale: 8.0 +2023-03-21 09:45:29,312 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 09:45:29,668 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-21 09:45:33,422 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 09:45:33,991 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=111224.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:45:34,539 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=111225.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:45:35,519 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7649, 3.8733, 3.6231, 3.8584, 3.5720, 3.6060, 4.0111, 4.0680], + device='cuda:0'), covar=tensor([0.0277, 0.0189, 0.0318, 0.0216, 0.0367, 0.0483, 0.0336, 0.0277], + device='cuda:0'), in_proj_covar=tensor([0.0123, 0.0130, 0.0123, 0.0126, 0.0116, 0.0103, 0.0100, 0.0103], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:45:36,052 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7228, 1.5836, 1.9972, 2.2376, 2.0811, 2.1532, 1.7758, 2.2228], + device='cuda:0'), covar=tensor([0.3911, 0.3496, 0.2182, 0.1559, 0.1277, 0.1816, 0.2244, 0.1354], + device='cuda:0'), in_proj_covar=tensor([0.0078, 0.0078, 0.0070, 0.0063, 0.0064, 0.0063, 0.0104, 0.0068], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:45:40,522 INFO [train.py:901] (0/2) Epoch 40, batch 1100, loss[loss=0.13, simple_loss=0.2106, pruned_loss=0.02474, over 7322.00 frames. ], tot_loss[loss=0.1294, simple_loss=0.2107, pruned_loss=0.02405, over 1438049.43 frames. ], batch size: 59, lr: 4.13e-03, grad_scale: 8.0 +2023-03-21 09:45:45,124 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9176, 3.2216, 3.8794, 4.0098, 4.0254, 3.9859, 4.0174, 3.8686], + device='cuda:0'), covar=tensor([0.0036, 0.0117, 0.0032, 0.0028, 0.0029, 0.0028, 0.0040, 0.0049], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0070, 0.0058, 0.0056, 0.0055, 0.0061, 0.0049, 0.0078], + device='cuda:0'), out_proj_covar=tensor([8.1012e-05, 1.4139e-04, 1.0411e-04, 9.6156e-05, 9.4257e-05, 1.0594e-04, + 9.1952e-05, 1.4347e-04], device='cuda:0') +2023-03-21 09:45:47,522 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.323e+02 1.688e+02 1.957e+02 2.307e+02 4.408e+02, threshold=3.914e+02, percent-clipped=3.0 +2023-03-21 09:45:56,426 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=111267.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:46:02,023 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 09:46:02,035 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 09:46:06,571 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=111286.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:46:06,949 INFO [train.py:901] (0/2) Epoch 40, batch 1150, loss[loss=0.1286, simple_loss=0.2044, pruned_loss=0.02638, over 7310.00 frames. ], tot_loss[loss=0.1295, simple_loss=0.2107, pruned_loss=0.02416, over 1440972.79 frames. ], batch size: 49, lr: 4.13e-03, grad_scale: 8.0 +2023-03-21 09:46:14,996 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 09:46:15,506 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 09:46:17,758 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.60 vs. limit=5.0 +2023-03-21 09:46:20,507 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=111314.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:46:20,990 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=111315.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:46:29,373 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.72 vs. limit=2.0 +2023-03-21 09:46:32,018 INFO [train.py:901] (0/2) Epoch 40, batch 1200, loss[loss=0.1412, simple_loss=0.2243, pruned_loss=0.02907, over 7284.00 frames. ], tot_loss[loss=0.13, simple_loss=0.2114, pruned_loss=0.02437, over 1442093.11 frames. ], batch size: 68, lr: 4.13e-03, grad_scale: 8.0 +2023-03-21 09:46:39,642 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.301e+02 1.737e+02 2.081e+02 2.449e+02 8.398e+02, threshold=4.161e+02, percent-clipped=1.0 +2023-03-21 09:46:47,662 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 09:46:50,767 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9660, 4.4212, 4.2447, 4.8986, 4.7077, 4.8630, 4.3832, 4.4980], + device='cuda:0'), covar=tensor([0.0766, 0.2714, 0.2414, 0.1072, 0.0954, 0.1211, 0.0672, 0.1237], + device='cuda:0'), in_proj_covar=tensor([0.0258, 0.0390, 0.0293, 0.0307, 0.0227, 0.0365, 0.0228, 0.0274], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:46:58,233 INFO [train.py:901] (0/2) Epoch 40, batch 1250, loss[loss=0.1304, simple_loss=0.2099, pruned_loss=0.02539, over 7306.00 frames. ], tot_loss[loss=0.1298, simple_loss=0.2111, pruned_loss=0.02425, over 1442030.17 frames. ], batch size: 49, lr: 4.13e-03, grad_scale: 8.0 +2023-03-21 09:46:58,854 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6468, 2.8150, 3.6454, 3.7529, 3.7431, 3.7852, 3.7956, 3.6034], + device='cuda:0'), covar=tensor([0.0044, 0.0171, 0.0044, 0.0044, 0.0039, 0.0037, 0.0048, 0.0058], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0070, 0.0058, 0.0056, 0.0055, 0.0061, 0.0049, 0.0078], + device='cuda:0'), out_proj_covar=tensor([8.1406e-05, 1.4154e-04, 1.0455e-04, 9.6684e-05, 9.4688e-05, 1.0650e-04, + 9.2350e-05, 1.4432e-04], device='cuda:0') +2023-03-21 09:47:00,838 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=111392.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 09:47:11,214 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 09:47:15,162 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 09:47:16,195 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 09:47:17,351 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9060, 3.2003, 2.7619, 3.0623, 3.1362, 2.7450, 3.0159, 2.8492], + device='cuda:0'), covar=tensor([0.0577, 0.0746, 0.1126, 0.1138, 0.0968, 0.0674, 0.1245, 0.1558], + device='cuda:0'), in_proj_covar=tensor([0.0059, 0.0059, 0.0068, 0.0060, 0.0058, 0.0063, 0.0057, 0.0054], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:47:24,391 INFO [train.py:901] (0/2) Epoch 40, batch 1300, loss[loss=0.1388, simple_loss=0.2132, pruned_loss=0.03215, over 7211.00 frames. ], tot_loss[loss=0.1301, simple_loss=0.2115, pruned_loss=0.02432, over 1442638.74 frames. ], batch size: 50, lr: 4.13e-03, grad_scale: 8.0 +2023-03-21 09:47:26,555 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=111440.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:47:30,800 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.87 vs. limit=5.0 +2023-03-21 09:47:31,943 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.074e+02 1.753e+02 2.042e+02 2.404e+02 3.175e+02, threshold=4.084e+02, percent-clipped=0.0 +2023-03-21 09:47:39,676 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 09:47:40,834 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=111468.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:47:41,742 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 09:47:45,346 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 09:47:50,251 INFO [train.py:901] (0/2) Epoch 40, batch 1350, loss[loss=0.1377, simple_loss=0.2203, pruned_loss=0.02756, over 7343.00 frames. ], tot_loss[loss=0.1304, simple_loss=0.2119, pruned_loss=0.02444, over 1441637.67 frames. ], batch size: 63, lr: 4.13e-03, grad_scale: 8.0 +2023-03-21 09:47:54,142 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-03-21 09:47:54,278 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 09:48:01,389 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4030, 3.3188, 2.7065, 3.8536, 2.9694, 3.3767, 1.7849, 2.6816], + device='cuda:0'), covar=tensor([0.0460, 0.0701, 0.2108, 0.0453, 0.0455, 0.0563, 0.3794, 0.1637], + device='cuda:0'), in_proj_covar=tensor([0.0276, 0.0256, 0.0279, 0.0268, 0.0269, 0.0264, 0.0232, 0.0257], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 09:48:09,996 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=111524.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:48:12,527 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=111529.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:48:16,297 INFO [train.py:901] (0/2) Epoch 40, batch 1400, loss[loss=0.1423, simple_loss=0.221, pruned_loss=0.03176, over 7327.00 frames. ], tot_loss[loss=0.131, simple_loss=0.2121, pruned_loss=0.02489, over 1441308.32 frames. ], batch size: 61, lr: 4.13e-03, grad_scale: 8.0 +2023-03-21 09:48:23,420 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.146e+02 1.668e+02 1.991e+02 2.540e+02 5.106e+02, threshold=3.982e+02, percent-clipped=1.0 +2023-03-21 09:48:27,083 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 09:48:34,167 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=111572.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:48:38,706 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1817, 4.6849, 4.7504, 4.7135, 4.6795, 4.2112, 4.7650, 4.5970], + device='cuda:0'), covar=tensor([0.0543, 0.0386, 0.0334, 0.0447, 0.0330, 0.0428, 0.0321, 0.0465], + device='cuda:0'), in_proj_covar=tensor([0.0220, 0.0259, 0.0202, 0.0201, 0.0160, 0.0231, 0.0211, 0.0148], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:48:38,710 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=111581.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:48:41,625 INFO [train.py:901] (0/2) Epoch 40, batch 1450, loss[loss=0.1284, simple_loss=0.2149, pruned_loss=0.02093, over 7327.00 frames. ], tot_loss[loss=0.131, simple_loss=0.212, pruned_loss=0.02497, over 1441255.92 frames. ], batch size: 75, lr: 4.13e-03, grad_scale: 8.0 +2023-03-21 09:48:47,302 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4949, 3.1679, 3.4697, 3.5356, 3.3135, 3.2254, 3.6615, 2.6810], + device='cuda:0'), covar=tensor([0.0498, 0.0566, 0.0671, 0.0650, 0.0694, 0.1104, 0.0593, 0.2156], + device='cuda:0'), in_proj_covar=tensor([0.0329, 0.0333, 0.0272, 0.0355, 0.0285, 0.0285, 0.0344, 0.0243], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:48:51,035 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 09:48:56,872 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=111614.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:48:57,084 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-03-21 09:49:07,258 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 09:49:08,254 INFO [train.py:901] (0/2) Epoch 40, batch 1500, loss[loss=0.1211, simple_loss=0.2031, pruned_loss=0.01954, over 7260.00 frames. ], tot_loss[loss=0.1309, simple_loss=0.2122, pruned_loss=0.02483, over 1442533.93 frames. ], batch size: 89, lr: 4.12e-03, grad_scale: 8.0 +2023-03-21 09:49:15,116 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.165e+02 1.685e+02 1.982e+02 2.275e+02 4.883e+02, threshold=3.965e+02, percent-clipped=1.0 +2023-03-21 09:49:20,687 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=111662.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:49:28,529 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.84 vs. limit=5.0 +2023-03-21 09:49:30,616 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 09:49:33,131 INFO [train.py:901] (0/2) Epoch 40, batch 1550, loss[loss=0.1339, simple_loss=0.2197, pruned_loss=0.0241, over 7215.00 frames. ], tot_loss[loss=0.1303, simple_loss=0.2116, pruned_loss=0.0245, over 1444737.09 frames. ], batch size: 93, lr: 4.12e-03, grad_scale: 8.0 +2023-03-21 09:49:59,469 INFO [train.py:901] (0/2) Epoch 40, batch 1600, loss[loss=0.09932, simple_loss=0.1713, pruned_loss=0.01368, over 6983.00 frames. ], tot_loss[loss=0.1298, simple_loss=0.2111, pruned_loss=0.02425, over 1444938.13 frames. ], batch size: 35, lr: 4.12e-03, grad_scale: 8.0 +2023-03-21 09:50:02,520 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 09:50:03,036 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 09:50:06,059 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 09:50:06,509 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.192e+02 1.612e+02 1.846e+02 2.181e+02 5.378e+02, threshold=3.692e+02, percent-clipped=1.0 +2023-03-21 09:50:15,101 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 09:50:19,126 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 09:50:23,904 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1861, 4.6935, 4.7581, 4.6843, 4.6673, 4.2398, 4.7940, 4.5888], + device='cuda:0'), covar=tensor([0.0560, 0.0399, 0.0349, 0.0539, 0.0318, 0.0428, 0.0312, 0.0422], + device='cuda:0'), in_proj_covar=tensor([0.0219, 0.0258, 0.0201, 0.0201, 0.0159, 0.0229, 0.0210, 0.0148], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:50:25,967 INFO [train.py:901] (0/2) Epoch 40, batch 1650, loss[loss=0.1285, simple_loss=0.2128, pruned_loss=0.02207, over 7262.00 frames. ], tot_loss[loss=0.1301, simple_loss=0.2114, pruned_loss=0.0244, over 1443478.67 frames. ], batch size: 89, lr: 4.12e-03, grad_scale: 8.0 +2023-03-21 09:50:28,090 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=111791.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 09:50:28,500 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 09:50:28,594 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1290, 3.5785, 4.1112, 3.9926, 4.0709, 4.1335, 4.2277, 3.9885], + device='cuda:0'), covar=tensor([0.0029, 0.0093, 0.0027, 0.0027, 0.0028, 0.0028, 0.0028, 0.0047], + device='cuda:0'), in_proj_covar=tensor([0.0048, 0.0070, 0.0058, 0.0056, 0.0055, 0.0061, 0.0049, 0.0078], + device='cuda:0'), out_proj_covar=tensor([8.1289e-05, 1.4147e-04, 1.0399e-04, 9.6628e-05, 9.3913e-05, 1.0544e-04, + 9.1563e-05, 1.4346e-04], device='cuda:0') +2023-03-21 09:50:41,909 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9573, 2.8669, 2.1346, 3.1994, 2.3331, 2.8276, 1.4427, 2.0919], + device='cuda:0'), covar=tensor([0.0629, 0.0914, 0.3055, 0.0877, 0.0624, 0.0661, 0.4037, 0.1999], + device='cuda:0'), in_proj_covar=tensor([0.0277, 0.0256, 0.0281, 0.0269, 0.0271, 0.0266, 0.0233, 0.0258], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 09:50:44,738 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=111824.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:50:45,658 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 09:50:49,598 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 09:50:51,071 INFO [train.py:901] (0/2) Epoch 40, batch 1700, loss[loss=0.1307, simple_loss=0.2133, pruned_loss=0.024, over 7307.00 frames. ], tot_loss[loss=0.1299, simple_loss=0.2112, pruned_loss=0.02429, over 1443828.61 frames. ], batch size: 80, lr: 4.12e-03, grad_scale: 8.0 +2023-03-21 09:50:55,591 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7153, 5.2944, 5.3653, 5.3196, 5.1079, 4.8655, 5.3826, 5.1955], + device='cuda:0'), covar=tensor([0.0447, 0.0317, 0.0283, 0.0422, 0.0321, 0.0371, 0.0293, 0.0392], + device='cuda:0'), in_proj_covar=tensor([0.0219, 0.0258, 0.0202, 0.0201, 0.0160, 0.0229, 0.0212, 0.0148], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:50:58,018 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.116e+02 1.630e+02 2.034e+02 2.442e+02 3.825e+02, threshold=4.068e+02, percent-clipped=2.0 +2023-03-21 09:50:58,612 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=111852.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 09:50:59,953 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 09:51:12,289 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2387, 2.6090, 1.9759, 3.2561, 2.8608, 3.3360, 2.7737, 2.9193], + device='cuda:0'), covar=tensor([0.2311, 0.1185, 0.4000, 0.0669, 0.0353, 0.0278, 0.0443, 0.0414], + device='cuda:0'), in_proj_covar=tensor([0.0248, 0.0229, 0.0244, 0.0257, 0.0198, 0.0198, 0.0217, 0.0222], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 09:51:13,135 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4155, 4.9762, 5.0702, 5.0119, 4.8447, 4.4803, 5.0571, 4.8500], + device='cuda:0'), covar=tensor([0.0504, 0.0363, 0.0312, 0.0441, 0.0339, 0.0406, 0.0345, 0.0455], + device='cuda:0'), in_proj_covar=tensor([0.0220, 0.0260, 0.0203, 0.0202, 0.0161, 0.0230, 0.0213, 0.0148], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:51:14,167 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=111881.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:51:17,113 INFO [train.py:901] (0/2) Epoch 40, batch 1750, loss[loss=0.119, simple_loss=0.2073, pruned_loss=0.01535, over 7298.00 frames. ], tot_loss[loss=0.1295, simple_loss=0.2109, pruned_loss=0.02401, over 1443769.07 frames. ], batch size: 68, lr: 4.12e-03, grad_scale: 16.0 +2023-03-21 09:51:25,212 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 09:51:26,234 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 09:51:38,281 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=111929.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:51:42,239 INFO [train.py:901] (0/2) Epoch 40, batch 1800, loss[loss=0.1248, simple_loss=0.2082, pruned_loss=0.02074, over 7283.00 frames. ], tot_loss[loss=0.1298, simple_loss=0.2114, pruned_loss=0.0241, over 1442658.17 frames. ], batch size: 68, lr: 4.12e-03, grad_scale: 16.0 +2023-03-21 09:51:47,687 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 09:51:49,140 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.360e+02 1.770e+02 2.065e+02 2.371e+02 4.234e+02, threshold=4.130e+02, percent-clipped=1.0 +2023-03-21 09:52:02,334 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 09:52:08,229 INFO [train.py:901] (0/2) Epoch 40, batch 1850, loss[loss=0.1323, simple_loss=0.2135, pruned_loss=0.02558, over 7290.00 frames. ], tot_loss[loss=0.1292, simple_loss=0.2106, pruned_loss=0.02392, over 1441301.19 frames. ], batch size: 77, lr: 4.12e-03, grad_scale: 8.0 +2023-03-21 09:52:12,738 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 09:52:15,079 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-112000.pt +2023-03-21 09:52:23,479 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.86 vs. limit=2.0 +2023-03-21 09:52:33,214 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 09:52:37,753 INFO [train.py:901] (0/2) Epoch 40, batch 1900, loss[loss=0.1472, simple_loss=0.2348, pruned_loss=0.0298, over 7362.00 frames. ], tot_loss[loss=0.1296, simple_loss=0.2107, pruned_loss=0.02423, over 1440659.22 frames. ], batch size: 73, lr: 4.12e-03, grad_scale: 8.0 +2023-03-21 09:52:45,233 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.221e+02 1.755e+02 2.099e+02 2.615e+02 5.453e+02, threshold=4.198e+02, percent-clipped=1.0 +2023-03-21 09:52:52,982 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=112066.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:52:57,913 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 09:52:58,520 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3464, 2.3931, 2.7444, 2.1581, 2.4739, 2.4786, 2.1789, 1.9644], + device='cuda:0'), covar=tensor([0.0568, 0.0541, 0.0323, 0.0331, 0.0490, 0.0446, 0.0410, 0.0270], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0039, 0.0040, 0.0039, 0.0037, 0.0038, 0.0043, 0.0042], + device='cuda:0'), out_proj_covar=tensor([1.0251e-04, 1.0170e-04, 1.0132e-04, 9.9273e-05, 9.7913e-05, 9.7583e-05, + 1.0614e-04, 1.0660e-04], device='cuda:0') +2023-03-21 09:53:03,468 INFO [train.py:901] (0/2) Epoch 40, batch 1950, loss[loss=0.133, simple_loss=0.2189, pruned_loss=0.02355, over 7143.00 frames. ], tot_loss[loss=0.1298, simple_loss=0.211, pruned_loss=0.02431, over 1442305.94 frames. ], batch size: 98, lr: 4.12e-03, grad_scale: 8.0 +2023-03-21 09:53:09,346 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 09:53:13,852 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 09:53:14,811 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 09:53:21,975 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=112124.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:53:23,498 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=112127.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:53:28,393 INFO [train.py:901] (0/2) Epoch 40, batch 2000, loss[loss=0.1248, simple_loss=0.2034, pruned_loss=0.02308, over 7259.00 frames. ], tot_loss[loss=0.1304, simple_loss=0.2117, pruned_loss=0.02451, over 1443416.40 frames. ], batch size: 55, lr: 4.11e-03, grad_scale: 8.0 +2023-03-21 09:53:30,838 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 09:53:34,578 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=112147.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 09:53:36,961 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.013e+02 1.664e+02 1.933e+02 2.332e+02 4.499e+02, threshold=3.866e+02, percent-clipped=2.0 +2023-03-21 09:53:40,058 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5224, 4.0695, 3.8695, 4.5193, 4.3052, 4.3980, 3.7242, 4.0448], + device='cuda:0'), covar=tensor([0.0832, 0.2526, 0.2571, 0.1044, 0.0921, 0.1180, 0.0949, 0.1369], + device='cuda:0'), in_proj_covar=tensor([0.0260, 0.0389, 0.0292, 0.0304, 0.0226, 0.0364, 0.0229, 0.0273], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:53:43,103 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 09:53:47,173 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=112172.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:53:51,147 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 09:53:54,639 INFO [train.py:901] (0/2) Epoch 40, batch 2050, loss[loss=0.1504, simple_loss=0.2177, pruned_loss=0.04156, over 7324.00 frames. ], tot_loss[loss=0.1301, simple_loss=0.211, pruned_loss=0.0246, over 1442768.09 frames. ], batch size: 49, lr: 4.11e-03, grad_scale: 8.0 +2023-03-21 09:53:55,862 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6753, 3.4249, 3.5322, 3.6728, 3.1596, 3.1860, 3.8079, 2.6485], + device='cuda:0'), covar=tensor([0.0448, 0.0587, 0.0644, 0.0702, 0.0731, 0.1004, 0.0690, 0.2289], + device='cuda:0'), in_proj_covar=tensor([0.0330, 0.0336, 0.0273, 0.0355, 0.0286, 0.0286, 0.0347, 0.0244], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:54:21,541 INFO [train.py:901] (0/2) Epoch 40, batch 2100, loss[loss=0.124, simple_loss=0.2057, pruned_loss=0.02118, over 7330.00 frames. ], tot_loss[loss=0.1302, simple_loss=0.2112, pruned_loss=0.0246, over 1442964.14 frames. ], batch size: 44, lr: 4.11e-03, grad_scale: 8.0 +2023-03-21 09:54:24,968 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 09:54:27,971 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 09:54:28,982 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.301e+02 1.776e+02 2.099e+02 2.556e+02 5.854e+02, threshold=4.198e+02, percent-clipped=3.0 +2023-03-21 09:54:31,663 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7004, 3.6528, 2.8395, 4.1060, 3.4384, 3.7241, 1.9737, 2.8949], + device='cuda:0'), covar=tensor([0.0484, 0.0804, 0.2280, 0.0479, 0.0476, 0.0618, 0.3562, 0.1729], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0256, 0.0279, 0.0267, 0.0271, 0.0264, 0.0233, 0.0256], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 09:54:46,387 INFO [train.py:901] (0/2) Epoch 40, batch 2150, loss[loss=0.1177, simple_loss=0.2005, pruned_loss=0.01738, over 7312.00 frames. ], tot_loss[loss=0.1303, simple_loss=0.2111, pruned_loss=0.02473, over 1442937.48 frames. ], batch size: 75, lr: 4.11e-03, grad_scale: 8.0 +2023-03-21 09:55:07,798 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0164, 2.5435, 3.0681, 2.9331, 3.0622, 2.7987, 2.6392, 2.7870], + device='cuda:0'), covar=tensor([0.1076, 0.0672, 0.1091, 0.1211, 0.0762, 0.1089, 0.1738, 0.2072], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0069, 0.0052, 0.0052, 0.0051, 0.0051, 0.0070, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 09:55:12,690 INFO [train.py:901] (0/2) Epoch 40, batch 2200, loss[loss=0.1397, simple_loss=0.2224, pruned_loss=0.02844, over 7326.00 frames. ], tot_loss[loss=0.1299, simple_loss=0.2106, pruned_loss=0.02458, over 1441321.24 frames. ], batch size: 83, lr: 4.11e-03, grad_scale: 8.0 +2023-03-21 09:55:13,222 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 09:55:20,093 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.190e+02 1.737e+02 2.025e+02 2.418e+02 3.757e+02, threshold=4.050e+02, percent-clipped=0.0 +2023-03-21 09:55:28,772 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4258, 3.2493, 3.4605, 3.4502, 3.2864, 3.1027, 3.5897, 2.5617], + device='cuda:0'), covar=tensor([0.0541, 0.0607, 0.0696, 0.0719, 0.0739, 0.1017, 0.0775, 0.2445], + device='cuda:0'), in_proj_covar=tensor([0.0331, 0.0336, 0.0273, 0.0356, 0.0285, 0.0286, 0.0349, 0.0244], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:55:30,445 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-03-21 09:55:37,761 INFO [train.py:901] (0/2) Epoch 40, batch 2250, loss[loss=0.1346, simple_loss=0.2177, pruned_loss=0.0258, over 7126.00 frames. ], tot_loss[loss=0.1305, simple_loss=0.2114, pruned_loss=0.02483, over 1441925.37 frames. ], batch size: 98, lr: 4.11e-03, grad_scale: 8.0 +2023-03-21 09:55:38,960 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=112389.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:55:45,823 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 09:55:46,887 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 09:55:51,094 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3882, 3.1841, 3.3635, 3.4544, 3.2786, 3.1536, 3.6293, 2.6090], + device='cuda:0'), covar=tensor([0.0477, 0.0601, 0.0721, 0.0614, 0.0712, 0.1007, 0.0669, 0.2300], + device='cuda:0'), in_proj_covar=tensor([0.0330, 0.0335, 0.0272, 0.0355, 0.0284, 0.0285, 0.0348, 0.0244], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:55:57,005 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=112422.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:55:58,927 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 09:56:04,556 INFO [train.py:901] (0/2) Epoch 40, batch 2300, loss[loss=0.1478, simple_loss=0.2262, pruned_loss=0.03469, over 7253.00 frames. ], tot_loss[loss=0.1306, simple_loss=0.2116, pruned_loss=0.02484, over 1439768.76 frames. ], batch size: 52, lr: 4.11e-03, grad_scale: 8.0 +2023-03-21 09:56:09,702 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=112447.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 09:56:09,907 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-21 09:56:11,287 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=112450.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:56:12,106 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.304e+02 1.811e+02 2.116e+02 2.493e+02 3.897e+02, threshold=4.232e+02, percent-clipped=0.0 +2023-03-21 09:56:17,362 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9581, 2.7630, 2.0515, 3.0290, 2.4567, 2.8580, 1.4157, 2.1870], + device='cuda:0'), covar=tensor([0.0835, 0.1374, 0.3223, 0.1029, 0.0607, 0.0742, 0.3837, 0.1881], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0255, 0.0279, 0.0265, 0.0271, 0.0265, 0.0232, 0.0255], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 09:56:26,815 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6396, 1.9420, 1.7995, 1.8322, 2.0839, 1.6537, 1.6537, 1.4990], + device='cuda:0'), covar=tensor([0.0202, 0.0288, 0.0286, 0.0285, 0.0187, 0.0305, 0.0306, 0.0253], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0037, 0.0036, 0.0038, 0.0036, 0.0034, 0.0037, 0.0045], + device='cuda:0'), out_proj_covar=tensor([4.3161e-05, 4.0647e-05, 4.0209e-05, 4.1395e-05, 4.0076e-05, 3.7799e-05, + 4.1998e-05, 5.0111e-05], device='cuda:0') +2023-03-21 09:56:29,544 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-03-21 09:56:29,678 INFO [train.py:901] (0/2) Epoch 40, batch 2350, loss[loss=0.1311, simple_loss=0.212, pruned_loss=0.02507, over 7281.00 frames. ], tot_loss[loss=0.1299, simple_loss=0.2108, pruned_loss=0.02449, over 1440730.32 frames. ], batch size: 77, lr: 4.11e-03, grad_scale: 8.0 +2023-03-21 09:56:33,598 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.21 vs. limit=2.0 +2023-03-21 09:56:34,401 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=112495.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 09:56:41,418 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-03-21 09:56:46,022 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 09:56:52,550 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 09:56:55,507 INFO [train.py:901] (0/2) Epoch 40, batch 2400, loss[loss=0.1647, simple_loss=0.2373, pruned_loss=0.04603, over 6615.00 frames. ], tot_loss[loss=0.1299, simple_loss=0.2111, pruned_loss=0.02437, over 1441981.69 frames. ], batch size: 106, lr: 4.11e-03, grad_scale: 8.0 +2023-03-21 09:57:02,550 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 09:57:03,014 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.085e+02 1.709e+02 1.972e+02 2.399e+02 3.691e+02, threshold=3.945e+02, percent-clipped=0.0 +2023-03-21 09:57:05,550 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 09:57:08,156 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=112562.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:57:15,510 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=112577.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:57:21,720 INFO [train.py:901] (0/2) Epoch 40, batch 2450, loss[loss=0.1208, simple_loss=0.2067, pruned_loss=0.01741, over 7229.00 frames. ], tot_loss[loss=0.1294, simple_loss=0.2106, pruned_loss=0.02407, over 1442383.75 frames. ], batch size: 93, lr: 4.11e-03, grad_scale: 8.0 +2023-03-21 09:57:34,621 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 09:57:37,187 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.1087, 4.6011, 4.4279, 5.0715, 4.8758, 4.9761, 4.3769, 4.6840], + device='cuda:0'), covar=tensor([0.0834, 0.2612, 0.2518, 0.0921, 0.0887, 0.1147, 0.0903, 0.1088], + device='cuda:0'), in_proj_covar=tensor([0.0262, 0.0394, 0.0294, 0.0307, 0.0228, 0.0369, 0.0231, 0.0275], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:57:37,256 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=112617.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:57:40,372 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=112623.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:57:47,167 INFO [train.py:901] (0/2) Epoch 40, batch 2500, loss[loss=0.1317, simple_loss=0.2116, pruned_loss=0.02588, over 7304.00 frames. ], tot_loss[loss=0.1292, simple_loss=0.2103, pruned_loss=0.02409, over 1443364.90 frames. ], batch size: 83, lr: 4.11e-03, grad_scale: 8.0 +2023-03-21 09:57:47,836 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=112638.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 09:57:54,736 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.154e+02 1.621e+02 1.878e+02 2.077e+02 3.587e+02, threshold=3.756e+02, percent-clipped=0.0 +2023-03-21 09:57:58,850 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.4513, 4.9433, 4.8542, 5.4394, 5.2847, 5.3877, 4.8225, 5.0218], + device='cuda:0'), covar=tensor([0.0736, 0.2491, 0.2260, 0.0890, 0.0810, 0.1077, 0.0777, 0.1089], + device='cuda:0'), in_proj_covar=tensor([0.0263, 0.0396, 0.0295, 0.0307, 0.0228, 0.0370, 0.0232, 0.0276], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:57:59,301 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 09:58:04,723 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1090, 2.2570, 2.3737, 3.4676, 1.8877, 3.2274, 1.4671, 3.1845], + device='cuda:0'), covar=tensor([0.0179, 0.1468, 0.1779, 0.0213, 0.3666, 0.0311, 0.1259, 0.0490], + device='cuda:0'), in_proj_covar=tensor([0.0203, 0.0243, 0.0253, 0.0205, 0.0246, 0.0213, 0.0221, 0.0224], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 09:58:09,163 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=112678.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 09:58:13,472 INFO [train.py:901] (0/2) Epoch 40, batch 2550, loss[loss=0.1274, simple_loss=0.2031, pruned_loss=0.02584, over 7220.00 frames. ], tot_loss[loss=0.1287, simple_loss=0.2097, pruned_loss=0.0239, over 1443670.96 frames. ], batch size: 45, lr: 4.10e-03, grad_scale: 8.0 +2023-03-21 09:58:31,056 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=112722.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:58:38,323 INFO [train.py:901] (0/2) Epoch 40, batch 2600, loss[loss=0.1281, simple_loss=0.2093, pruned_loss=0.02341, over 7277.00 frames. ], tot_loss[loss=0.1283, simple_loss=0.2094, pruned_loss=0.0236, over 1444620.15 frames. ], batch size: 64, lr: 4.10e-03, grad_scale: 8.0 +2023-03-21 09:58:42,290 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=112745.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:58:45,638 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.153e+02 1.606e+02 1.959e+02 2.340e+02 7.874e+02, threshold=3.917e+02, percent-clipped=2.0 +2023-03-21 09:58:54,534 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=112770.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 09:59:02,874 INFO [train.py:901] (0/2) Epoch 40, batch 2650, loss[loss=0.148, simple_loss=0.2274, pruned_loss=0.03429, over 7261.00 frames. ], tot_loss[loss=0.1288, simple_loss=0.21, pruned_loss=0.02381, over 1444812.21 frames. ], batch size: 55, lr: 4.10e-03, grad_scale: 8.0 +2023-03-21 09:59:24,629 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1019, 3.9072, 3.3989, 3.6771, 3.1656, 2.3626, 2.0175, 4.0024], + device='cuda:0'), covar=tensor([0.0046, 0.0068, 0.0129, 0.0072, 0.0174, 0.0535, 0.0614, 0.0054], + device='cuda:0'), in_proj_covar=tensor([0.0115, 0.0095, 0.0112, 0.0094, 0.0130, 0.0134, 0.0128, 0.0106], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 09:59:27,081 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1897, 2.2616, 2.5715, 2.0523, 2.3634, 2.2692, 2.1468, 1.8261], + device='cuda:0'), covar=tensor([0.0477, 0.0517, 0.0261, 0.0317, 0.0410, 0.0521, 0.0371, 0.0349], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0039, 0.0040, 0.0039, 0.0036, 0.0037, 0.0042, 0.0042], + device='cuda:0'), out_proj_covar=tensor([1.0304e-04, 1.0085e-04, 9.9858e-05, 9.8374e-05, 9.6336e-05, 9.6314e-05, + 1.0482e-04, 1.0545e-04], device='cuda:0') +2023-03-21 09:59:28,801 INFO [train.py:901] (0/2) Epoch 40, batch 2700, loss[loss=0.1529, simple_loss=0.2257, pruned_loss=0.04009, over 7275.00 frames. ], tot_loss[loss=0.1288, simple_loss=0.2099, pruned_loss=0.02386, over 1443267.05 frames. ], batch size: 57, lr: 4.10e-03, grad_scale: 8.0 +2023-03-21 09:59:36,114 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.272e+02 1.707e+02 2.016e+02 2.345e+02 3.632e+02, threshold=4.032e+02, percent-clipped=0.0 +2023-03-21 09:59:37,154 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2263, 3.8439, 3.8539, 3.8968, 3.8581, 3.7165, 4.0617, 3.6711], + device='cuda:0'), covar=tensor([0.0185, 0.0201, 0.0143, 0.0200, 0.0496, 0.0146, 0.0171, 0.0187], + device='cuda:0'), in_proj_covar=tensor([0.0102, 0.0108, 0.0108, 0.0095, 0.0186, 0.0113, 0.0109, 0.0119], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 09:59:51,989 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7547, 5.2253, 5.3189, 5.2987, 5.0451, 4.7786, 5.3183, 5.1341], + device='cuda:0'), covar=tensor([0.0471, 0.0352, 0.0373, 0.0409, 0.0330, 0.0361, 0.0315, 0.0424], + device='cuda:0'), in_proj_covar=tensor([0.0220, 0.0260, 0.0202, 0.0204, 0.0162, 0.0230, 0.0214, 0.0149], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 09:59:53,424 INFO [train.py:901] (0/2) Epoch 40, batch 2750, loss[loss=0.1478, simple_loss=0.2327, pruned_loss=0.03139, over 6747.00 frames. ], tot_loss[loss=0.1293, simple_loss=0.2104, pruned_loss=0.02408, over 1441913.48 frames. ], batch size: 107, lr: 4.10e-03, grad_scale: 8.0 +2023-03-21 10:00:05,501 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4198, 1.6185, 1.4149, 1.4679, 1.5974, 1.4075, 1.5415, 1.1990], + device='cuda:0'), covar=tensor([0.0174, 0.0168, 0.0188, 0.0137, 0.0135, 0.0138, 0.0132, 0.0237], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0037, 0.0036, 0.0038, 0.0036, 0.0034, 0.0037, 0.0046], + device='cuda:0'), out_proj_covar=tensor([4.3232e-05, 4.0795e-05, 4.0441e-05, 4.1510e-05, 3.9939e-05, 3.7996e-05, + 4.2147e-05, 5.0287e-05], device='cuda:0') +2023-03-21 10:00:08,762 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=112918.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:00:10,285 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=112921.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:00:16,118 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=112933.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 10:00:17,999 INFO [train.py:901] (0/2) Epoch 40, batch 2800, loss[loss=0.1249, simple_loss=0.2012, pruned_loss=0.02426, over 7235.00 frames. ], tot_loss[loss=0.1294, simple_loss=0.2108, pruned_loss=0.02405, over 1441865.14 frames. ], batch size: 47, lr: 4.10e-03, grad_scale: 8.0 +2023-03-21 10:00:25,333 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.134e+02 1.722e+02 1.974e+02 2.339e+02 4.614e+02, threshold=3.948e+02, percent-clipped=1.0 +2023-03-21 10:00:28,256 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.53 vs. limit=2.0 +2023-03-21 10:00:30,577 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-40.pt +2023-03-21 10:00:45,479 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 10:00:45,539 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 10:00:45,565 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 10:00:45,582 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 10:00:45,761 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 10:00:45,777 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 10:00:45,788 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 10:00:45,888 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 10:00:46,304 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 10:00:46,319 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 10:00:46,340 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 10:00:46,503 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 10:00:46,509 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 10:00:46,517 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 10:00:46,661 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 10:00:46,664 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 10:00:46,723 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 10:00:49,022 INFO [train.py:901] (0/2) Epoch 41, batch 0, loss[loss=0.1313, simple_loss=0.2138, pruned_loss=0.02434, over 7354.00 frames. ], tot_loss[loss=0.1313, simple_loss=0.2138, pruned_loss=0.02434, over 7354.00 frames. ], batch size: 63, lr: 4.05e-03, grad_scale: 8.0 +2023-03-21 10:00:49,023 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 10:00:55,671 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0404, 3.9261, 3.4075, 3.7007, 3.3473, 2.4310, 1.9008, 4.0505], + device='cuda:0'), covar=tensor([0.0062, 0.0082, 0.0130, 0.0072, 0.0147, 0.0636, 0.0766, 0.0057], + device='cuda:0'), in_proj_covar=tensor([0.0116, 0.0096, 0.0113, 0.0094, 0.0131, 0.0135, 0.0129, 0.0107], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 10:01:04,930 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6332, 3.7673, 3.5546, 3.8209, 3.5046, 3.7388, 3.9988, 4.0579], + device='cuda:0'), covar=tensor([0.0213, 0.0173, 0.0222, 0.0149, 0.0318, 0.0221, 0.0207, 0.0138], + device='cuda:0'), in_proj_covar=tensor([0.0124, 0.0129, 0.0123, 0.0125, 0.0115, 0.0102, 0.0100, 0.0104], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 10:01:12,268 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2967, 3.9878, 3.9360, 3.9657, 3.9844, 3.8140, 4.1140, 3.8104], + device='cuda:0'), covar=tensor([0.0123, 0.0154, 0.0133, 0.0178, 0.0524, 0.0138, 0.0167, 0.0176], + device='cuda:0'), in_proj_covar=tensor([0.0101, 0.0107, 0.0108, 0.0095, 0.0186, 0.0113, 0.0109, 0.0119], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 10:01:15,064 INFO [train.py:935] (0/2) Epoch 41, validation: loss=0.1652, simple_loss=0.2574, pruned_loss=0.03654, over 1622729.00 frames. +2023-03-21 10:01:15,065 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 10:01:21,209 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=112973.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 10:01:22,108 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 10:01:25,897 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=112982.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:01:32,681 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 10:01:39,999 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 10:01:40,509 INFO [train.py:901] (0/2) Epoch 41, batch 50, loss[loss=0.1281, simple_loss=0.2117, pruned_loss=0.02221, over 7363.00 frames. ], tot_loss[loss=0.1288, simple_loss=0.2108, pruned_loss=0.02343, over 327926.51 frames. ], batch size: 63, lr: 4.05e-03, grad_scale: 8.0 +2023-03-21 10:01:42,068 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 10:01:44,570 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 10:01:58,879 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=113045.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:02:02,324 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.308e+02 1.789e+02 2.061e+02 2.499e+02 5.277e+02, threshold=4.122e+02, percent-clipped=1.0 +2023-03-21 10:02:02,869 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 10:02:03,354 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 10:02:06,876 INFO [train.py:901] (0/2) Epoch 41, batch 100, loss[loss=0.128, simple_loss=0.2149, pruned_loss=0.02053, over 7279.00 frames. ], tot_loss[loss=0.1296, simple_loss=0.2116, pruned_loss=0.0238, over 576147.82 frames. ], batch size: 52, lr: 4.05e-03, grad_scale: 8.0 +2023-03-21 10:02:22,753 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=113093.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:02:23,065 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.77 vs. limit=2.0 +2023-03-21 10:02:32,459 INFO [train.py:901] (0/2) Epoch 41, batch 150, loss[loss=0.113, simple_loss=0.1927, pruned_loss=0.01662, over 7212.00 frames. ], tot_loss[loss=0.1294, simple_loss=0.2109, pruned_loss=0.02394, over 765069.69 frames. ], batch size: 39, lr: 4.05e-03, grad_scale: 8.0 +2023-03-21 10:02:33,652 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4023, 3.3472, 1.9711, 3.7121, 3.7202, 3.7499, 3.3075, 3.3991], + device='cuda:0'), covar=tensor([0.2494, 0.0831, 0.4917, 0.0479, 0.0288, 0.0262, 0.0502, 0.0408], + device='cuda:0'), in_proj_covar=tensor([0.0252, 0.0233, 0.0247, 0.0260, 0.0201, 0.0203, 0.0220, 0.0228], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:02:53,726 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.262e+02 1.658e+02 2.006e+02 2.352e+02 4.117e+02, threshold=4.012e+02, percent-clipped=0.0 +2023-03-21 10:02:54,983 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6490, 3.6753, 2.6563, 4.0005, 3.1244, 3.7425, 1.8463, 2.6407], + device='cuda:0'), covar=tensor([0.0500, 0.0792, 0.2952, 0.0554, 0.0512, 0.0752, 0.4073, 0.1987], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0256, 0.0279, 0.0266, 0.0270, 0.0267, 0.0232, 0.0257], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:02:58,221 INFO [train.py:901] (0/2) Epoch 41, batch 200, loss[loss=0.1288, simple_loss=0.2116, pruned_loss=0.02299, over 7353.00 frames. ], tot_loss[loss=0.129, simple_loss=0.2109, pruned_loss=0.02355, over 917097.26 frames. ], batch size: 63, lr: 4.05e-03, grad_scale: 8.0 +2023-03-21 10:03:02,223 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 10:03:03,887 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7796, 1.5860, 1.8824, 2.0929, 1.9950, 2.0662, 1.6324, 2.0945], + device='cuda:0'), covar=tensor([0.2418, 0.3090, 0.1815, 0.0955, 0.1572, 0.1553, 0.1837, 0.2709], + device='cuda:0'), in_proj_covar=tensor([0.0078, 0.0079, 0.0072, 0.0064, 0.0065, 0.0064, 0.0103, 0.0069], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:03:06,784 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 10:03:13,290 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 10:03:24,895 INFO [train.py:901] (0/2) Epoch 41, batch 250, loss[loss=0.1382, simple_loss=0.2169, pruned_loss=0.02976, over 7271.00 frames. ], tot_loss[loss=0.1282, simple_loss=0.2101, pruned_loss=0.02313, over 1033436.30 frames. ], batch size: 70, lr: 4.05e-03, grad_scale: 8.0 +2023-03-21 10:03:26,900 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 10:03:28,455 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=113218.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:03:36,076 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=113233.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:03:36,604 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=113234.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:03:46,067 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.202e+02 1.702e+02 2.019e+02 2.239e+02 3.429e+02, threshold=4.038e+02, percent-clipped=0.0 +2023-03-21 10:03:47,664 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 10:03:50,100 INFO [train.py:901] (0/2) Epoch 41, batch 300, loss[loss=0.1426, simple_loss=0.2337, pruned_loss=0.02571, over 7143.00 frames. ], tot_loss[loss=0.1285, simple_loss=0.2106, pruned_loss=0.02326, over 1124103.04 frames. ], batch size: 98, lr: 4.04e-03, grad_scale: 4.0 +2023-03-21 10:03:52,706 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=113266.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:03:56,335 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=113273.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 10:03:56,727 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 10:03:58,973 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=113277.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:04:00,949 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=113281.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:04:08,770 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=113295.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:04:15,548 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.64 vs. limit=2.0 +2023-03-21 10:04:16,638 INFO [train.py:901] (0/2) Epoch 41, batch 350, loss[loss=0.1281, simple_loss=0.213, pruned_loss=0.02165, over 7241.00 frames. ], tot_loss[loss=0.1285, simple_loss=0.2105, pruned_loss=0.02325, over 1194105.84 frames. ], batch size: 89, lr: 4.04e-03, grad_scale: 4.0 +2023-03-21 10:04:21,722 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=113321.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:04:31,715 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 10:04:37,769 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.063e+02 1.665e+02 1.929e+02 2.230e+02 3.805e+02, threshold=3.858e+02, percent-clipped=0.0 +2023-03-21 10:04:41,795 INFO [train.py:901] (0/2) Epoch 41, batch 400, loss[loss=0.1337, simple_loss=0.22, pruned_loss=0.02374, over 7266.00 frames. ], tot_loss[loss=0.1279, simple_loss=0.2099, pruned_loss=0.02295, over 1248950.18 frames. ], batch size: 89, lr: 4.04e-03, grad_scale: 8.0 +2023-03-21 10:04:53,199 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-21 10:05:08,137 INFO [train.py:901] (0/2) Epoch 41, batch 450, loss[loss=0.1303, simple_loss=0.2145, pruned_loss=0.02303, over 7325.00 frames. ], tot_loss[loss=0.1285, simple_loss=0.2101, pruned_loss=0.0235, over 1291542.22 frames. ], batch size: 59, lr: 4.04e-03, grad_scale: 8.0 +2023-03-21 10:05:13,165 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 10:05:13,669 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 10:05:14,809 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9582, 3.2478, 3.8922, 4.0483, 3.9702, 3.9934, 3.9802, 3.7368], + device='cuda:0'), covar=tensor([0.0042, 0.0147, 0.0047, 0.0039, 0.0042, 0.0040, 0.0048, 0.0070], + device='cuda:0'), in_proj_covar=tensor([0.0049, 0.0071, 0.0058, 0.0057, 0.0055, 0.0061, 0.0048, 0.0078], + device='cuda:0'), out_proj_covar=tensor([8.1881e-05, 1.4334e-04, 1.0501e-04, 9.7256e-05, 9.3465e-05, 1.0572e-04, + 9.0222e-05, 1.4354e-04], device='cuda:0') +2023-03-21 10:05:29,863 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.168e+02 1.845e+02 2.130e+02 2.461e+02 6.042e+02, threshold=4.260e+02, percent-clipped=2.0 +2023-03-21 10:05:33,940 INFO [train.py:901] (0/2) Epoch 41, batch 500, loss[loss=0.1336, simple_loss=0.2153, pruned_loss=0.02598, over 7224.00 frames. ], tot_loss[loss=0.129, simple_loss=0.2101, pruned_loss=0.02394, over 1324385.98 frames. ], batch size: 50, lr: 4.04e-03, grad_scale: 8.0 +2023-03-21 10:05:34,571 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7508, 2.2336, 2.8009, 2.8495, 2.8405, 2.7102, 2.3404, 2.8503], + device='cuda:0'), covar=tensor([0.1421, 0.1141, 0.1019, 0.0938, 0.0832, 0.0988, 0.2081, 0.1008], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0070, 0.0052, 0.0052, 0.0051, 0.0050, 0.0071, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 10:05:48,389 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 10:05:49,943 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 10:05:50,441 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 10:05:52,523 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 10:05:54,929 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.67 vs. limit=2.0 +2023-03-21 10:05:57,027 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 10:06:00,022 INFO [train.py:901] (0/2) Epoch 41, batch 550, loss[loss=0.1448, simple_loss=0.2283, pruned_loss=0.03067, over 6683.00 frames. ], tot_loss[loss=0.1281, simple_loss=0.2091, pruned_loss=0.02357, over 1348738.18 frames. ], batch size: 106, lr: 4.04e-03, grad_scale: 8.0 +2023-03-21 10:06:02,978 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.10 vs. limit=5.0 +2023-03-21 10:06:09,573 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 10:06:12,554 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.90 vs. limit=2.0 +2023-03-21 10:06:18,338 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 10:06:22,448 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.173e+02 1.584e+02 1.802e+02 2.149e+02 4.424e+02, threshold=3.603e+02, percent-clipped=1.0 +2023-03-21 10:06:22,492 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 10:06:26,489 INFO [train.py:901] (0/2) Epoch 41, batch 600, loss[loss=0.1376, simple_loss=0.2194, pruned_loss=0.02789, over 7286.00 frames. ], tot_loss[loss=0.1282, simple_loss=0.2091, pruned_loss=0.02367, over 1370778.45 frames. ], batch size: 68, lr: 4.04e-03, grad_scale: 8.0 +2023-03-21 10:06:28,998 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 10:06:34,688 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=113577.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:06:41,241 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=113590.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:06:44,185 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 10:06:52,134 INFO [train.py:901] (0/2) Epoch 41, batch 650, loss[loss=0.1037, simple_loss=0.1762, pruned_loss=0.01565, over 7180.00 frames. ], tot_loss[loss=0.1285, simple_loss=0.2096, pruned_loss=0.02375, over 1385933.13 frames. ], batch size: 39, lr: 4.04e-03, grad_scale: 8.0 +2023-03-21 10:06:52,678 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 10:06:55,251 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=113617.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:06:58,374 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2811, 3.1461, 2.3391, 3.4319, 2.4440, 3.1243, 1.5638, 2.3938], + device='cuda:0'), covar=tensor([0.0578, 0.0977, 0.2827, 0.0748, 0.0625, 0.0895, 0.4052, 0.1772], + device='cuda:0'), in_proj_covar=tensor([0.0271, 0.0253, 0.0275, 0.0262, 0.0267, 0.0262, 0.0229, 0.0254], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:06:59,263 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=113625.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:07:11,048 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 10:07:11,159 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=113646.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:07:14,447 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.108e+02 1.694e+02 1.988e+02 2.330e+02 3.626e+02, threshold=3.977e+02, percent-clipped=1.0 +2023-03-21 10:07:18,571 INFO [train.py:901] (0/2) Epoch 41, batch 700, loss[loss=0.1283, simple_loss=0.2078, pruned_loss=0.02438, over 7311.00 frames. ], tot_loss[loss=0.1286, simple_loss=0.2097, pruned_loss=0.02373, over 1397782.84 frames. ], batch size: 49, lr: 4.04e-03, grad_scale: 8.0 +2023-03-21 10:07:19,599 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 10:07:22,679 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5591, 1.7892, 1.4436, 1.7721, 1.8394, 1.6874, 1.7974, 1.5071], + device='cuda:0'), covar=tensor([0.0209, 0.0226, 0.0262, 0.0245, 0.0139, 0.0141, 0.0117, 0.0176], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0037, 0.0036, 0.0038, 0.0036, 0.0034, 0.0038, 0.0045], + device='cuda:0'), out_proj_covar=tensor([4.3087e-05, 4.1358e-05, 4.0621e-05, 4.1566e-05, 3.9937e-05, 3.7954e-05, + 4.2210e-05, 5.0235e-05], device='cuda:0') +2023-03-21 10:07:24,065 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2927, 3.9242, 3.9201, 3.9778, 3.9144, 3.8210, 4.0898, 3.6529], + device='cuda:0'), covar=tensor([0.0129, 0.0168, 0.0123, 0.0173, 0.0443, 0.0116, 0.0153, 0.0186], + device='cuda:0'), in_proj_covar=tensor([0.0099, 0.0106, 0.0107, 0.0092, 0.0183, 0.0111, 0.0108, 0.0117], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 10:07:24,140 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2132, 3.3222, 2.3956, 3.5417, 2.6938, 3.2706, 1.6086, 2.4404], + device='cuda:0'), covar=tensor([0.0452, 0.0734, 0.2515, 0.0609, 0.0493, 0.0540, 0.4015, 0.1861], + device='cuda:0'), in_proj_covar=tensor([0.0273, 0.0254, 0.0277, 0.0264, 0.0269, 0.0264, 0.0231, 0.0256], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:07:27,125 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=113678.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:07:39,096 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7463, 4.2529, 4.0473, 4.6475, 4.4634, 4.6199, 4.1406, 4.2747], + device='cuda:0'), covar=tensor([0.0843, 0.2625, 0.2502, 0.1102, 0.0937, 0.1221, 0.0825, 0.1146], + device='cuda:0'), in_proj_covar=tensor([0.0263, 0.0396, 0.0296, 0.0309, 0.0230, 0.0368, 0.0232, 0.0273], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 10:07:41,545 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 10:07:41,675 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=113707.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:07:42,075 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 10:07:43,471 INFO [train.py:901] (0/2) Epoch 41, batch 750, loss[loss=0.1343, simple_loss=0.2194, pruned_loss=0.0246, over 7267.00 frames. ], tot_loss[loss=0.129, simple_loss=0.2104, pruned_loss=0.02378, over 1408289.02 frames. ], batch size: 89, lr: 4.04e-03, grad_scale: 8.0 +2023-03-21 10:07:57,654 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 10:08:02,020 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 10:08:05,447 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.130e+02 1.674e+02 1.983e+02 2.352e+02 7.977e+02, threshold=3.966e+02, percent-clipped=3.0 +2023-03-21 10:08:08,032 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 10:08:09,521 INFO [train.py:901] (0/2) Epoch 41, batch 800, loss[loss=0.1278, simple_loss=0.2121, pruned_loss=0.0218, over 7239.00 frames. ], tot_loss[loss=0.1294, simple_loss=0.2107, pruned_loss=0.02398, over 1419233.84 frames. ], batch size: 55, lr: 4.04e-03, grad_scale: 8.0 +2023-03-21 10:08:09,534 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 10:08:15,597 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7593, 2.3299, 2.8714, 2.8213, 2.8514, 2.7887, 2.4529, 2.9301], + device='cuda:0'), covar=tensor([0.1417, 0.0889, 0.1392, 0.1039, 0.0815, 0.0896, 0.1897, 0.1230], + device='cuda:0'), in_proj_covar=tensor([0.0052, 0.0069, 0.0052, 0.0051, 0.0051, 0.0050, 0.0069, 0.0051], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 10:08:19,503 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 10:08:27,344 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-03-21 10:08:33,692 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4638, 3.1879, 3.3958, 3.3382, 3.1392, 2.9631, 3.5259, 2.4706], + device='cuda:0'), covar=tensor([0.0457, 0.0574, 0.0749, 0.0582, 0.0769, 0.1016, 0.0582, 0.2674], + device='cuda:0'), in_proj_covar=tensor([0.0329, 0.0332, 0.0272, 0.0355, 0.0284, 0.0286, 0.0346, 0.0244], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:08:36,105 INFO [train.py:901] (0/2) Epoch 41, batch 850, loss[loss=0.135, simple_loss=0.2165, pruned_loss=0.02675, over 7249.00 frames. ], tot_loss[loss=0.129, simple_loss=0.2099, pruned_loss=0.02401, over 1420997.74 frames. ], batch size: 47, lr: 4.03e-03, grad_scale: 8.0 +2023-03-21 10:08:39,692 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 10:08:40,191 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 10:08:45,252 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 10:08:48,877 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 10:08:49,987 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7555, 3.2667, 2.7813, 3.0763, 3.0318, 2.7588, 3.0740, 2.8539], + device='cuda:0'), covar=tensor([0.0685, 0.0525, 0.0778, 0.0785, 0.0979, 0.0456, 0.0773, 0.1236], + device='cuda:0'), in_proj_covar=tensor([0.0059, 0.0059, 0.0067, 0.0060, 0.0057, 0.0061, 0.0056, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 10:08:57,503 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.207e+02 1.665e+02 1.936e+02 2.284e+02 3.420e+02, threshold=3.871e+02, percent-clipped=0.0 +2023-03-21 10:09:01,606 INFO [train.py:901] (0/2) Epoch 41, batch 900, loss[loss=0.126, simple_loss=0.2155, pruned_loss=0.01824, over 7328.00 frames. ], tot_loss[loss=0.1291, simple_loss=0.2102, pruned_loss=0.02405, over 1426275.61 frames. ], batch size: 61, lr: 4.03e-03, grad_scale: 8.0 +2023-03-21 10:09:17,115 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=113890.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:09:27,307 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 10:09:28,304 INFO [train.py:901] (0/2) Epoch 41, batch 950, loss[loss=0.09253, simple_loss=0.1674, pruned_loss=0.008839, over 7185.00 frames. ], tot_loss[loss=0.1294, simple_loss=0.2108, pruned_loss=0.02405, over 1429298.91 frames. ], batch size: 39, lr: 4.03e-03, grad_scale: 8.0 +2023-03-21 10:09:42,019 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=113938.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:09:49,398 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.076e+02 1.644e+02 2.020e+02 2.352e+02 7.244e+02, threshold=4.039e+02, percent-clipped=3.0 +2023-03-21 10:09:51,458 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 10:09:53,452 INFO [train.py:901] (0/2) Epoch 41, batch 1000, loss[loss=0.1262, simple_loss=0.2088, pruned_loss=0.0218, over 7354.00 frames. ], tot_loss[loss=0.1292, simple_loss=0.2104, pruned_loss=0.02402, over 1431316.09 frames. ], batch size: 44, lr: 4.03e-03, grad_scale: 8.0 +2023-03-21 10:09:54,753 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.43 vs. limit=5.0 +2023-03-21 10:10:00,237 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=113973.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:10:11,898 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 10:10:15,742 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=114002.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:10:20,134 INFO [train.py:901] (0/2) Epoch 41, batch 1050, loss[loss=0.1138, simple_loss=0.1979, pruned_loss=0.01485, over 7127.00 frames. ], tot_loss[loss=0.1294, simple_loss=0.2105, pruned_loss=0.02412, over 1431998.53 frames. ], batch size: 41, lr: 4.03e-03, grad_scale: 8.0 +2023-03-21 10:10:20,760 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7392, 4.1460, 4.3590, 4.3662, 4.2940, 4.2537, 4.5280, 4.0208], + device='cuda:0'), covar=tensor([0.0163, 0.0196, 0.0116, 0.0168, 0.0424, 0.0112, 0.0146, 0.0207], + device='cuda:0'), in_proj_covar=tensor([0.0098, 0.0106, 0.0105, 0.0091, 0.0180, 0.0110, 0.0107, 0.0116], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 10:10:27,151 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8109, 2.1489, 1.6825, 2.1997, 2.2944, 1.8973, 1.6887, 1.4422], + device='cuda:0'), covar=tensor([0.0158, 0.0234, 0.0371, 0.0243, 0.0120, 0.0162, 0.0258, 0.0276], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0037, 0.0035, 0.0037, 0.0036, 0.0034, 0.0037, 0.0045], + device='cuda:0'), out_proj_covar=tensor([4.2799e-05, 4.1212e-05, 4.0011e-05, 4.1499e-05, 3.9467e-05, 3.7539e-05, + 4.2039e-05, 5.0183e-05], device='cuda:0') +2023-03-21 10:10:33,474 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 10:10:37,443 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 10:10:40,997 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.136e+02 1.735e+02 2.009e+02 2.345e+02 4.363e+02, threshold=4.019e+02, percent-clipped=1.0 +2023-03-21 10:10:45,262 INFO [train.py:901] (0/2) Epoch 41, batch 1100, loss[loss=0.1187, simple_loss=0.2092, pruned_loss=0.01415, over 7309.00 frames. ], tot_loss[loss=0.1297, simple_loss=0.2107, pruned_loss=0.02434, over 1434749.80 frames. ], batch size: 80, lr: 4.03e-03, grad_scale: 8.0 +2023-03-21 10:10:57,726 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2690, 2.3445, 2.5288, 2.2486, 2.3521, 2.4441, 2.1745, 1.7877], + device='cuda:0'), covar=tensor([0.0332, 0.0519, 0.0246, 0.0230, 0.0567, 0.0449, 0.0257, 0.0355], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0040, 0.0040, 0.0039, 0.0037, 0.0037, 0.0042, 0.0042], + device='cuda:0'), out_proj_covar=tensor([1.0421e-04, 1.0224e-04, 1.0172e-04, 9.9119e-05, 9.8262e-05, 9.6876e-05, + 1.0603e-04, 1.0697e-04], device='cuda:0') +2023-03-21 10:11:07,430 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3546, 2.7733, 2.8132, 2.5496, 2.6536, 2.5745, 2.3307, 2.0662], + device='cuda:0'), covar=tensor([0.0683, 0.0475, 0.0368, 0.0268, 0.0477, 0.0599, 0.0402, 0.0462], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0040, 0.0041, 0.0039, 0.0038, 0.0037, 0.0043, 0.0043], + device='cuda:0'), out_proj_covar=tensor([1.0453e-04, 1.0273e-04, 1.0205e-04, 9.9505e-05, 9.8774e-05, 9.7229e-05, + 1.0646e-04, 1.0760e-04], device='cuda:0') +2023-03-21 10:11:07,800 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 10:11:07,813 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 10:11:08,866 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8574, 4.1002, 3.7814, 4.0761, 3.6432, 4.0795, 4.3623, 4.3800], + device='cuda:0'), covar=tensor([0.0245, 0.0171, 0.0251, 0.0168, 0.0338, 0.0322, 0.0219, 0.0197], + device='cuda:0'), in_proj_covar=tensor([0.0124, 0.0127, 0.0121, 0.0124, 0.0114, 0.0101, 0.0100, 0.0102], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 10:11:11,751 INFO [train.py:901] (0/2) Epoch 41, batch 1150, loss[loss=0.1292, simple_loss=0.2138, pruned_loss=0.02228, over 7276.00 frames. ], tot_loss[loss=0.1295, simple_loss=0.2108, pruned_loss=0.02408, over 1437848.58 frames. ], batch size: 66, lr: 4.03e-03, grad_scale: 8.0 +2023-03-21 10:11:13,435 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4541, 3.1408, 3.2690, 3.2819, 2.9970, 2.9555, 3.2946, 2.3795], + device='cuda:0'), covar=tensor([0.0435, 0.0551, 0.0751, 0.0631, 0.0695, 0.1011, 0.0644, 0.2537], + device='cuda:0'), in_proj_covar=tensor([0.0330, 0.0334, 0.0273, 0.0356, 0.0284, 0.0286, 0.0348, 0.0244], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:11:18,590 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.05 vs. limit=5.0 +2023-03-21 10:11:19,853 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=114127.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:11:20,777 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 10:11:21,272 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 10:11:33,372 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.308e+02 1.735e+02 1.978e+02 2.469e+02 4.023e+02, threshold=3.955e+02, percent-clipped=1.0 +2023-03-21 10:11:38,052 INFO [train.py:901] (0/2) Epoch 41, batch 1200, loss[loss=0.1329, simple_loss=0.2192, pruned_loss=0.02334, over 7314.00 frames. ], tot_loss[loss=0.1288, simple_loss=0.2098, pruned_loss=0.02386, over 1437071.22 frames. ], batch size: 59, lr: 4.03e-03, grad_scale: 8.0 +2023-03-21 10:11:48,747 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.1395, 5.4738, 5.3975, 5.3589, 5.2688, 5.1599, 5.5131, 5.3466], + device='cuda:0'), covar=tensor([0.0580, 0.0731, 0.0759, 0.0952, 0.0626, 0.0550, 0.0681, 0.0887], + device='cuda:0'), in_proj_covar=tensor([0.0223, 0.0267, 0.0206, 0.0206, 0.0163, 0.0232, 0.0217, 0.0151], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:11:51,754 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=114188.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:11:53,001 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.58 vs. limit=2.0 +2023-03-21 10:11:54,155 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 10:12:03,598 INFO [train.py:901] (0/2) Epoch 41, batch 1250, loss[loss=0.1405, simple_loss=0.2227, pruned_loss=0.02918, over 7128.00 frames. ], tot_loss[loss=0.1293, simple_loss=0.2103, pruned_loss=0.02419, over 1439157.43 frames. ], batch size: 98, lr: 4.03e-03, grad_scale: 8.0 +2023-03-21 10:12:18,192 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 10:12:22,630 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 10:12:24,692 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 10:12:25,677 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.306e+02 1.779e+02 2.107e+02 2.486e+02 4.646e+02, threshold=4.214e+02, percent-clipped=2.0 +2023-03-21 10:12:27,958 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9802, 3.1340, 2.2091, 3.3305, 2.4125, 2.9076, 1.4619, 2.3683], + device='cuda:0'), covar=tensor([0.0558, 0.1079, 0.3197, 0.0842, 0.0601, 0.0687, 0.4312, 0.1882], + device='cuda:0'), in_proj_covar=tensor([0.0277, 0.0256, 0.0279, 0.0267, 0.0271, 0.0265, 0.0234, 0.0258], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:12:29,811 INFO [train.py:901] (0/2) Epoch 41, batch 1300, loss[loss=0.1363, simple_loss=0.2165, pruned_loss=0.02803, over 7276.00 frames. ], tot_loss[loss=0.1287, simple_loss=0.2095, pruned_loss=0.024, over 1440013.51 frames. ], batch size: 77, lr: 4.03e-03, grad_scale: 8.0 +2023-03-21 10:12:35,071 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4474, 3.6233, 2.7395, 3.8192, 3.1212, 3.4230, 1.7968, 2.7373], + device='cuda:0'), covar=tensor([0.0513, 0.0615, 0.2343, 0.0524, 0.0495, 0.0662, 0.3799, 0.1856], + device='cuda:0'), in_proj_covar=tensor([0.0276, 0.0256, 0.0279, 0.0267, 0.0270, 0.0265, 0.0233, 0.0258], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:12:36,034 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=114273.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:12:43,673 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0857, 3.7738, 3.8947, 3.9219, 3.7064, 3.6427, 4.2428, 2.6992], + device='cuda:0'), covar=tensor([0.0435, 0.0555, 0.0478, 0.0560, 0.0818, 0.0886, 0.0753, 0.1953], + device='cuda:0'), in_proj_covar=tensor([0.0328, 0.0333, 0.0271, 0.0354, 0.0283, 0.0285, 0.0347, 0.0242], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:12:48,997 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 10:12:50,550 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=114302.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:12:50,890 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.91 vs. limit=5.0 +2023-03-21 10:12:50,983 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 10:12:53,667 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0555, 2.6029, 2.5649, 4.0128, 2.0734, 3.7529, 1.5485, 3.5708], + device='cuda:0'), covar=tensor([0.0257, 0.1502, 0.1865, 0.0286, 0.4121, 0.0314, 0.1400, 0.0498], + device='cuda:0'), in_proj_covar=tensor([0.0206, 0.0248, 0.0261, 0.0210, 0.0251, 0.0218, 0.0224, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:12:54,038 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 10:12:55,004 INFO [train.py:901] (0/2) Epoch 41, batch 1350, loss[loss=0.1066, simple_loss=0.1857, pruned_loss=0.01369, over 7324.00 frames. ], tot_loss[loss=0.1291, simple_loss=0.2096, pruned_loss=0.02428, over 1437411.83 frames. ], batch size: 44, lr: 4.03e-03, grad_scale: 8.0 +2023-03-21 10:13:00,104 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=114321.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:13:04,120 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 10:13:15,959 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=114350.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:13:17,364 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.320e+02 1.723e+02 1.987e+02 2.358e+02 3.719e+02, threshold=3.975e+02, percent-clipped=0.0 +2023-03-21 10:13:21,355 INFO [train.py:901] (0/2) Epoch 41, batch 1400, loss[loss=0.1019, simple_loss=0.1652, pruned_loss=0.01929, over 5834.00 frames. ], tot_loss[loss=0.1292, simple_loss=0.2102, pruned_loss=0.02406, over 1436979.59 frames. ], batch size: 25, lr: 4.02e-03, grad_scale: 8.0 +2023-03-21 10:13:36,861 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 10:13:46,777 INFO [train.py:901] (0/2) Epoch 41, batch 1450, loss[loss=0.1443, simple_loss=0.2267, pruned_loss=0.03092, over 7143.00 frames. ], tot_loss[loss=0.1298, simple_loss=0.2107, pruned_loss=0.02443, over 1438697.03 frames. ], batch size: 98, lr: 4.02e-03, grad_scale: 8.0 +2023-03-21 10:14:02,600 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 10:14:05,073 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.79 vs. limit=2.0 +2023-03-21 10:14:09,122 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.348e+02 1.681e+02 1.902e+02 2.503e+02 5.939e+02, threshold=3.804e+02, percent-clipped=2.0 +2023-03-21 10:14:13,120 INFO [train.py:901] (0/2) Epoch 41, batch 1500, loss[loss=0.1377, simple_loss=0.2229, pruned_loss=0.02625, over 7268.00 frames. ], tot_loss[loss=0.129, simple_loss=0.2099, pruned_loss=0.02407, over 1438921.62 frames. ], batch size: 70, lr: 4.02e-03, grad_scale: 8.0 +2023-03-21 10:14:19,189 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 10:14:24,341 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=114483.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:14:39,029 INFO [train.py:901] (0/2) Epoch 41, batch 1550, loss[loss=0.114, simple_loss=0.1994, pruned_loss=0.01433, over 7325.00 frames. ], tot_loss[loss=0.1294, simple_loss=0.2105, pruned_loss=0.02414, over 1440317.20 frames. ], batch size: 61, lr: 4.02e-03, grad_scale: 8.0 +2023-03-21 10:14:43,745 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 10:15:00,785 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.328e+02 1.777e+02 2.044e+02 2.457e+02 4.625e+02, threshold=4.087e+02, percent-clipped=1.0 +2023-03-21 10:15:04,695 INFO [train.py:901] (0/2) Epoch 41, batch 1600, loss[loss=0.1376, simple_loss=0.2199, pruned_loss=0.02762, over 7289.00 frames. ], tot_loss[loss=0.129, simple_loss=0.2098, pruned_loss=0.02408, over 1439292.58 frames. ], batch size: 86, lr: 4.02e-03, grad_scale: 8.0 +2023-03-21 10:15:14,001 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 10:15:14,520 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 10:15:16,960 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 10:15:28,008 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 10:15:31,065 INFO [train.py:901] (0/2) Epoch 41, batch 1650, loss[loss=0.1323, simple_loss=0.2174, pruned_loss=0.02365, over 7299.00 frames. ], tot_loss[loss=0.1297, simple_loss=0.2105, pruned_loss=0.02447, over 1442203.33 frames. ], batch size: 86, lr: 4.02e-03, grad_scale: 8.0 +2023-03-21 10:15:32,568 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 10:15:40,579 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 10:15:47,736 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1299, 3.1862, 2.1448, 3.4918, 2.5471, 3.0829, 1.5254, 2.1437], + device='cuda:0'), covar=tensor([0.0555, 0.0966, 0.3555, 0.0767, 0.0622, 0.0728, 0.4307, 0.2167], + device='cuda:0'), in_proj_covar=tensor([0.0276, 0.0257, 0.0278, 0.0267, 0.0270, 0.0265, 0.0232, 0.0257], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:15:52,151 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.270e+02 1.753e+02 2.036e+02 2.309e+02 4.360e+02, threshold=4.073e+02, percent-clipped=1.0 +2023-03-21 10:15:56,188 INFO [train.py:901] (0/2) Epoch 41, batch 1700, loss[loss=0.1349, simple_loss=0.2178, pruned_loss=0.026, over 7312.00 frames. ], tot_loss[loss=0.1298, simple_loss=0.2107, pruned_loss=0.02444, over 1443174.99 frames. ], batch size: 83, lr: 4.02e-03, grad_scale: 8.0 +2023-03-21 10:15:57,197 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 10:16:01,238 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 10:16:13,204 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6572, 3.7091, 2.8260, 4.0711, 3.1555, 3.5083, 1.8333, 2.7254], + device='cuda:0'), covar=tensor([0.0561, 0.0824, 0.2543, 0.0543, 0.0585, 0.0940, 0.4414, 0.2217], + device='cuda:0'), in_proj_covar=tensor([0.0277, 0.0257, 0.0278, 0.0268, 0.0271, 0.0265, 0.0233, 0.0257], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:16:13,549 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 10:16:22,616 INFO [train.py:901] (0/2) Epoch 41, batch 1750, loss[loss=0.1493, simple_loss=0.2405, pruned_loss=0.02903, over 6695.00 frames. ], tot_loss[loss=0.1299, simple_loss=0.211, pruned_loss=0.02438, over 1443998.96 frames. ], batch size: 107, lr: 4.02e-03, grad_scale: 8.0 +2023-03-21 10:16:25,344 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=114716.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:16:25,601 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=2.06 vs. limit=2.0 +2023-03-21 10:16:37,412 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 10:16:38,411 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 10:16:43,882 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.146e+02 1.738e+02 2.103e+02 2.462e+02 3.793e+02, threshold=4.205e+02, percent-clipped=0.0 +2023-03-21 10:16:47,914 INFO [train.py:901] (0/2) Epoch 41, batch 1800, loss[loss=0.1406, simple_loss=0.2147, pruned_loss=0.0332, over 7308.00 frames. ], tot_loss[loss=0.1298, simple_loss=0.2109, pruned_loss=0.02437, over 1442779.82 frames. ], batch size: 49, lr: 4.02e-03, grad_scale: 8.0 +2023-03-21 10:16:56,751 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=114777.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:17:00,409 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=114783.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:17:01,336 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 10:17:14,683 INFO [train.py:901] (0/2) Epoch 41, batch 1850, loss[loss=0.1289, simple_loss=0.2149, pruned_loss=0.02143, over 7325.00 frames. ], tot_loss[loss=0.1294, simple_loss=0.2106, pruned_loss=0.02411, over 1442773.65 frames. ], batch size: 59, lr: 4.02e-03, grad_scale: 8.0 +2023-03-21 10:17:14,692 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 10:17:24,313 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 10:17:24,863 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=114831.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:17:36,017 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.195e+02 1.665e+02 1.955e+02 2.157e+02 5.091e+02, threshold=3.909e+02, percent-clipped=1.0 +2023-03-21 10:17:40,525 INFO [train.py:901] (0/2) Epoch 41, batch 1900, loss[loss=0.09671, simple_loss=0.1578, pruned_loss=0.01782, over 6261.00 frames. ], tot_loss[loss=0.1295, simple_loss=0.2109, pruned_loss=0.02408, over 1443156.85 frames. ], batch size: 27, lr: 4.02e-03, grad_scale: 8.0 +2023-03-21 10:17:41,073 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 10:17:46,370 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=114871.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:17:49,832 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8791, 3.5864, 3.5995, 3.6132, 3.5395, 3.4584, 3.7557, 3.4333], + device='cuda:0'), covar=tensor([0.0127, 0.0192, 0.0118, 0.0189, 0.0429, 0.0116, 0.0162, 0.0146], + device='cuda:0'), in_proj_covar=tensor([0.0097, 0.0104, 0.0105, 0.0090, 0.0179, 0.0109, 0.0108, 0.0115], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 10:18:06,229 INFO [train.py:901] (0/2) Epoch 41, batch 1950, loss[loss=0.1292, simple_loss=0.2123, pruned_loss=0.02307, over 7290.00 frames. ], tot_loss[loss=0.1296, simple_loss=0.211, pruned_loss=0.02409, over 1443077.37 frames. ], batch size: 66, lr: 4.02e-03, grad_scale: 8.0 +2023-03-21 10:18:06,265 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 10:18:07,442 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1997, 2.6119, 1.9895, 3.0295, 2.7942, 2.8184, 2.6057, 2.6126], + device='cuda:0'), covar=tensor([0.2227, 0.1148, 0.3949, 0.0573, 0.0241, 0.0284, 0.0355, 0.0336], + device='cuda:0'), in_proj_covar=tensor([0.0248, 0.0230, 0.0243, 0.0255, 0.0196, 0.0201, 0.0218, 0.0223], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:18:09,667 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.95 vs. limit=2.0 +2023-03-21 10:18:16,834 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 10:18:16,975 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=114932.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:18:21,740 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 10:18:22,269 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 10:18:25,744 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.28 vs. limit=5.0 +2023-03-21 10:18:27,919 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.257e+02 1.719e+02 2.045e+02 2.389e+02 4.701e+02, threshold=4.090e+02, percent-clipped=2.0 +2023-03-21 10:18:32,593 INFO [train.py:901] (0/2) Epoch 41, batch 2000, loss[loss=0.1171, simple_loss=0.2009, pruned_loss=0.01663, over 7200.00 frames. ], tot_loss[loss=0.1292, simple_loss=0.2107, pruned_loss=0.0239, over 1442629.54 frames. ], batch size: 50, lr: 4.01e-03, grad_scale: 8.0 +2023-03-21 10:18:34,774 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0934, 3.9105, 3.2421, 3.6816, 2.9399, 2.3162, 1.9386, 4.1071], + device='cuda:0'), covar=tensor([0.0039, 0.0069, 0.0145, 0.0060, 0.0165, 0.0518, 0.0580, 0.0043], + device='cuda:0'), in_proj_covar=tensor([0.0115, 0.0094, 0.0113, 0.0095, 0.0131, 0.0136, 0.0128, 0.0106], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 10:18:39,232 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 10:18:42,409 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5791, 1.8963, 1.7233, 2.0161, 2.0807, 1.7562, 1.7998, 1.6290], + device='cuda:0'), covar=tensor([0.0253, 0.0280, 0.0415, 0.0184, 0.0140, 0.0157, 0.0170, 0.0210], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0037, 0.0035, 0.0037, 0.0035, 0.0033, 0.0037, 0.0045], + device='cuda:0'), out_proj_covar=tensor([4.2293e-05, 4.1102e-05, 3.9627e-05, 4.0887e-05, 3.9101e-05, 3.7010e-05, + 4.1695e-05, 4.9597e-05], device='cuda:0') +2023-03-21 10:18:49,373 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 10:18:57,886 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 10:18:58,360 INFO [train.py:901] (0/2) Epoch 41, batch 2050, loss[loss=0.1305, simple_loss=0.2106, pruned_loss=0.02516, over 7332.00 frames. ], tot_loss[loss=0.1301, simple_loss=0.2115, pruned_loss=0.02439, over 1441309.46 frames. ], batch size: 44, lr: 4.01e-03, grad_scale: 8.0 +2023-03-21 10:19:20,707 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.213e+02 1.710e+02 1.996e+02 2.355e+02 4.786e+02, threshold=3.993e+02, percent-clipped=2.0 +2023-03-21 10:19:24,719 INFO [train.py:901] (0/2) Epoch 41, batch 2100, loss[loss=0.1336, simple_loss=0.2207, pruned_loss=0.02327, over 7323.00 frames. ], tot_loss[loss=0.1301, simple_loss=0.2116, pruned_loss=0.02427, over 1442366.20 frames. ], batch size: 83, lr: 4.01e-03, grad_scale: 8.0 +2023-03-21 10:19:30,205 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.84 vs. limit=2.0 +2023-03-21 10:19:30,473 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=115072.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:19:32,000 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 10:19:34,608 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 10:19:37,710 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=115086.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:19:50,076 INFO [train.py:901] (0/2) Epoch 41, batch 2150, loss[loss=0.1364, simple_loss=0.217, pruned_loss=0.02796, over 7318.00 frames. ], tot_loss[loss=0.1303, simple_loss=0.2116, pruned_loss=0.02445, over 1441432.23 frames. ], batch size: 75, lr: 4.01e-03, grad_scale: 8.0 +2023-03-21 10:19:56,436 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8756, 4.3475, 4.4555, 4.3604, 4.4017, 3.9858, 4.4196, 4.3365], + device='cuda:0'), covar=tensor([0.0499, 0.0411, 0.0358, 0.0491, 0.0318, 0.0461, 0.0361, 0.0418], + device='cuda:0'), in_proj_covar=tensor([0.0223, 0.0264, 0.0203, 0.0205, 0.0161, 0.0233, 0.0216, 0.0151], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:20:09,470 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=115147.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:20:10,918 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6907, 5.2100, 5.3465, 5.2493, 5.0561, 4.7368, 5.3112, 5.1211], + device='cuda:0'), covar=tensor([0.0436, 0.0365, 0.0335, 0.0428, 0.0344, 0.0402, 0.0318, 0.0442], + device='cuda:0'), in_proj_covar=tensor([0.0222, 0.0264, 0.0203, 0.0205, 0.0161, 0.0232, 0.0215, 0.0151], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:20:12,342 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.139e+02 1.680e+02 1.975e+02 2.353e+02 5.510e+02, threshold=3.949e+02, percent-clipped=1.0 +2023-03-21 10:20:16,350 INFO [train.py:901] (0/2) Epoch 41, batch 2200, loss[loss=0.1148, simple_loss=0.1923, pruned_loss=0.01865, over 7265.00 frames. ], tot_loss[loss=0.1301, simple_loss=0.2117, pruned_loss=0.02424, over 1440879.85 frames. ], batch size: 42, lr: 4.01e-03, grad_scale: 8.0 +2023-03-21 10:20:17,947 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 10:20:40,728 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5688, 1.9506, 1.6603, 1.9389, 2.0403, 1.7514, 1.8764, 1.4538], + device='cuda:0'), covar=tensor([0.0213, 0.0252, 0.0293, 0.0157, 0.0154, 0.0163, 0.0134, 0.0193], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0038, 0.0036, 0.0038, 0.0036, 0.0034, 0.0038, 0.0046], + device='cuda:0'), out_proj_covar=tensor([4.3460e-05, 4.2060e-05, 4.0768e-05, 4.1510e-05, 3.9825e-05, 3.7956e-05, + 4.2398e-05, 5.0637e-05], device='cuda:0') +2023-03-21 10:20:43,074 INFO [train.py:901] (0/2) Epoch 41, batch 2250, loss[loss=0.1391, simple_loss=0.2205, pruned_loss=0.0289, over 7262.00 frames. ], tot_loss[loss=0.1298, simple_loss=0.2114, pruned_loss=0.02411, over 1442656.41 frames. ], batch size: 64, lr: 4.01e-03, grad_scale: 8.0 +2023-03-21 10:20:51,203 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=115227.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:20:51,725 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0830, 2.4639, 3.1207, 2.9826, 3.1070, 2.8731, 2.6089, 2.9641], + device='cuda:0'), covar=tensor([0.1173, 0.0747, 0.0832, 0.1064, 0.0665, 0.0954, 0.1659, 0.1370], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0068, 0.0051, 0.0051, 0.0050, 0.0050, 0.0068, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0004, 0.0003], + device='cuda:0') +2023-03-21 10:20:52,164 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 10:20:52,178 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 10:20:52,441 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.70 vs. limit=2.0 +2023-03-21 10:21:01,462 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5119, 2.4089, 2.4359, 3.7597, 2.0430, 3.4660, 1.4459, 3.4331], + device='cuda:0'), covar=tensor([0.0206, 0.1523, 0.1881, 0.0275, 0.4025, 0.0289, 0.1326, 0.0559], + device='cuda:0'), in_proj_covar=tensor([0.0207, 0.0248, 0.0259, 0.0213, 0.0252, 0.0219, 0.0226, 0.0231], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:21:04,392 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.338e+02 1.786e+02 2.050e+02 2.347e+02 4.305e+02, threshold=4.099e+02, percent-clipped=1.0 +2023-03-21 10:21:04,917 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 10:21:08,410 INFO [train.py:901] (0/2) Epoch 41, batch 2300, loss[loss=0.1391, simple_loss=0.2124, pruned_loss=0.03288, over 7261.00 frames. ], tot_loss[loss=0.13, simple_loss=0.2117, pruned_loss=0.02419, over 1443260.48 frames. ], batch size: 52, lr: 4.01e-03, grad_scale: 16.0 +2023-03-21 10:21:34,781 INFO [train.py:901] (0/2) Epoch 41, batch 2350, loss[loss=0.1322, simple_loss=0.2151, pruned_loss=0.02466, over 7256.00 frames. ], tot_loss[loss=0.1295, simple_loss=0.2113, pruned_loss=0.02391, over 1443133.66 frames. ], batch size: 55, lr: 4.01e-03, grad_scale: 16.0 +2023-03-21 10:21:52,021 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 10:21:56,117 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.249e+02 1.722e+02 2.004e+02 2.418e+02 5.558e+02, threshold=4.008e+02, percent-clipped=3.0 +2023-03-21 10:21:58,723 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 10:22:00,267 INFO [train.py:901] (0/2) Epoch 41, batch 2400, loss[loss=0.1193, simple_loss=0.1994, pruned_loss=0.01961, over 7348.00 frames. ], tot_loss[loss=0.1297, simple_loss=0.2115, pruned_loss=0.0239, over 1444255.69 frames. ], batch size: 44, lr: 4.01e-03, grad_scale: 16.0 +2023-03-21 10:22:05,673 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=115370.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:22:06,165 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=115371.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:22:06,682 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=115372.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:22:09,737 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 10:22:12,826 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 10:22:27,157 INFO [train.py:901] (0/2) Epoch 41, batch 2450, loss[loss=0.132, simple_loss=0.2131, pruned_loss=0.02549, over 7270.00 frames. ], tot_loss[loss=0.13, simple_loss=0.2117, pruned_loss=0.02413, over 1443497.31 frames. ], batch size: 55, lr: 4.01e-03, grad_scale: 16.0 +2023-03-21 10:22:31,874 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=115420.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:22:37,375 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=115431.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:22:37,871 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=115432.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:22:38,254 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 10:22:42,769 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=115442.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:22:43,338 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7189, 3.1983, 2.8392, 3.2092, 3.1209, 2.8407, 3.1343, 2.8670], + device='cuda:0'), covar=tensor([0.1253, 0.0912, 0.1147, 0.0870, 0.0863, 0.0829, 0.0902, 0.1143], + device='cuda:0'), in_proj_covar=tensor([0.0060, 0.0060, 0.0068, 0.0061, 0.0057, 0.0064, 0.0057, 0.0055], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:22:48,840 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.303e+02 1.715e+02 1.917e+02 2.250e+02 3.232e+02, threshold=3.834e+02, percent-clipped=0.0 +2023-03-21 10:22:53,620 INFO [train.py:901] (0/2) Epoch 41, batch 2500, loss[loss=0.1256, simple_loss=0.2081, pruned_loss=0.0216, over 7319.00 frames. ], tot_loss[loss=0.13, simple_loss=0.2118, pruned_loss=0.02415, over 1444450.00 frames. ], batch size: 83, lr: 4.01e-03, grad_scale: 16.0 +2023-03-21 10:23:03,575 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 10:23:16,567 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.0736, 4.5388, 4.3903, 5.0630, 4.7701, 4.9200, 4.3869, 4.4712], + device='cuda:0'), covar=tensor([0.0833, 0.2698, 0.2534, 0.1042, 0.1077, 0.1225, 0.0895, 0.1317], + device='cuda:0'), in_proj_covar=tensor([0.0271, 0.0410, 0.0310, 0.0323, 0.0240, 0.0383, 0.0240, 0.0287], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 10:23:18,446 INFO [train.py:901] (0/2) Epoch 41, batch 2550, loss[loss=0.139, simple_loss=0.222, pruned_loss=0.02797, over 7354.00 frames. ], tot_loss[loss=0.1299, simple_loss=0.2114, pruned_loss=0.02418, over 1441853.97 frames. ], batch size: 73, lr: 4.00e-03, grad_scale: 8.0 +2023-03-21 10:23:26,760 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=115527.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:23:41,548 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.217e+02 1.617e+02 1.970e+02 2.385e+02 4.458e+02, threshold=3.941e+02, percent-clipped=1.0 +2023-03-21 10:23:45,041 INFO [train.py:901] (0/2) Epoch 41, batch 2600, loss[loss=0.1323, simple_loss=0.2117, pruned_loss=0.02651, over 7313.00 frames. ], tot_loss[loss=0.1297, simple_loss=0.211, pruned_loss=0.02422, over 1439440.09 frames. ], batch size: 59, lr: 4.00e-03, grad_scale: 8.0 +2023-03-21 10:23:48,642 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1491, 3.4261, 2.4457, 3.7187, 2.9305, 3.0838, 1.5480, 2.6229], + device='cuda:0'), covar=tensor([0.0454, 0.0880, 0.2502, 0.0607, 0.0431, 0.0677, 0.3807, 0.1753], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0257, 0.0276, 0.0266, 0.0270, 0.0263, 0.0232, 0.0254], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:23:52,036 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=115575.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:24:01,889 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2586, 2.3611, 2.4332, 2.2693, 2.3593, 2.2626, 2.0923, 1.8008], + device='cuda:0'), covar=tensor([0.0421, 0.0542, 0.0478, 0.0257, 0.0358, 0.0379, 0.0459, 0.0346], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0040, 0.0041, 0.0039, 0.0038, 0.0038, 0.0043, 0.0043], + device='cuda:0'), out_proj_covar=tensor([1.0434e-04, 1.0309e-04, 1.0370e-04, 1.0048e-04, 9.9346e-05, 9.8812e-05, + 1.0778e-04, 1.0816e-04], device='cuda:0') +2023-03-21 10:24:09,933 INFO [train.py:901] (0/2) Epoch 41, batch 2650, loss[loss=0.1338, simple_loss=0.2166, pruned_loss=0.02551, over 7252.00 frames. ], tot_loss[loss=0.1299, simple_loss=0.2113, pruned_loss=0.02423, over 1442893.48 frames. ], batch size: 47, lr: 4.00e-03, grad_scale: 8.0 +2023-03-21 10:24:28,110 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9005, 3.0344, 2.0914, 3.2365, 2.5206, 2.7967, 1.4084, 2.1168], + device='cuda:0'), covar=tensor([0.0648, 0.0947, 0.3199, 0.0814, 0.0619, 0.0733, 0.3935, 0.1948], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0257, 0.0276, 0.0267, 0.0271, 0.0263, 0.0232, 0.0254], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:24:31,801 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.159e+02 1.689e+02 1.943e+02 2.398e+02 3.998e+02, threshold=3.887e+02, percent-clipped=1.0 +2023-03-21 10:24:34,764 INFO [train.py:901] (0/2) Epoch 41, batch 2700, loss[loss=0.1256, simple_loss=0.2077, pruned_loss=0.0217, over 7286.00 frames. ], tot_loss[loss=0.13, simple_loss=0.2114, pruned_loss=0.02427, over 1442776.24 frames. ], batch size: 68, lr: 4.00e-03, grad_scale: 4.0 +2023-03-21 10:24:59,546 INFO [train.py:901] (0/2) Epoch 41, batch 2750, loss[loss=0.1375, simple_loss=0.217, pruned_loss=0.02902, over 7260.00 frames. ], tot_loss[loss=0.1299, simple_loss=0.2114, pruned_loss=0.02416, over 1443958.91 frames. ], batch size: 64, lr: 4.00e-03, grad_scale: 4.0 +2023-03-21 10:25:07,108 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=115726.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:25:07,583 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=115727.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:25:14,887 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=115742.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:25:21,069 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.281e+02 1.781e+02 2.069e+02 2.360e+02 3.753e+02, threshold=4.137e+02, percent-clipped=0.0 +2023-03-21 10:25:23,958 INFO [train.py:901] (0/2) Epoch 41, batch 2800, loss[loss=0.1315, simple_loss=0.2144, pruned_loss=0.02436, over 7297.00 frames. ], tot_loss[loss=0.1296, simple_loss=0.2109, pruned_loss=0.02418, over 1442077.78 frames. ], batch size: 86, lr: 4.00e-03, grad_scale: 8.0 +2023-03-21 10:25:28,957 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5816, 1.9445, 1.4967, 1.8966, 1.9861, 1.7537, 1.6473, 1.5547], + device='cuda:0'), covar=tensor([0.0206, 0.0249, 0.0453, 0.0267, 0.0132, 0.0169, 0.0245, 0.0250], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0038, 0.0036, 0.0038, 0.0036, 0.0034, 0.0038, 0.0046], + device='cuda:0'), out_proj_covar=tensor([4.3565e-05, 4.2295e-05, 4.1218e-05, 4.1643e-05, 4.0007e-05, 3.8253e-05, + 4.2756e-05, 5.0445e-05], device='cuda:0') +2023-03-21 10:25:36,647 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-41.pt +2023-03-21 10:25:53,285 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 10:25:53,340 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 10:25:53,365 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 10:25:53,381 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 10:25:53,889 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 10:25:53,904 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 10:25:53,916 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 10:25:54,013 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 10:25:54,088 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 10:25:54,102 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 10:25:54,120 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 10:25:54,280 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 10:25:54,287 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 10:25:54,294 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 10:25:54,433 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 10:25:54,436 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 10:25:54,837 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 10:25:56,687 INFO [train.py:901] (0/2) Epoch 42, batch 0, loss[loss=0.1275, simple_loss=0.2123, pruned_loss=0.02134, over 7285.00 frames. ], tot_loss[loss=0.1275, simple_loss=0.2123, pruned_loss=0.02134, over 7285.00 frames. ], batch size: 66, lr: 3.95e-03, grad_scale: 8.0 +2023-03-21 10:25:56,688 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 10:26:06,879 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7001, 4.0165, 3.3486, 4.1297, 3.8441, 3.9562, 2.2985, 3.4564], + device='cuda:0'), covar=tensor([0.0387, 0.0797, 0.2039, 0.0468, 0.0360, 0.0472, 0.3139, 0.1458], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0258, 0.0277, 0.0267, 0.0270, 0.0264, 0.0232, 0.0254], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:26:12,728 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.3843, 5.1137, 4.7547, 5.4655, 5.1392, 5.4613, 5.1226, 5.1367], + device='cuda:0'), covar=tensor([0.0551, 0.1815, 0.1592, 0.0894, 0.0785, 0.0671, 0.0499, 0.0764], + device='cuda:0'), in_proj_covar=tensor([0.0270, 0.0405, 0.0307, 0.0321, 0.0238, 0.0381, 0.0238, 0.0284], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 10:26:20,060 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4942, 1.9189, 1.5310, 1.7829, 1.8759, 1.7479, 1.8529, 1.4490], + device='cuda:0'), covar=tensor([0.0316, 0.0172, 0.0249, 0.0216, 0.0147, 0.0266, 0.0144, 0.0197], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0038, 0.0036, 0.0038, 0.0036, 0.0034, 0.0038, 0.0046], + device='cuda:0'), out_proj_covar=tensor([4.3531e-05, 4.2313e-05, 4.1211e-05, 4.1674e-05, 4.0039e-05, 3.8308e-05, + 4.2724e-05, 5.0491e-05], device='cuda:0') +2023-03-21 10:26:22,423 INFO [train.py:935] (0/2) Epoch 42, validation: loss=0.166, simple_loss=0.2576, pruned_loss=0.03718, over 1622729.00 frames. +2023-03-21 10:26:22,424 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 10:26:25,059 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=115790.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:26:26,703 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8515, 2.6783, 2.7728, 2.7701, 2.6818, 2.6329, 2.8821, 2.0380], + device='cuda:0'), covar=tensor([0.0717, 0.0879, 0.0735, 0.0870, 0.0749, 0.1004, 0.0876, 0.2390], + device='cuda:0'), in_proj_covar=tensor([0.0327, 0.0334, 0.0269, 0.0352, 0.0283, 0.0283, 0.0346, 0.0241], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:26:29,997 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 10:26:41,189 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 10:26:47,218 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 10:26:47,358 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4093, 2.3379, 2.3725, 3.5855, 1.9089, 3.5038, 1.3995, 3.3548], + device='cuda:0'), covar=tensor([0.0210, 0.1522, 0.1881, 0.0244, 0.4069, 0.0247, 0.1445, 0.0367], + device='cuda:0'), in_proj_covar=tensor([0.0207, 0.0245, 0.0258, 0.0212, 0.0250, 0.0216, 0.0224, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:26:49,217 INFO [train.py:901] (0/2) Epoch 42, batch 50, loss[loss=0.1337, simple_loss=0.2173, pruned_loss=0.02505, over 7313.00 frames. ], tot_loss[loss=0.1309, simple_loss=0.212, pruned_loss=0.02488, over 326259.99 frames. ], batch size: 75, lr: 3.95e-03, grad_scale: 8.0 +2023-03-21 10:26:49,240 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 10:26:52,198 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 10:26:59,257 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.182e+02 1.683e+02 1.944e+02 2.243e+02 4.177e+02, threshold=3.887e+02, percent-clipped=1.0 +2023-03-21 10:27:00,433 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3960, 1.7743, 1.4676, 1.7646, 1.7590, 1.7883, 1.7058, 1.5261], + device='cuda:0'), covar=tensor([0.0338, 0.0213, 0.0369, 0.0239, 0.0184, 0.0187, 0.0199, 0.0207], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0038, 0.0036, 0.0038, 0.0036, 0.0034, 0.0038, 0.0045], + device='cuda:0'), out_proj_covar=tensor([4.3229e-05, 4.2164e-05, 4.0972e-05, 4.1525e-05, 3.9846e-05, 3.8130e-05, + 4.2362e-05, 5.0040e-05], device='cuda:0') +2023-03-21 10:27:09,408 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 10:27:09,868 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 10:27:12,095 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0178, 2.4387, 1.9381, 2.8778, 2.7982, 2.7589, 2.5561, 2.6905], + device='cuda:0'), covar=tensor([0.2150, 0.1153, 0.3759, 0.0656, 0.0316, 0.0331, 0.0466, 0.0490], + device='cuda:0'), in_proj_covar=tensor([0.0251, 0.0232, 0.0248, 0.0258, 0.0201, 0.0204, 0.0222, 0.0228], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:27:14,896 INFO [train.py:901] (0/2) Epoch 42, batch 100, loss[loss=0.125, simple_loss=0.2069, pruned_loss=0.02156, over 7300.00 frames. ], tot_loss[loss=0.1314, simple_loss=0.2123, pruned_loss=0.02525, over 573509.60 frames. ], batch size: 80, lr: 3.95e-03, grad_scale: 8.0 +2023-03-21 10:27:18,712 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0471, 2.9316, 2.9368, 2.9818, 2.7737, 2.6743, 3.1246, 2.2182], + device='cuda:0'), covar=tensor([0.0575, 0.0647, 0.0847, 0.0818, 0.0716, 0.1094, 0.0882, 0.2540], + device='cuda:0'), in_proj_covar=tensor([0.0326, 0.0333, 0.0269, 0.0352, 0.0283, 0.0284, 0.0345, 0.0241], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:27:19,169 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=115892.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 10:27:28,075 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6660, 2.7085, 2.5549, 3.6970, 2.0128, 3.6350, 1.5020, 3.2918], + device='cuda:0'), covar=tensor([0.0257, 0.1386, 0.2214, 0.0282, 0.4662, 0.0310, 0.1528, 0.0491], + device='cuda:0'), in_proj_covar=tensor([0.0207, 0.0245, 0.0259, 0.0213, 0.0251, 0.0217, 0.0225, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:27:40,730 INFO [train.py:901] (0/2) Epoch 42, batch 150, loss[loss=0.1327, simple_loss=0.2164, pruned_loss=0.02453, over 7277.00 frames. ], tot_loss[loss=0.1282, simple_loss=0.2095, pruned_loss=0.0234, over 767254.65 frames. ], batch size: 57, lr: 3.95e-03, grad_scale: 8.0 +2023-03-21 10:27:42,612 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.84 vs. limit=5.0 +2023-03-21 10:27:43,887 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3622, 3.5712, 3.2993, 3.5500, 3.2616, 3.3770, 3.6471, 3.7222], + device='cuda:0'), covar=tensor([0.0284, 0.0171, 0.0268, 0.0177, 0.0338, 0.0612, 0.0326, 0.0232], + device='cuda:0'), in_proj_covar=tensor([0.0121, 0.0124, 0.0119, 0.0122, 0.0111, 0.0099, 0.0098, 0.0099], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 10:27:49,873 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=115953.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 10:27:50,693 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.221e+02 1.790e+02 2.023e+02 2.356e+02 3.555e+02, threshold=4.045e+02, percent-clipped=0.0 +2023-03-21 10:27:55,022 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.63 vs. limit=2.0 +2023-03-21 10:28:07,101 INFO [train.py:901] (0/2) Epoch 42, batch 200, loss[loss=0.1389, simple_loss=0.2162, pruned_loss=0.0308, over 7291.00 frames. ], tot_loss[loss=0.1285, simple_loss=0.2099, pruned_loss=0.02357, over 914723.22 frames. ], batch size: 57, lr: 3.95e-03, grad_scale: 8.0 +2023-03-21 10:28:12,623 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 10:28:14,956 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-116000.pt +2023-03-21 10:28:21,048 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 10:28:21,835 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.90 vs. limit=2.0 +2023-03-21 10:28:25,698 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=116014.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:28:27,061 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 10:28:31,698 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=116026.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:28:32,145 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=116027.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:28:36,000 INFO [train.py:901] (0/2) Epoch 42, batch 250, loss[loss=0.154, simple_loss=0.2383, pruned_loss=0.0348, over 7213.00 frames. ], tot_loss[loss=0.1287, simple_loss=0.2096, pruned_loss=0.02386, over 1028634.93 frames. ], batch size: 93, lr: 3.95e-03, grad_scale: 8.0 +2023-03-21 10:28:40,504 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 10:28:46,480 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.130e+02 1.767e+02 2.086e+02 2.345e+02 3.684e+02, threshold=4.171e+02, percent-clipped=0.0 +2023-03-21 10:28:56,703 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=116074.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:28:57,202 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=116075.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:28:57,283 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=116075.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 10:29:00,578 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 10:29:01,158 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7399, 2.5658, 2.6490, 3.8165, 2.0609, 3.7361, 1.6708, 3.4603], + device='cuda:0'), covar=tensor([0.0198, 0.1342, 0.1824, 0.0231, 0.4008, 0.0284, 0.1254, 0.0397], + device='cuda:0'), in_proj_covar=tensor([0.0207, 0.0244, 0.0258, 0.0213, 0.0251, 0.0217, 0.0225, 0.0228], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:29:02,018 INFO [train.py:901] (0/2) Epoch 42, batch 300, loss[loss=0.1349, simple_loss=0.2151, pruned_loss=0.02737, over 7352.00 frames. ], tot_loss[loss=0.129, simple_loss=0.2101, pruned_loss=0.02396, over 1121623.29 frames. ], batch size: 73, lr: 3.95e-03, grad_scale: 8.0 +2023-03-21 10:29:08,739 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 10:29:11,883 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=116105.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:29:17,441 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=116116.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:29:26,750 INFO [train.py:901] (0/2) Epoch 42, batch 350, loss[loss=0.1387, simple_loss=0.2249, pruned_loss=0.02619, over 7324.00 frames. ], tot_loss[loss=0.1289, simple_loss=0.21, pruned_loss=0.02392, over 1193293.53 frames. ], batch size: 83, lr: 3.95e-03, grad_scale: 8.0 +2023-03-21 10:29:37,068 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.345e+02 1.856e+02 2.238e+02 2.554e+02 4.573e+02, threshold=4.477e+02, percent-clipped=1.0 +2023-03-21 10:29:39,064 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.81 vs. limit=2.0 +2023-03-21 10:29:43,316 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 10:29:43,458 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=116166.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:29:48,932 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5421, 5.0615, 5.1465, 5.1151, 4.9176, 4.6126, 5.1543, 4.9622], + device='cuda:0'), covar=tensor([0.0447, 0.0377, 0.0358, 0.0441, 0.0338, 0.0421, 0.0343, 0.0502], + device='cuda:0'), in_proj_covar=tensor([0.0224, 0.0265, 0.0205, 0.0207, 0.0161, 0.0235, 0.0217, 0.0151], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:29:48,996 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=116177.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:29:52,895 INFO [train.py:901] (0/2) Epoch 42, batch 400, loss[loss=0.1429, simple_loss=0.2239, pruned_loss=0.03095, over 7247.00 frames. ], tot_loss[loss=0.1296, simple_loss=0.2106, pruned_loss=0.02427, over 1247877.09 frames. ], batch size: 89, lr: 3.95e-03, grad_scale: 8.0 +2023-03-21 10:30:12,028 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9427, 2.6892, 2.9208, 2.9480, 2.6293, 2.6979, 3.1857, 2.2781], + device='cuda:0'), covar=tensor([0.0604, 0.0758, 0.0819, 0.0692, 0.0735, 0.1097, 0.0749, 0.2504], + device='cuda:0'), in_proj_covar=tensor([0.0326, 0.0333, 0.0269, 0.0350, 0.0282, 0.0283, 0.0345, 0.0240], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:30:18,819 INFO [train.py:901] (0/2) Epoch 42, batch 450, loss[loss=0.1379, simple_loss=0.2138, pruned_loss=0.03104, over 7314.00 frames. ], tot_loss[loss=0.1298, simple_loss=0.2109, pruned_loss=0.02437, over 1292803.36 frames. ], batch size: 49, lr: 3.94e-03, grad_scale: 8.0 +2023-03-21 10:30:22,442 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.58 vs. limit=2.0 +2023-03-21 10:30:25,490 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 10:30:25,504 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 10:30:26,090 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=116248.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 10:30:29,974 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.379e+02 1.720e+02 2.089e+02 2.447e+02 6.240e+02, threshold=4.178e+02, percent-clipped=2.0 +2023-03-21 10:30:44,578 INFO [train.py:901] (0/2) Epoch 42, batch 500, loss[loss=0.1178, simple_loss=0.2009, pruned_loss=0.01733, over 7321.00 frames. ], tot_loss[loss=0.1295, simple_loss=0.2103, pruned_loss=0.02434, over 1324832.17 frames. ], batch size: 59, lr: 3.94e-03, grad_scale: 4.0 +2023-03-21 10:30:57,102 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 10:30:59,089 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 10:31:00,100 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 10:31:01,631 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 10:31:04,298 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=116323.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:31:07,763 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 10:31:10,708 INFO [train.py:901] (0/2) Epoch 42, batch 550, loss[loss=0.1253, simple_loss=0.2104, pruned_loss=0.02008, over 7209.00 frames. ], tot_loss[loss=0.1295, simple_loss=0.2103, pruned_loss=0.02439, over 1349870.68 frames. ], batch size: 93, lr: 3.94e-03, grad_scale: 4.0 +2023-03-21 10:31:12,381 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2983, 4.7756, 4.8210, 4.8020, 4.7194, 4.3618, 4.8518, 4.7057], + device='cuda:0'), covar=tensor([0.0434, 0.0378, 0.0394, 0.0465, 0.0292, 0.0410, 0.0332, 0.0405], + device='cuda:0'), in_proj_covar=tensor([0.0220, 0.0261, 0.0203, 0.0205, 0.0159, 0.0232, 0.0215, 0.0149], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:31:19,566 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 10:31:21,543 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.129e+02 1.746e+02 2.072e+02 2.428e+02 3.204e+02, threshold=4.144e+02, percent-clipped=0.0 +2023-03-21 10:31:27,586 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 10:31:28,683 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=116370.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 10:31:31,142 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 10:31:32,438 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.49 vs. limit=5.0 +2023-03-21 10:31:35,736 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=116384.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:31:36,112 INFO [train.py:901] (0/2) Epoch 42, batch 600, loss[loss=0.1312, simple_loss=0.2054, pruned_loss=0.02846, over 7327.00 frames. ], tot_loss[loss=0.1296, simple_loss=0.2102, pruned_loss=0.02445, over 1369167.00 frames. ], batch size: 51, lr: 3.94e-03, grad_scale: 4.0 +2023-03-21 10:31:37,626 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 10:31:45,685 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7778, 1.5864, 2.0598, 2.1358, 2.0123, 2.1004, 1.9321, 2.3215], + device='cuda:0'), covar=tensor([0.2941, 0.3411, 0.1813, 0.1488, 0.1890, 0.2542, 0.1679, 0.3029], + device='cuda:0'), in_proj_covar=tensor([0.0081, 0.0081, 0.0074, 0.0067, 0.0067, 0.0066, 0.0107, 0.0071], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:31:48,241 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4461, 2.3471, 2.3476, 3.6225, 1.9494, 3.3920, 1.3460, 3.4294], + device='cuda:0'), covar=tensor([0.0250, 0.1433, 0.1862, 0.0259, 0.3897, 0.0333, 0.1345, 0.0364], + device='cuda:0'), in_proj_covar=tensor([0.0209, 0.0246, 0.0259, 0.0213, 0.0252, 0.0219, 0.0227, 0.0230], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:31:54,734 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 10:31:55,563 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.43 vs. limit=2.0 +2023-03-21 10:32:02,799 INFO [train.py:901] (0/2) Epoch 42, batch 650, loss[loss=0.1196, simple_loss=0.207, pruned_loss=0.0161, over 7297.00 frames. ], tot_loss[loss=0.1294, simple_loss=0.2103, pruned_loss=0.02421, over 1384608.65 frames. ], batch size: 80, lr: 3.94e-03, grad_scale: 4.0 +2023-03-21 10:32:03,772 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 10:32:08,272 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=116446.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:32:13,285 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.271e+02 1.770e+02 2.062e+02 2.474e+02 7.007e+02, threshold=4.123e+02, percent-clipped=2.0 +2023-03-21 10:32:15,957 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=116461.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:32:20,997 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 10:32:21,542 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=116472.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:32:27,958 INFO [train.py:901] (0/2) Epoch 42, batch 700, loss[loss=0.1264, simple_loss=0.2088, pruned_loss=0.02201, over 7316.00 frames. ], tot_loss[loss=0.1293, simple_loss=0.2103, pruned_loss=0.02418, over 1396364.64 frames. ], batch size: 83, lr: 3.94e-03, grad_scale: 4.0 +2023-03-21 10:32:28,524 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 10:32:34,938 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.13 vs. limit=5.0 +2023-03-21 10:32:39,428 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=116505.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:32:40,487 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=116507.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 10:32:53,175 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.62 vs. limit=5.0 +2023-03-21 10:32:54,347 INFO [train.py:901] (0/2) Epoch 42, batch 750, loss[loss=0.1281, simple_loss=0.2163, pruned_loss=0.0199, over 7317.00 frames. ], tot_loss[loss=0.1291, simple_loss=0.2105, pruned_loss=0.02387, over 1407721.16 frames. ], batch size: 59, lr: 3.94e-03, grad_scale: 4.0 +2023-03-21 10:32:54,351 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 10:32:54,889 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 10:32:58,520 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3791, 3.6513, 2.6513, 3.9339, 3.2213, 3.6138, 1.8589, 2.7367], + device='cuda:0'), covar=tensor([0.0560, 0.0944, 0.2595, 0.0608, 0.0468, 0.0749, 0.4012, 0.1912], + device='cuda:0'), in_proj_covar=tensor([0.0280, 0.0262, 0.0282, 0.0270, 0.0274, 0.0267, 0.0235, 0.0259], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:33:00,948 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=116548.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 10:33:04,816 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.244e+02 1.798e+02 2.124e+02 2.350e+02 4.773e+02, threshold=4.247e+02, percent-clipped=1.0 +2023-03-21 10:33:08,810 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 10:33:09,932 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=116566.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:33:12,847 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 10:33:18,806 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.42 vs. limit=2.0 +2023-03-21 10:33:19,499 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 10:33:19,996 INFO [train.py:901] (0/2) Epoch 42, batch 800, loss[loss=0.123, simple_loss=0.2013, pruned_loss=0.02236, over 7327.00 frames. ], tot_loss[loss=0.1293, simple_loss=0.2104, pruned_loss=0.02405, over 1416848.30 frames. ], batch size: 61, lr: 3.94e-03, grad_scale: 8.0 +2023-03-21 10:33:21,163 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 10:33:26,290 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=116596.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 10:33:31,701 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 10:33:33,273 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0016, 3.7455, 3.6805, 3.6841, 3.6709, 3.5619, 3.8347, 3.4095], + device='cuda:0'), covar=tensor([0.0140, 0.0188, 0.0134, 0.0216, 0.0397, 0.0125, 0.0170, 0.0225], + device='cuda:0'), in_proj_covar=tensor([0.0096, 0.0104, 0.0104, 0.0089, 0.0178, 0.0108, 0.0108, 0.0114], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 10:33:35,817 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5852, 2.3572, 2.3770, 3.7509, 2.0075, 3.4974, 1.4372, 3.3194], + device='cuda:0'), covar=tensor([0.0195, 0.1397, 0.1904, 0.0238, 0.3586, 0.0290, 0.1358, 0.0403], + device='cuda:0'), in_proj_covar=tensor([0.0207, 0.0243, 0.0258, 0.0211, 0.0249, 0.0218, 0.0225, 0.0228], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:33:37,753 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=116618.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:33:43,884 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.76 vs. limit=2.0 +2023-03-21 10:33:46,114 INFO [train.py:901] (0/2) Epoch 42, batch 850, loss[loss=0.1341, simple_loss=0.2129, pruned_loss=0.02767, over 7216.00 frames. ], tot_loss[loss=0.1287, simple_loss=0.2097, pruned_loss=0.02383, over 1422602.90 frames. ], batch size: 50, lr: 3.94e-03, grad_scale: 8.0 +2023-03-21 10:33:50,648 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 10:33:50,657 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 10:33:56,160 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 10:33:56,612 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.288e+02 1.744e+02 2.049e+02 2.430e+02 4.238e+02, threshold=4.097e+02, percent-clipped=0.0 +2023-03-21 10:34:00,121 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 10:34:04,268 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=116670.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 10:34:09,207 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=116679.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:34:09,262 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=116679.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:34:12,095 INFO [train.py:901] (0/2) Epoch 42, batch 900, loss[loss=0.1305, simple_loss=0.2126, pruned_loss=0.02419, over 7251.00 frames. ], tot_loss[loss=0.1291, simple_loss=0.2102, pruned_loss=0.02397, over 1428617.33 frames. ], batch size: 55, lr: 3.94e-03, grad_scale: 8.0 +2023-03-21 10:34:21,790 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=116704.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:34:28,559 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=116718.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:34:37,085 INFO [train.py:901] (0/2) Epoch 42, batch 950, loss[loss=0.1312, simple_loss=0.217, pruned_loss=0.02273, over 7305.00 frames. ], tot_loss[loss=0.1297, simple_loss=0.2109, pruned_loss=0.02422, over 1432151.58 frames. ], batch size: 83, lr: 3.94e-03, grad_scale: 8.0 +2023-03-21 10:34:37,106 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 10:34:48,299 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.041e+02 1.811e+02 2.090e+02 2.390e+02 5.316e+02, threshold=4.179e+02, percent-clipped=3.0 +2023-03-21 10:34:51,609 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=116761.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:34:53,686 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=116765.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:34:57,213 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=116772.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:35:02,130 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 10:35:03,562 INFO [train.py:901] (0/2) Epoch 42, batch 1000, loss[loss=0.1295, simple_loss=0.2115, pruned_loss=0.02378, over 7288.00 frames. ], tot_loss[loss=0.1292, simple_loss=0.2105, pruned_loss=0.02397, over 1435561.78 frames. ], batch size: 57, lr: 3.94e-03, grad_scale: 8.0 +2023-03-21 10:35:12,500 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=116802.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 10:35:12,541 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=116802.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:35:16,086 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=116809.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:35:21,585 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 10:35:21,640 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=116820.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:35:29,791 INFO [train.py:901] (0/2) Epoch 42, batch 1050, loss[loss=0.1276, simple_loss=0.2153, pruned_loss=0.01999, over 7274.00 frames. ], tot_loss[loss=0.1287, simple_loss=0.2099, pruned_loss=0.0237, over 1436245.85 frames. ], batch size: 52, lr: 3.93e-03, grad_scale: 8.0 +2023-03-21 10:35:30,913 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.0404, 4.4044, 4.6921, 4.6344, 4.6071, 4.4581, 4.8293, 4.2206], + device='cuda:0'), covar=tensor([0.0099, 0.0169, 0.0094, 0.0155, 0.0349, 0.0121, 0.0120, 0.0195], + device='cuda:0'), in_proj_covar=tensor([0.0096, 0.0105, 0.0105, 0.0090, 0.0180, 0.0109, 0.0108, 0.0116], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 10:35:40,883 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.209e+02 1.773e+02 2.097e+02 2.514e+02 5.231e+02, threshold=4.194e+02, percent-clipped=1.0 +2023-03-21 10:35:43,578 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=116861.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:35:44,033 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 10:35:44,691 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=116863.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:35:48,135 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 10:35:55,650 INFO [train.py:901] (0/2) Epoch 42, batch 1100, loss[loss=0.1152, simple_loss=0.1942, pruned_loss=0.01807, over 7361.00 frames. ], tot_loss[loss=0.1286, simple_loss=0.2098, pruned_loss=0.02368, over 1434554.42 frames. ], batch size: 73, lr: 3.93e-03, grad_scale: 8.0 +2023-03-21 10:36:16,931 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 10:36:16,944 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 10:36:21,980 INFO [train.py:901] (0/2) Epoch 42, batch 1150, loss[loss=0.1201, simple_loss=0.2002, pruned_loss=0.01998, over 7341.00 frames. ], tot_loss[loss=0.129, simple_loss=0.2104, pruned_loss=0.02383, over 1437714.56 frames. ], batch size: 44, lr: 3.93e-03, grad_scale: 8.0 +2023-03-21 10:36:29,611 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 10:36:30,144 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 10:36:32,583 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.106e+02 1.641e+02 1.959e+02 2.194e+02 3.203e+02, threshold=3.918e+02, percent-clipped=0.0 +2023-03-21 10:36:41,704 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=116974.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:36:44,308 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=116979.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:36:47,270 INFO [train.py:901] (0/2) Epoch 42, batch 1200, loss[loss=0.1327, simple_loss=0.2155, pruned_loss=0.02495, over 7320.00 frames. ], tot_loss[loss=0.1288, simple_loss=0.2102, pruned_loss=0.02369, over 1439283.98 frames. ], batch size: 49, lr: 3.93e-03, grad_scale: 8.0 +2023-03-21 10:36:49,108 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.62 vs. limit=2.0 +2023-03-21 10:37:02,609 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.43 vs. limit=2.0 +2023-03-21 10:37:02,932 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9921, 2.1667, 2.2824, 2.5117, 2.2569, 2.4257, 2.5747, 2.4456], + device='cuda:0'), covar=tensor([0.3450, 0.4224, 0.2301, 0.1919, 0.4345, 0.2493, 0.1934, 0.5563], + device='cuda:0'), in_proj_covar=tensor([0.0081, 0.0081, 0.0074, 0.0068, 0.0068, 0.0066, 0.0107, 0.0071], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:37:03,344 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 10:37:10,012 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=117027.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:37:14,028 INFO [train.py:901] (0/2) Epoch 42, batch 1250, loss[loss=0.1363, simple_loss=0.2206, pruned_loss=0.02599, over 7389.00 frames. ], tot_loss[loss=0.1287, simple_loss=0.2103, pruned_loss=0.0236, over 1438915.34 frames. ], batch size: 65, lr: 3.93e-03, grad_scale: 8.0 +2023-03-21 10:37:16,073 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1382, 4.3705, 4.1419, 4.3797, 3.9769, 4.3066, 4.6325, 4.6693], + device='cuda:0'), covar=tensor([0.0210, 0.0127, 0.0193, 0.0156, 0.0285, 0.0285, 0.0213, 0.0179], + device='cuda:0'), in_proj_covar=tensor([0.0125, 0.0126, 0.0121, 0.0126, 0.0114, 0.0102, 0.0099, 0.0102], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 10:37:18,395 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.61 vs. limit=2.0 +2023-03-21 10:37:24,586 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.243e+02 1.700e+02 1.978e+02 2.314e+02 4.583e+02, threshold=3.956e+02, percent-clipped=1.0 +2023-03-21 10:37:26,664 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 10:37:26,731 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=117060.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:37:30,569 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 10:37:31,579 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 10:37:35,447 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.79 vs. limit=2.0 +2023-03-21 10:37:39,135 INFO [train.py:901] (0/2) Epoch 42, batch 1300, loss[loss=0.1305, simple_loss=0.2145, pruned_loss=0.02324, over 7320.00 frames. ], tot_loss[loss=0.1287, simple_loss=0.21, pruned_loss=0.02366, over 1438871.47 frames. ], batch size: 61, lr: 3.93e-03, grad_scale: 8.0 +2023-03-21 10:37:48,346 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=117102.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:37:56,070 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 10:37:58,649 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 10:38:02,271 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 10:38:05,802 INFO [train.py:901] (0/2) Epoch 42, batch 1350, loss[loss=0.1342, simple_loss=0.2144, pruned_loss=0.027, over 7211.00 frames. ], tot_loss[loss=0.1292, simple_loss=0.2103, pruned_loss=0.02398, over 1439068.05 frames. ], batch size: 50, lr: 3.93e-03, grad_scale: 8.0 +2023-03-21 10:38:12,865 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 10:38:13,416 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=117150.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:38:16,420 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.067e+02 1.654e+02 1.903e+02 2.260e+02 3.154e+02, threshold=3.806e+02, percent-clipped=0.0 +2023-03-21 10:38:17,592 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=117158.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:38:19,093 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=117161.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:38:31,669 INFO [train.py:901] (0/2) Epoch 42, batch 1400, loss[loss=0.1369, simple_loss=0.225, pruned_loss=0.02439, over 7262.00 frames. ], tot_loss[loss=0.1286, simple_loss=0.2097, pruned_loss=0.0237, over 1437928.61 frames. ], batch size: 89, lr: 3.93e-03, grad_scale: 8.0 +2023-03-21 10:38:44,771 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=117209.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:38:47,283 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 10:38:50,882 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=117221.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:38:57,826 INFO [train.py:901] (0/2) Epoch 42, batch 1450, loss[loss=0.1147, simple_loss=0.1822, pruned_loss=0.02357, over 7036.00 frames. ], tot_loss[loss=0.1289, simple_loss=0.2098, pruned_loss=0.02397, over 1437091.16 frames. ], batch size: 35, lr: 3.93e-03, grad_scale: 8.0 +2023-03-21 10:39:05,196 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.97 vs. limit=5.0 +2023-03-21 10:39:08,249 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.082e+02 1.818e+02 2.119e+02 2.503e+02 5.217e+02, threshold=4.239e+02, percent-clipped=4.0 +2023-03-21 10:39:08,854 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2001, 4.4880, 4.1955, 4.4731, 4.0153, 4.3348, 4.7420, 4.7624], + device='cuda:0'), covar=tensor([0.0234, 0.0143, 0.0237, 0.0149, 0.0400, 0.0321, 0.0202, 0.0172], + device='cuda:0'), in_proj_covar=tensor([0.0124, 0.0126, 0.0121, 0.0125, 0.0114, 0.0102, 0.0098, 0.0101], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 10:39:09,825 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 10:39:18,062 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=117274.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:39:22,740 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=117282.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:39:24,106 INFO [train.py:901] (0/2) Epoch 42, batch 1500, loss[loss=0.1295, simple_loss=0.2156, pruned_loss=0.02166, over 7263.00 frames. ], tot_loss[loss=0.1294, simple_loss=0.2104, pruned_loss=0.02413, over 1439125.94 frames. ], batch size: 89, lr: 3.93e-03, grad_scale: 8.0 +2023-03-21 10:39:27,144 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 10:39:42,603 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=117322.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:39:46,143 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=117329.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:39:48,936 INFO [train.py:901] (0/2) Epoch 42, batch 1550, loss[loss=0.1512, simple_loss=0.2392, pruned_loss=0.03155, over 6702.00 frames. ], tot_loss[loss=0.1298, simple_loss=0.2111, pruned_loss=0.02423, over 1441359.62 frames. ], batch size: 106, lr: 3.93e-03, grad_scale: 8.0 +2023-03-21 10:39:48,950 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 10:40:00,047 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.131e+02 1.700e+02 1.983e+02 2.415e+02 3.192e+02, threshold=3.965e+02, percent-clipped=0.0 +2023-03-21 10:40:02,264 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=117360.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:40:15,306 INFO [train.py:901] (0/2) Epoch 42, batch 1600, loss[loss=0.1316, simple_loss=0.2115, pruned_loss=0.02584, over 7367.00 frames. ], tot_loss[loss=0.1296, simple_loss=0.2109, pruned_loss=0.02409, over 1440819.07 frames. ], batch size: 73, lr: 3.92e-03, grad_scale: 8.0 +2023-03-21 10:40:17,906 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=117390.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:40:20,373 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 10:40:20,866 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 10:40:24,266 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 10:40:27,403 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=117408.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:40:31,955 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6859, 1.7309, 1.7749, 2.0473, 1.7643, 1.8649, 1.5754, 2.1303], + device='cuda:0'), covar=tensor([0.2201, 0.2708, 0.1140, 0.0828, 0.1706, 0.1815, 0.1876, 0.1118], + device='cuda:0'), in_proj_covar=tensor([0.0081, 0.0081, 0.0074, 0.0067, 0.0067, 0.0065, 0.0106, 0.0071], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:40:33,770 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 10:40:38,336 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 10:40:41,381 INFO [train.py:901] (0/2) Epoch 42, batch 1650, loss[loss=0.1388, simple_loss=0.2245, pruned_loss=0.02659, over 7336.00 frames. ], tot_loss[loss=0.1297, simple_loss=0.211, pruned_loss=0.02423, over 1440632.89 frames. ], batch size: 73, lr: 3.92e-03, grad_scale: 8.0 +2023-03-21 10:40:45,792 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 10:40:51,673 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.314e+02 1.848e+02 2.191e+02 2.608e+02 4.497e+02, threshold=4.382e+02, percent-clipped=1.0 +2023-03-21 10:40:53,384 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=117458.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:40:59,697 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.74 vs. limit=2.0 +2023-03-21 10:41:01,857 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8436, 4.2727, 4.2383, 4.7602, 4.6146, 4.7350, 4.3998, 4.3820], + device='cuda:0'), covar=tensor([0.0865, 0.2697, 0.2234, 0.1145, 0.0906, 0.1162, 0.0707, 0.1204], + device='cuda:0'), in_proj_covar=tensor([0.0268, 0.0403, 0.0306, 0.0320, 0.0236, 0.0377, 0.0234, 0.0284], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 10:41:03,873 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 10:41:06,957 INFO [train.py:901] (0/2) Epoch 42, batch 1700, loss[loss=0.142, simple_loss=0.2226, pruned_loss=0.03067, over 7272.00 frames. ], tot_loss[loss=0.1302, simple_loss=0.2114, pruned_loss=0.02446, over 1440603.15 frames. ], batch size: 64, lr: 3.92e-03, grad_scale: 8.0 +2023-03-21 10:41:07,977 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 10:41:12,501 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2404, 4.0676, 3.4034, 3.8029, 3.0532, 2.2254, 2.1175, 4.2487], + device='cuda:0'), covar=tensor([0.0045, 0.0064, 0.0142, 0.0074, 0.0176, 0.0594, 0.0604, 0.0048], + device='cuda:0'), in_proj_covar=tensor([0.0117, 0.0095, 0.0114, 0.0097, 0.0133, 0.0136, 0.0130, 0.0107], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 10:41:17,489 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=117506.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:41:19,032 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 10:41:32,723 INFO [train.py:901] (0/2) Epoch 42, batch 1750, loss[loss=0.1349, simple_loss=0.2174, pruned_loss=0.02621, over 7199.00 frames. ], tot_loss[loss=0.1298, simple_loss=0.2111, pruned_loss=0.02428, over 1442144.57 frames. ], batch size: 50, lr: 3.92e-03, grad_scale: 8.0 +2023-03-21 10:41:43,727 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.161e+02 1.708e+02 1.965e+02 2.274e+02 3.472e+02, threshold=3.931e+02, percent-clipped=0.0 +2023-03-21 10:41:43,775 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 10:41:44,787 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 10:41:47,618 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-03-21 10:41:54,394 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=117577.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:41:56,093 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.27 vs. limit=2.0 +2023-03-21 10:41:58,332 INFO [train.py:901] (0/2) Epoch 42, batch 1800, loss[loss=0.1325, simple_loss=0.2169, pruned_loss=0.02405, over 7291.00 frames. ], tot_loss[loss=0.1298, simple_loss=0.2109, pruned_loss=0.02437, over 1442309.46 frames. ], batch size: 68, lr: 3.92e-03, grad_scale: 8.0 +2023-03-21 10:42:06,707 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 10:42:20,327 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 10:42:24,843 INFO [train.py:901] (0/2) Epoch 42, batch 1850, loss[loss=0.1331, simple_loss=0.2158, pruned_loss=0.02524, over 7186.00 frames. ], tot_loss[loss=0.1296, simple_loss=0.2106, pruned_loss=0.02428, over 1442986.66 frames. ], batch size: 99, lr: 3.92e-03, grad_scale: 8.0 +2023-03-21 10:42:26,211 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.62 vs. limit=2.0 +2023-03-21 10:42:29,254 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.81 vs. limit=2.0 +2023-03-21 10:42:30,312 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 10:42:35,311 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.137e+02 1.716e+02 2.093e+02 2.337e+02 4.292e+02, threshold=4.185e+02, percent-clipped=1.0 +2023-03-21 10:42:35,954 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2555, 4.0694, 3.6476, 3.8213, 2.9387, 2.0500, 1.8997, 4.2339], + device='cuda:0'), covar=tensor([0.0067, 0.0096, 0.0154, 0.0085, 0.0258, 0.0773, 0.0792, 0.0068], + device='cuda:0'), in_proj_covar=tensor([0.0117, 0.0095, 0.0114, 0.0097, 0.0133, 0.0136, 0.0130, 0.0107], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 10:42:36,450 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9036, 3.2243, 3.9641, 3.8081, 3.9856, 3.9706, 3.9882, 3.8901], + device='cuda:0'), covar=tensor([0.0034, 0.0109, 0.0027, 0.0033, 0.0027, 0.0026, 0.0034, 0.0045], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0071, 0.0058, 0.0058, 0.0056, 0.0061, 0.0050, 0.0079], + device='cuda:0'), out_proj_covar=tensor([8.4035e-05, 1.4287e-04, 1.0386e-04, 9.8955e-05, 9.4158e-05, 1.0569e-04, + 9.3611e-05, 1.4348e-04], device='cuda:0') +2023-03-21 10:42:45,436 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 10:42:47,723 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.75 vs. limit=2.0 +2023-03-21 10:42:49,990 INFO [train.py:901] (0/2) Epoch 42, batch 1900, loss[loss=0.1195, simple_loss=0.2043, pruned_loss=0.01731, over 7283.00 frames. ], tot_loss[loss=0.1307, simple_loss=0.2121, pruned_loss=0.02463, over 1443529.50 frames. ], batch size: 66, lr: 3.92e-03, grad_scale: 8.0 +2023-03-21 10:42:50,077 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=117685.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:43:05,570 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-03-21 10:43:10,337 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 10:43:16,249 INFO [train.py:901] (0/2) Epoch 42, batch 1950, loss[loss=0.1345, simple_loss=0.2231, pruned_loss=0.023, over 6741.00 frames. ], tot_loss[loss=0.1309, simple_loss=0.2123, pruned_loss=0.02481, over 1441692.68 frames. ], batch size: 106, lr: 3.92e-03, grad_scale: 8.0 +2023-03-21 10:43:21,216 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 10:43:26,199 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 10:43:26,666 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.244e+02 1.776e+02 2.090e+02 2.427e+02 5.955e+02, threshold=4.180e+02, percent-clipped=1.0 +2023-03-21 10:43:26,709 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 10:43:41,889 INFO [train.py:901] (0/2) Epoch 42, batch 2000, loss[loss=0.1385, simple_loss=0.2214, pruned_loss=0.02782, over 7325.00 frames. ], tot_loss[loss=0.13, simple_loss=0.2111, pruned_loss=0.02443, over 1441566.81 frames. ], batch size: 75, lr: 3.92e-03, grad_scale: 8.0 +2023-03-21 10:43:43,969 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 10:43:52,812 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2683, 2.6288, 2.1310, 2.9959, 3.1063, 2.8829, 2.5591, 2.6736], + device='cuda:0'), covar=tensor([0.2198, 0.1221, 0.3951, 0.0636, 0.0314, 0.0395, 0.0483, 0.0568], + device='cuda:0'), in_proj_covar=tensor([0.0251, 0.0231, 0.0245, 0.0257, 0.0199, 0.0203, 0.0219, 0.0228], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:43:56,172 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 10:44:04,196 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 10:44:08,068 INFO [train.py:901] (0/2) Epoch 42, batch 2050, loss[loss=0.09598, simple_loss=0.1696, pruned_loss=0.01117, over 6963.00 frames. ], tot_loss[loss=0.1284, simple_loss=0.2094, pruned_loss=0.02374, over 1439978.28 frames. ], batch size: 35, lr: 3.92e-03, grad_scale: 8.0 +2023-03-21 10:44:17,862 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.58 vs. limit=2.0 +2023-03-21 10:44:18,467 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.201e+02 1.814e+02 2.172e+02 2.477e+02 4.017e+02, threshold=4.343e+02, percent-clipped=0.0 +2023-03-21 10:44:26,665 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3144, 3.9575, 3.9698, 3.9514, 3.9267, 3.8154, 4.1520, 3.5172], + device='cuda:0'), covar=tensor([0.0136, 0.0175, 0.0138, 0.0186, 0.0428, 0.0150, 0.0165, 0.0245], + device='cuda:0'), in_proj_covar=tensor([0.0096, 0.0105, 0.0105, 0.0090, 0.0180, 0.0110, 0.0109, 0.0116], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 10:44:29,764 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=117877.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:44:33,672 INFO [train.py:901] (0/2) Epoch 42, batch 2100, loss[loss=0.1384, simple_loss=0.2261, pruned_loss=0.02538, over 7293.00 frames. ], tot_loss[loss=0.1284, simple_loss=0.2096, pruned_loss=0.02364, over 1442170.56 frames. ], batch size: 68, lr: 3.92e-03, grad_scale: 8.0 +2023-03-21 10:44:38,095 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 10:44:40,684 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 10:44:51,735 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8575, 4.3729, 4.2029, 4.7932, 4.6005, 4.7309, 4.2191, 4.3288], + device='cuda:0'), covar=tensor([0.0910, 0.2655, 0.2472, 0.1040, 0.0986, 0.1101, 0.0909, 0.1202], + device='cuda:0'), in_proj_covar=tensor([0.0270, 0.0401, 0.0306, 0.0321, 0.0238, 0.0379, 0.0234, 0.0284], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 10:44:54,186 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=117925.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:44:54,559 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.96 vs. limit=5.0 +2023-03-21 10:44:59,105 INFO [train.py:901] (0/2) Epoch 42, batch 2150, loss[loss=0.1171, simple_loss=0.2025, pruned_loss=0.01583, over 7308.00 frames. ], tot_loss[loss=0.1282, simple_loss=0.2096, pruned_loss=0.02345, over 1442058.74 frames. ], batch size: 80, lr: 3.92e-03, grad_scale: 8.0 +2023-03-21 10:45:10,031 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.110e+02 1.728e+02 2.032e+02 2.462e+02 4.551e+02, threshold=4.063e+02, percent-clipped=1.0 +2023-03-21 10:45:25,245 INFO [train.py:901] (0/2) Epoch 42, batch 2200, loss[loss=0.1486, simple_loss=0.2412, pruned_loss=0.02804, over 6734.00 frames. ], tot_loss[loss=0.1283, simple_loss=0.2097, pruned_loss=0.02343, over 1442040.99 frames. ], batch size: 107, lr: 3.91e-03, grad_scale: 8.0 +2023-03-21 10:45:25,260 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 10:45:25,351 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=117985.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:45:50,271 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=118033.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:45:51,239 INFO [train.py:901] (0/2) Epoch 42, batch 2250, loss[loss=0.1324, simple_loss=0.2091, pruned_loss=0.02781, over 7264.00 frames. ], tot_loss[loss=0.1288, simple_loss=0.2104, pruned_loss=0.02361, over 1445432.32 frames. ], batch size: 52, lr: 3.91e-03, grad_scale: 8.0 +2023-03-21 10:45:53,864 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=118040.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:45:54,310 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.3090, 4.7361, 4.6523, 5.2518, 5.0612, 5.1922, 4.6983, 4.8123], + device='cuda:0'), covar=tensor([0.0887, 0.2695, 0.2442, 0.1199, 0.1012, 0.1242, 0.0942, 0.1221], + device='cuda:0'), in_proj_covar=tensor([0.0268, 0.0396, 0.0301, 0.0318, 0.0235, 0.0374, 0.0232, 0.0280], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 10:45:59,331 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 10:45:59,841 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 10:46:01,749 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.280e+02 1.865e+02 2.098e+02 2.420e+02 4.681e+02, threshold=4.196e+02, percent-clipped=2.0 +2023-03-21 10:46:13,097 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 10:46:17,175 INFO [train.py:901] (0/2) Epoch 42, batch 2300, loss[loss=0.14, simple_loss=0.2183, pruned_loss=0.03085, over 7208.00 frames. ], tot_loss[loss=0.1284, simple_loss=0.2096, pruned_loss=0.02353, over 1447052.26 frames. ], batch size: 50, lr: 3.91e-03, grad_scale: 8.0 +2023-03-21 10:46:17,995 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.81 vs. limit=2.0 +2023-03-21 10:46:25,278 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=118101.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:46:27,318 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5139, 3.2035, 3.3156, 3.4226, 3.0622, 2.9347, 3.7128, 2.3906], + device='cuda:0'), covar=tensor([0.0739, 0.0637, 0.0758, 0.0788, 0.0812, 0.1062, 0.0707, 0.2564], + device='cuda:0'), in_proj_covar=tensor([0.0329, 0.0337, 0.0268, 0.0352, 0.0281, 0.0286, 0.0344, 0.0242], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:46:29,295 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8783, 2.4692, 3.0593, 2.7750, 3.0760, 2.8909, 2.6103, 3.0433], + device='cuda:0'), covar=tensor([0.1290, 0.0775, 0.0848, 0.1278, 0.0637, 0.0813, 0.1705, 0.1213], + device='cuda:0'), in_proj_covar=tensor([0.0052, 0.0070, 0.0053, 0.0052, 0.0052, 0.0051, 0.0070, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 10:46:30,334 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6865, 2.5738, 2.4526, 3.7144, 1.9138, 3.7326, 1.6447, 3.3427], + device='cuda:0'), covar=tensor([0.0235, 0.1547, 0.2248, 0.0281, 0.4781, 0.0315, 0.1411, 0.0542], + device='cuda:0'), in_proj_covar=tensor([0.0205, 0.0245, 0.0255, 0.0210, 0.0248, 0.0217, 0.0221, 0.0227], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:46:42,792 INFO [train.py:901] (0/2) Epoch 42, batch 2350, loss[loss=0.135, simple_loss=0.2167, pruned_loss=0.02664, over 7344.00 frames. ], tot_loss[loss=0.1284, simple_loss=0.2096, pruned_loss=0.02356, over 1445528.65 frames. ], batch size: 54, lr: 3.91e-03, grad_scale: 8.0 +2023-03-21 10:46:53,312 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.246e+02 1.681e+02 1.961e+02 2.398e+02 5.180e+02, threshold=3.922e+02, percent-clipped=2.0 +2023-03-21 10:46:55,675 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-03-21 10:46:59,500 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 10:47:00,089 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=118168.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:47:02,624 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8267, 1.7117, 2.0509, 2.1937, 2.0289, 2.0311, 1.7139, 2.1969], + device='cuda:0'), covar=tensor([0.1902, 0.3386, 0.1889, 0.1182, 0.0881, 0.3609, 0.2307, 0.2471], + device='cuda:0'), in_proj_covar=tensor([0.0079, 0.0079, 0.0072, 0.0066, 0.0066, 0.0064, 0.0105, 0.0068], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:47:06,091 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 10:47:08,472 INFO [train.py:901] (0/2) Epoch 42, batch 2400, loss[loss=0.1411, simple_loss=0.222, pruned_loss=0.03009, over 7281.00 frames. ], tot_loss[loss=0.1285, simple_loss=0.2098, pruned_loss=0.02354, over 1447194.88 frames. ], batch size: 77, lr: 3.91e-03, grad_scale: 8.0 +2023-03-21 10:47:17,342 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 10:47:20,521 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.4539, 4.9626, 4.7568, 5.3813, 5.1400, 5.3485, 4.7460, 4.9678], + device='cuda:0'), covar=tensor([0.0843, 0.2381, 0.2354, 0.1073, 0.1061, 0.1087, 0.0904, 0.1071], + device='cuda:0'), in_proj_covar=tensor([0.0267, 0.0399, 0.0301, 0.0318, 0.0235, 0.0374, 0.0232, 0.0281], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 10:47:20,931 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 10:47:31,477 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=118229.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 10:47:34,331 INFO [train.py:901] (0/2) Epoch 42, batch 2450, loss[loss=0.1396, simple_loss=0.2236, pruned_loss=0.02781, over 7284.00 frames. ], tot_loss[loss=0.1284, simple_loss=0.2095, pruned_loss=0.02365, over 1443279.88 frames. ], batch size: 66, lr: 3.91e-03, grad_scale: 8.0 +2023-03-21 10:47:45,321 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.245e+02 1.774e+02 2.090e+02 2.469e+02 5.091e+02, threshold=4.180e+02, percent-clipped=0.0 +2023-03-21 10:47:47,419 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 10:47:52,023 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=118269.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:47:59,883 INFO [train.py:901] (0/2) Epoch 42, batch 2500, loss[loss=0.1108, simple_loss=0.1886, pruned_loss=0.01655, over 7210.00 frames. ], tot_loss[loss=0.1287, simple_loss=0.2096, pruned_loss=0.02392, over 1440052.63 frames. ], batch size: 39, lr: 3.91e-03, grad_scale: 16.0 +2023-03-21 10:48:02,656 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=118289.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:48:12,615 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 10:48:22,715 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=118329.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:48:23,262 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=118330.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:48:25,609 INFO [train.py:901] (0/2) Epoch 42, batch 2550, loss[loss=0.1467, simple_loss=0.2266, pruned_loss=0.03336, over 7291.00 frames. ], tot_loss[loss=0.1293, simple_loss=0.21, pruned_loss=0.02432, over 1437319.86 frames. ], batch size: 86, lr: 3.91e-03, grad_scale: 8.0 +2023-03-21 10:48:33,835 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=118350.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:48:37,182 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.126e+02 1.748e+02 2.085e+02 2.434e+02 8.128e+02, threshold=4.170e+02, percent-clipped=2.0 +2023-03-21 10:48:51,753 INFO [train.py:901] (0/2) Epoch 42, batch 2600, loss[loss=0.1118, simple_loss=0.1987, pruned_loss=0.01247, over 7163.00 frames. ], tot_loss[loss=0.1285, simple_loss=0.2096, pruned_loss=0.02368, over 1439060.18 frames. ], batch size: 41, lr: 3.91e-03, grad_scale: 8.0 +2023-03-21 10:48:54,388 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=118390.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:48:57,335 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=118396.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:49:14,867 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3004, 3.7666, 4.3101, 4.3041, 4.2669, 4.3093, 4.3911, 4.2865], + device='cuda:0'), covar=tensor([0.0027, 0.0093, 0.0026, 0.0025, 0.0030, 0.0029, 0.0025, 0.0040], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0072, 0.0059, 0.0058, 0.0056, 0.0061, 0.0050, 0.0080], + device='cuda:0'), out_proj_covar=tensor([8.4113e-05, 1.4371e-04, 1.0416e-04, 9.8463e-05, 9.5304e-05, 1.0623e-04, + 9.3683e-05, 1.4577e-04], device='cuda:0') +2023-03-21 10:49:16,729 INFO [train.py:901] (0/2) Epoch 42, batch 2650, loss[loss=0.1402, simple_loss=0.2296, pruned_loss=0.02542, over 7126.00 frames. ], tot_loss[loss=0.1284, simple_loss=0.2097, pruned_loss=0.02352, over 1440402.02 frames. ], batch size: 98, lr: 3.91e-03, grad_scale: 8.0 +2023-03-21 10:49:27,697 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.174e+02 1.667e+02 1.916e+02 2.431e+02 3.839e+02, threshold=3.833e+02, percent-clipped=0.0 +2023-03-21 10:49:41,251 INFO [train.py:901] (0/2) Epoch 42, batch 2700, loss[loss=0.1262, simple_loss=0.2109, pruned_loss=0.02074, over 7326.00 frames. ], tot_loss[loss=0.1288, simple_loss=0.21, pruned_loss=0.02383, over 1439453.90 frames. ], batch size: 83, lr: 3.91e-03, grad_scale: 8.0 +2023-03-21 10:50:00,162 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=118524.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 10:50:05,546 INFO [train.py:901] (0/2) Epoch 42, batch 2750, loss[loss=0.1598, simple_loss=0.2347, pruned_loss=0.0425, over 6588.00 frames. ], tot_loss[loss=0.129, simple_loss=0.2099, pruned_loss=0.02403, over 1439309.94 frames. ], batch size: 106, lr: 3.91e-03, grad_scale: 8.0 +2023-03-21 10:50:14,714 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4869, 2.4501, 2.4644, 2.2478, 2.3447, 2.2931, 2.2100, 1.9041], + device='cuda:0'), covar=tensor([0.0344, 0.0477, 0.0392, 0.0270, 0.0761, 0.0648, 0.0362, 0.0375], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0040, 0.0041, 0.0039, 0.0037, 0.0038, 0.0044, 0.0042], + device='cuda:0'), out_proj_covar=tensor([1.0334e-04, 1.0274e-04, 1.0229e-04, 1.0045e-04, 9.8458e-05, 9.8309e-05, + 1.0812e-04, 1.0732e-04], device='cuda:0') +2023-03-21 10:50:16,536 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.177e+02 1.730e+02 2.063e+02 2.457e+02 5.328e+02, threshold=4.125e+02, percent-clipped=3.0 +2023-03-21 10:50:30,178 INFO [train.py:901] (0/2) Epoch 42, batch 2800, loss[loss=0.1007, simple_loss=0.1657, pruned_loss=0.01782, over 5920.00 frames. ], tot_loss[loss=0.129, simple_loss=0.2097, pruned_loss=0.02414, over 1437062.94 frames. ], batch size: 25, lr: 3.91e-03, grad_scale: 8.0 +2023-03-21 10:50:32,199 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0743, 2.5272, 2.0629, 2.9885, 2.7796, 2.7774, 2.4986, 2.6537], + device='cuda:0'), covar=tensor([0.2380, 0.1223, 0.3706, 0.0757, 0.0313, 0.0384, 0.0350, 0.0412], + device='cuda:0'), in_proj_covar=tensor([0.0250, 0.0230, 0.0243, 0.0258, 0.0198, 0.0202, 0.0218, 0.0227], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:50:43,129 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-42.pt +2023-03-21 10:50:55,176 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 10:50:55,232 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 10:50:55,258 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 10:50:55,274 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 10:50:55,448 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 10:50:55,464 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 10:50:55,475 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 10:50:55,573 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 10:50:55,649 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 10:50:55,665 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 10:50:55,681 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 10:50:56,208 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 10:50:56,214 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 10:50:56,221 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 10:50:56,361 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 10:50:56,364 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 10:50:56,423 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 10:50:58,608 INFO [train.py:901] (0/2) Epoch 43, batch 0, loss[loss=0.1223, simple_loss=0.2054, pruned_loss=0.01962, over 7267.00 frames. ], tot_loss[loss=0.1223, simple_loss=0.2054, pruned_loss=0.01962, over 7267.00 frames. ], batch size: 52, lr: 3.86e-03, grad_scale: 8.0 +2023-03-21 10:50:58,610 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 10:51:04,079 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0108, 2.5555, 3.3868, 3.0177, 3.3275, 3.1144, 2.8360, 3.2013], + device='cuda:0'), covar=tensor([0.1244, 0.0655, 0.0617, 0.1159, 0.0483, 0.0476, 0.1200, 0.0962], + device='cuda:0'), in_proj_covar=tensor([0.0052, 0.0070, 0.0053, 0.0052, 0.0052, 0.0050, 0.0070, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 10:51:24,840 INFO [train.py:935] (0/2) Epoch 43, validation: loss=0.1646, simple_loss=0.2571, pruned_loss=0.03606, over 1622729.00 frames. +2023-03-21 10:51:24,840 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 10:51:25,997 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9608, 2.4441, 1.9385, 2.7371, 2.6401, 2.5916, 2.2972, 2.4367], + device='cuda:0'), covar=tensor([0.2406, 0.1232, 0.3951, 0.0719, 0.0300, 0.0373, 0.0385, 0.0434], + device='cuda:0'), in_proj_covar=tensor([0.0250, 0.0230, 0.0244, 0.0258, 0.0198, 0.0202, 0.0218, 0.0227], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:51:30,891 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 10:51:33,550 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=118625.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:51:36,764 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=118631.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:51:42,067 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 10:51:43,653 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=118645.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:51:48,661 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 10:51:48,746 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2181, 4.3802, 4.2044, 4.3877, 4.0609, 4.3300, 4.6481, 4.6880], + device='cuda:0'), covar=tensor([0.0195, 0.0135, 0.0202, 0.0144, 0.0379, 0.0235, 0.0233, 0.0166], + device='cuda:0'), in_proj_covar=tensor([0.0124, 0.0128, 0.0123, 0.0127, 0.0114, 0.0103, 0.0100, 0.0103], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 10:51:49,631 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.053e+02 1.693e+02 1.954e+02 2.272e+02 3.982e+02, threshold=3.907e+02, percent-clipped=0.0 +2023-03-21 10:51:50,641 INFO [train.py:901] (0/2) Epoch 43, batch 50, loss[loss=0.1205, simple_loss=0.2045, pruned_loss=0.01823, over 7217.00 frames. ], tot_loss[loss=0.129, simple_loss=0.2118, pruned_loss=0.02315, over 326045.38 frames. ], batch size: 50, lr: 3.86e-03, grad_scale: 8.0 +2023-03-21 10:51:51,150 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 10:51:54,222 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 10:52:04,443 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=118685.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:52:07,990 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=118692.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:52:09,890 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=118696.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:52:11,733 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 10:52:12,215 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 10:52:13,864 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4716, 1.9118, 1.6375, 1.8310, 1.9641, 1.6984, 1.7991, 1.4814], + device='cuda:0'), covar=tensor([0.0319, 0.0248, 0.0301, 0.0286, 0.0203, 0.0197, 0.0194, 0.0303], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0040, 0.0037, 0.0038, 0.0037, 0.0036, 0.0039, 0.0048], + device='cuda:0'), out_proj_covar=tensor([4.4665e-05, 4.4270e-05, 4.2262e-05, 4.2213e-05, 4.1294e-05, 3.9853e-05, + 4.3942e-05, 5.2396e-05], device='cuda:0') +2023-03-21 10:52:16,186 INFO [train.py:901] (0/2) Epoch 43, batch 100, loss[loss=0.1371, simple_loss=0.2215, pruned_loss=0.02637, over 7293.00 frames. ], tot_loss[loss=0.1304, simple_loss=0.2121, pruned_loss=0.02436, over 574428.58 frames. ], batch size: 68, lr: 3.86e-03, grad_scale: 8.0 +2023-03-21 10:52:34,364 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=118744.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:52:39,375 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6397, 5.1979, 5.2531, 5.1731, 5.0003, 4.7619, 5.2661, 5.0205], + device='cuda:0'), covar=tensor([0.0441, 0.0368, 0.0374, 0.0537, 0.0341, 0.0384, 0.0340, 0.0434], + device='cuda:0'), in_proj_covar=tensor([0.0219, 0.0264, 0.0205, 0.0206, 0.0159, 0.0233, 0.0215, 0.0148], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:52:40,765 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.242e+02 1.769e+02 1.999e+02 2.347e+02 3.939e+02, threshold=3.999e+02, percent-clipped=2.0 +2023-03-21 10:52:41,762 INFO [train.py:901] (0/2) Epoch 43, batch 150, loss[loss=0.127, simple_loss=0.2171, pruned_loss=0.01842, over 7250.00 frames. ], tot_loss[loss=0.1294, simple_loss=0.2108, pruned_loss=0.024, over 764905.12 frames. ], batch size: 89, lr: 3.86e-03, grad_scale: 8.0 +2023-03-21 10:53:08,482 INFO [train.py:901] (0/2) Epoch 43, batch 200, loss[loss=0.1159, simple_loss=0.1954, pruned_loss=0.01818, over 7339.00 frames. ], tot_loss[loss=0.1292, simple_loss=0.2103, pruned_loss=0.02405, over 914272.92 frames. ], batch size: 44, lr: 3.86e-03, grad_scale: 8.0 +2023-03-21 10:53:13,163 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 10:53:16,307 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=118824.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:53:17,078 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.99 vs. limit=5.0 +2023-03-21 10:53:18,234 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 10:53:23,755 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 10:53:32,652 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.321e+02 1.696e+02 2.045e+02 2.404e+02 3.575e+02, threshold=4.091e+02, percent-clipped=0.0 +2023-03-21 10:53:33,665 INFO [train.py:901] (0/2) Epoch 43, batch 250, loss[loss=0.1228, simple_loss=0.206, pruned_loss=0.01979, over 7364.00 frames. ], tot_loss[loss=0.1294, simple_loss=0.2108, pruned_loss=0.02405, over 1032346.15 frames. ], batch size: 44, lr: 3.85e-03, grad_scale: 8.0 +2023-03-21 10:53:34,582 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.81 vs. limit=2.0 +2023-03-21 10:53:36,839 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 10:53:40,854 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=118872.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:53:55,766 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.90 vs. limit=5.0 +2023-03-21 10:53:58,072 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 10:54:00,030 INFO [train.py:901] (0/2) Epoch 43, batch 300, loss[loss=0.122, simple_loss=0.2086, pruned_loss=0.01772, over 7267.00 frames. ], tot_loss[loss=0.1288, simple_loss=0.2104, pruned_loss=0.02364, over 1122722.16 frames. ], batch size: 64, lr: 3.85e-03, grad_scale: 8.0 +2023-03-21 10:54:00,172 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7304, 3.0271, 2.6436, 2.8040, 2.9030, 2.4724, 2.8553, 2.8284], + device='cuda:0'), covar=tensor([0.0693, 0.0745, 0.1113, 0.1244, 0.1002, 0.0891, 0.0698, 0.0808], + device='cuda:0'), in_proj_covar=tensor([0.0060, 0.0060, 0.0069, 0.0061, 0.0058, 0.0064, 0.0057, 0.0056], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 10:54:00,423 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.52 vs. limit=2.0 +2023-03-21 10:54:06,594 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 10:54:08,125 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=118925.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:54:18,109 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=118945.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:54:24,620 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.093e+02 1.683e+02 1.946e+02 2.383e+02 3.462e+02, threshold=3.893e+02, percent-clipped=0.0 +2023-03-21 10:54:25,626 INFO [train.py:901] (0/2) Epoch 43, batch 350, loss[loss=0.1236, simple_loss=0.2037, pruned_loss=0.02171, over 7321.00 frames. ], tot_loss[loss=0.1288, simple_loss=0.2105, pruned_loss=0.02361, over 1196137.38 frames. ], batch size: 61, lr: 3.85e-03, grad_scale: 8.0 +2023-03-21 10:54:33,250 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=118973.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:54:39,215 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=118985.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:54:40,211 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=118987.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:54:42,161 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 10:54:43,160 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=118993.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:54:50,760 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0356, 3.9449, 3.2908, 3.5891, 3.0216, 2.2762, 1.9296, 4.0112], + device='cuda:0'), covar=tensor([0.0051, 0.0073, 0.0136, 0.0079, 0.0175, 0.0591, 0.0708, 0.0054], + device='cuda:0'), in_proj_covar=tensor([0.0118, 0.0096, 0.0115, 0.0098, 0.0135, 0.0138, 0.0131, 0.0109], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 10:54:51,690 INFO [train.py:901] (0/2) Epoch 43, batch 400, loss[loss=0.1062, simple_loss=0.1849, pruned_loss=0.01379, over 7116.00 frames. ], tot_loss[loss=0.1282, simple_loss=0.2098, pruned_loss=0.02328, over 1250691.73 frames. ], batch size: 41, lr: 3.85e-03, grad_scale: 8.0 +2023-03-21 10:55:03,869 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=119033.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:55:09,205 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.64 vs. limit=2.0 +2023-03-21 10:55:16,239 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.313e+02 1.703e+02 1.963e+02 2.470e+02 3.924e+02, threshold=3.926e+02, percent-clipped=1.0 +2023-03-21 10:55:17,238 INFO [train.py:901] (0/2) Epoch 43, batch 450, loss[loss=0.1323, simple_loss=0.2133, pruned_loss=0.02567, over 7275.00 frames. ], tot_loss[loss=0.1284, simple_loss=0.2099, pruned_loss=0.0234, over 1293600.24 frames. ], batch size: 66, lr: 3.85e-03, grad_scale: 8.0 +2023-03-21 10:55:23,770 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 10:55:24,277 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 10:55:25,849 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=119075.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:55:34,868 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=119093.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:55:42,747 INFO [train.py:901] (0/2) Epoch 43, batch 500, loss[loss=0.1382, simple_loss=0.2104, pruned_loss=0.03305, over 7306.00 frames. ], tot_loss[loss=0.1293, simple_loss=0.2105, pruned_loss=0.02406, over 1326071.86 frames. ], batch size: 49, lr: 3.85e-03, grad_scale: 8.0 +2023-03-21 10:55:45,159 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.06 vs. limit=2.0 +2023-03-21 10:55:50,031 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6158, 2.5052, 2.4680, 3.8532, 2.1078, 3.5839, 1.5629, 3.4838], + device='cuda:0'), covar=tensor([0.0202, 0.1497, 0.1855, 0.0248, 0.3713, 0.0288, 0.1228, 0.0415], + device='cuda:0'), in_proj_covar=tensor([0.0205, 0.0245, 0.0256, 0.0211, 0.0249, 0.0219, 0.0222, 0.0227], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:55:55,143 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2745, 3.0774, 2.0739, 3.7111, 3.5996, 3.7574, 3.3094, 3.2122], + device='cuda:0'), covar=tensor([0.2761, 0.0970, 0.4548, 0.0420, 0.0255, 0.0243, 0.0391, 0.0445], + device='cuda:0'), in_proj_covar=tensor([0.0253, 0.0232, 0.0245, 0.0260, 0.0200, 0.0204, 0.0218, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:55:57,076 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=119136.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:55:57,934 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 10:55:59,440 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 10:55:59,935 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 10:56:02,401 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 10:56:06,676 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=119154.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 10:56:07,550 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 10:56:08,020 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.252e+02 1.731e+02 2.047e+02 2.533e+02 3.684e+02, threshold=4.094e+02, percent-clipped=0.0 +2023-03-21 10:56:09,057 INFO [train.py:901] (0/2) Epoch 43, batch 550, loss[loss=0.1251, simple_loss=0.2118, pruned_loss=0.01917, over 7256.00 frames. ], tot_loss[loss=0.1287, simple_loss=0.2101, pruned_loss=0.0237, over 1353471.47 frames. ], batch size: 89, lr: 3.85e-03, grad_scale: 8.0 +2023-03-21 10:56:18,056 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 10:56:25,935 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 10:56:29,082 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6493, 1.9195, 1.6511, 1.8026, 1.9403, 1.8408, 1.8414, 1.5819], + device='cuda:0'), covar=tensor([0.0174, 0.0209, 0.0286, 0.0220, 0.0149, 0.0152, 0.0149, 0.0231], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0039, 0.0037, 0.0037, 0.0037, 0.0035, 0.0039, 0.0047], + device='cuda:0'), out_proj_covar=tensor([4.4220e-05, 4.3406e-05, 4.1795e-05, 4.1338e-05, 4.0744e-05, 3.9163e-05, + 4.3233e-05, 5.1686e-05], device='cuda:0') +2023-03-21 10:56:29,815 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 10:56:34,171 INFO [train.py:901] (0/2) Epoch 43, batch 600, loss[loss=0.1307, simple_loss=0.2137, pruned_loss=0.02383, over 7297.00 frames. ], tot_loss[loss=0.129, simple_loss=0.2103, pruned_loss=0.0239, over 1374615.06 frames. ], batch size: 83, lr: 3.85e-03, grad_scale: 8.0 +2023-03-21 10:56:36,823 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 10:56:53,668 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 10:56:59,206 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.206e+02 1.745e+02 2.072e+02 2.457e+02 4.750e+02, threshold=4.144e+02, percent-clipped=3.0 +2023-03-21 10:57:00,228 INFO [train.py:901] (0/2) Epoch 43, batch 650, loss[loss=0.1355, simple_loss=0.218, pruned_loss=0.02657, over 7291.00 frames. ], tot_loss[loss=0.1299, simple_loss=0.2112, pruned_loss=0.02437, over 1390285.84 frames. ], batch size: 66, lr: 3.85e-03, grad_scale: 8.0 +2023-03-21 10:57:02,287 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 10:57:14,753 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=119287.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:57:19,300 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 10:57:26,476 INFO [train.py:901] (0/2) Epoch 43, batch 700, loss[loss=0.1036, simple_loss=0.1863, pruned_loss=0.01045, over 7161.00 frames. ], tot_loss[loss=0.1295, simple_loss=0.2107, pruned_loss=0.02412, over 1399255.52 frames. ], batch size: 41, lr: 3.85e-03, grad_scale: 4.0 +2023-03-21 10:57:29,447 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 10:57:36,890 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-03-21 10:57:40,106 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=119335.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:57:51,954 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.167e+02 1.591e+02 1.956e+02 2.341e+02 4.303e+02, threshold=3.912e+02, percent-clipped=1.0 +2023-03-21 10:57:51,974 INFO [train.py:901] (0/2) Epoch 43, batch 750, loss[loss=0.1338, simple_loss=0.2125, pruned_loss=0.02756, over 7279.00 frames. ], tot_loss[loss=0.1292, simple_loss=0.2102, pruned_loss=0.02407, over 1407059.68 frames. ], batch size: 68, lr: 3.85e-03, grad_scale: 2.0 +2023-03-21 10:57:52,952 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 10:57:53,480 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 10:58:08,086 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 10:58:13,489 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 10:58:17,910 INFO [train.py:901] (0/2) Epoch 43, batch 800, loss[loss=0.1123, simple_loss=0.1955, pruned_loss=0.01457, over 7366.00 frames. ], tot_loss[loss=0.129, simple_loss=0.2099, pruned_loss=0.02406, over 1415809.06 frames. ], batch size: 73, lr: 3.85e-03, grad_scale: 4.0 +2023-03-21 10:58:19,376 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.56 vs. limit=2.0 +2023-03-21 10:58:19,992 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 10:58:21,007 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 10:58:29,557 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=119431.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:58:32,481 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 10:58:32,661 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4271, 3.5864, 2.5149, 3.9484, 3.1696, 3.3930, 1.8093, 2.7815], + device='cuda:0'), covar=tensor([0.0422, 0.0918, 0.2852, 0.0529, 0.0431, 0.0536, 0.3883, 0.1705], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0256, 0.0278, 0.0267, 0.0267, 0.0263, 0.0229, 0.0253], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 10:58:38,654 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=119449.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 10:58:43,520 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.213e+02 1.777e+02 2.026e+02 2.373e+02 4.143e+02, threshold=4.052e+02, percent-clipped=2.0 +2023-03-21 10:58:43,540 INFO [train.py:901] (0/2) Epoch 43, batch 850, loss[loss=0.1329, simple_loss=0.206, pruned_loss=0.02984, over 7260.00 frames. ], tot_loss[loss=0.1286, simple_loss=0.2096, pruned_loss=0.02375, over 1421309.59 frames. ], batch size: 47, lr: 3.85e-03, grad_scale: 4.0 +2023-03-21 10:58:51,180 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 10:58:51,190 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 10:58:56,194 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 10:59:00,233 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 10:59:09,899 INFO [train.py:901] (0/2) Epoch 43, batch 900, loss[loss=0.1376, simple_loss=0.2166, pruned_loss=0.02932, over 7331.00 frames. ], tot_loss[loss=0.1284, simple_loss=0.2094, pruned_loss=0.02373, over 1424419.12 frames. ], batch size: 61, lr: 3.84e-03, grad_scale: 4.0 +2023-03-21 10:59:12,522 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=119514.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 10:59:34,749 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-03-21 10:59:34,901 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.200e+02 1.779e+02 2.084e+02 2.521e+02 3.784e+02, threshold=4.168e+02, percent-clipped=0.0 +2023-03-21 10:59:34,921 INFO [train.py:901] (0/2) Epoch 43, batch 950, loss[loss=0.1528, simple_loss=0.2343, pruned_loss=0.03558, over 7332.00 frames. ], tot_loss[loss=0.1284, simple_loss=0.2093, pruned_loss=0.02376, over 1429573.80 frames. ], batch size: 61, lr: 3.84e-03, grad_scale: 4.0 +2023-03-21 10:59:38,010 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 10:59:43,678 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=119575.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:00:01,534 INFO [train.py:901] (0/2) Epoch 43, batch 1000, loss[loss=0.1372, simple_loss=0.2196, pruned_loss=0.02742, over 7293.00 frames. ], tot_loss[loss=0.1283, simple_loss=0.2093, pruned_loss=0.02371, over 1434815.32 frames. ], batch size: 86, lr: 3.84e-03, grad_scale: 4.0 +2023-03-21 11:00:01,549 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 11:00:20,409 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 11:00:27,111 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.326e+02 1.691e+02 2.007e+02 2.275e+02 7.458e+02, threshold=4.015e+02, percent-clipped=1.0 +2023-03-21 11:00:27,131 INFO [train.py:901] (0/2) Epoch 43, batch 1050, loss[loss=0.1263, simple_loss=0.2086, pruned_loss=0.02195, over 7330.00 frames. ], tot_loss[loss=0.1281, simple_loss=0.2092, pruned_loss=0.02347, over 1434810.38 frames. ], batch size: 75, lr: 3.84e-03, grad_scale: 4.0 +2023-03-21 11:00:30,303 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=119665.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:00:43,989 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 11:00:48,569 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 11:00:52,940 INFO [train.py:901] (0/2) Epoch 43, batch 1100, loss[loss=0.1254, simple_loss=0.1993, pruned_loss=0.02581, over 7212.00 frames. ], tot_loss[loss=0.128, simple_loss=0.2089, pruned_loss=0.02355, over 1435533.96 frames. ], batch size: 45, lr: 3.84e-03, grad_scale: 4.0 +2023-03-21 11:00:54,148 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.22 vs. limit=2.0 +2023-03-21 11:01:01,431 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=119726.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:01:03,829 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=119731.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:01:13,566 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=119749.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 11:01:18,025 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 11:01:18,487 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 11:01:18,974 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.320e+02 1.743e+02 2.054e+02 2.529e+02 4.457e+02, threshold=4.108e+02, percent-clipped=1.0 +2023-03-21 11:01:18,994 INFO [train.py:901] (0/2) Epoch 43, batch 1150, loss[loss=0.1246, simple_loss=0.2073, pruned_loss=0.02091, over 7259.00 frames. ], tot_loss[loss=0.1281, simple_loss=0.2092, pruned_loss=0.02351, over 1440046.20 frames. ], batch size: 55, lr: 3.84e-03, grad_scale: 4.0 +2023-03-21 11:01:28,901 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=119779.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:01:29,368 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 11:01:29,865 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 11:01:37,918 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=119797.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:01:44,229 INFO [train.py:901] (0/2) Epoch 43, batch 1200, loss[loss=0.131, simple_loss=0.2082, pruned_loss=0.02694, over 7345.00 frames. ], tot_loss[loss=0.1286, simple_loss=0.2098, pruned_loss=0.02365, over 1442266.63 frames. ], batch size: 54, lr: 3.84e-03, grad_scale: 8.0 +2023-03-21 11:01:56,676 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.74 vs. limit=2.0 +2023-03-21 11:02:00,111 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6874, 2.2096, 1.8204, 2.1055, 2.1380, 1.9474, 1.8759, 1.7590], + device='cuda:0'), covar=tensor([0.0156, 0.0131, 0.0225, 0.0136, 0.0156, 0.0148, 0.0332, 0.0195], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0039, 0.0038, 0.0039, 0.0037, 0.0036, 0.0039, 0.0048], + device='cuda:0'), out_proj_covar=tensor([4.4869e-05, 4.3750e-05, 4.2475e-05, 4.2677e-05, 4.0988e-05, 3.9922e-05, + 4.3666e-05, 5.3001e-05], device='cuda:0') +2023-03-21 11:02:03,022 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 11:02:10,429 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.203e+02 1.784e+02 2.054e+02 2.428e+02 4.631e+02, threshold=4.108e+02, percent-clipped=3.0 +2023-03-21 11:02:10,449 INFO [train.py:901] (0/2) Epoch 43, batch 1250, loss[loss=0.1301, simple_loss=0.2166, pruned_loss=0.02183, over 7312.00 frames. ], tot_loss[loss=0.1284, simple_loss=0.2098, pruned_loss=0.02351, over 1441572.15 frames. ], batch size: 49, lr: 3.84e-03, grad_scale: 8.0 +2023-03-21 11:02:15,984 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=119870.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:02:20,322 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.98 vs. limit=2.0 +2023-03-21 11:02:25,918 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 11:02:30,413 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 11:02:31,048 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0841, 3.8203, 3.7984, 3.7665, 3.7708, 3.6408, 3.9797, 3.6024], + device='cuda:0'), covar=tensor([0.0153, 0.0177, 0.0129, 0.0191, 0.0428, 0.0134, 0.0147, 0.0174], + device='cuda:0'), in_proj_covar=tensor([0.0098, 0.0105, 0.0106, 0.0092, 0.0183, 0.0111, 0.0109, 0.0117], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 11:02:31,435 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 11:02:35,440 INFO [train.py:901] (0/2) Epoch 43, batch 1300, loss[loss=0.1281, simple_loss=0.2106, pruned_loss=0.02283, over 7294.00 frames. ], tot_loss[loss=0.1276, simple_loss=0.2091, pruned_loss=0.02304, over 1442513.23 frames. ], batch size: 68, lr: 3.84e-03, grad_scale: 8.0 +2023-03-21 11:02:50,284 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8758, 3.3528, 2.8642, 3.1231, 3.2950, 2.8274, 3.1716, 3.1343], + device='cuda:0'), covar=tensor([0.0888, 0.0733, 0.1400, 0.1014, 0.1135, 0.0697, 0.1039, 0.0777], + device='cuda:0'), in_proj_covar=tensor([0.0061, 0.0060, 0.0070, 0.0061, 0.0058, 0.0065, 0.0058, 0.0056], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:02:51,237 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3453, 3.9878, 3.9583, 4.0483, 3.9677, 3.8394, 4.2103, 3.7334], + device='cuda:0'), covar=tensor([0.0147, 0.0188, 0.0133, 0.0166, 0.0472, 0.0142, 0.0153, 0.0196], + device='cuda:0'), in_proj_covar=tensor([0.0098, 0.0105, 0.0106, 0.0092, 0.0183, 0.0111, 0.0110, 0.0118], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 11:02:55,679 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 11:02:58,137 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 11:03:01,598 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.277e+02 1.695e+02 1.928e+02 2.177e+02 3.614e+02, threshold=3.856e+02, percent-clipped=0.0 +2023-03-21 11:03:01,618 INFO [train.py:901] (0/2) Epoch 43, batch 1350, loss[loss=0.1291, simple_loss=0.2185, pruned_loss=0.01982, over 7224.00 frames. ], tot_loss[loss=0.1277, simple_loss=0.2093, pruned_loss=0.02303, over 1444738.77 frames. ], batch size: 93, lr: 3.84e-03, grad_scale: 8.0 +2023-03-21 11:03:01,625 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 11:03:12,018 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 11:03:23,010 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-120000.pt +2023-03-21 11:03:31,239 INFO [train.py:901] (0/2) Epoch 43, batch 1400, loss[loss=0.13, simple_loss=0.2073, pruned_loss=0.02633, over 7308.00 frames. ], tot_loss[loss=0.128, simple_loss=0.2093, pruned_loss=0.02336, over 1441674.30 frames. ], batch size: 59, lr: 3.84e-03, grad_scale: 8.0 +2023-03-21 11:03:35,493 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2522, 3.0087, 3.0788, 3.2485, 2.9974, 2.8133, 3.2442, 2.3021], + device='cuda:0'), covar=tensor([0.0499, 0.0695, 0.0855, 0.0710, 0.0740, 0.1076, 0.0655, 0.2762], + device='cuda:0'), in_proj_covar=tensor([0.0328, 0.0338, 0.0271, 0.0353, 0.0283, 0.0287, 0.0344, 0.0241], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 11:03:37,935 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=120021.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:03:43,008 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3708, 2.7551, 2.2671, 3.1648, 2.7172, 2.8763, 2.5774, 2.5041], + device='cuda:0'), covar=tensor([0.2367, 0.1221, 0.3838, 0.0848, 0.0317, 0.0273, 0.0502, 0.0415], + device='cuda:0'), in_proj_covar=tensor([0.0249, 0.0230, 0.0242, 0.0258, 0.0200, 0.0204, 0.0217, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:03:49,308 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 11:03:49,745 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.51 vs. limit=2.0 +2023-03-21 11:03:56,944 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.060e+02 1.782e+02 2.120e+02 2.474e+02 4.600e+02, threshold=4.240e+02, percent-clipped=2.0 +2023-03-21 11:03:56,964 INFO [train.py:901] (0/2) Epoch 43, batch 1450, loss[loss=0.1166, simple_loss=0.1981, pruned_loss=0.01748, over 7277.00 frames. ], tot_loss[loss=0.1274, simple_loss=0.2086, pruned_loss=0.02311, over 1441141.88 frames. ], batch size: 77, lr: 3.84e-03, grad_scale: 8.0 +2023-03-21 11:04:11,175 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0304, 4.1982, 3.9603, 4.2023, 3.7092, 4.1186, 4.4580, 4.5258], + device='cuda:0'), covar=tensor([0.0228, 0.0161, 0.0258, 0.0167, 0.0436, 0.0274, 0.0233, 0.0169], + device='cuda:0'), in_proj_covar=tensor([0.0125, 0.0128, 0.0122, 0.0126, 0.0114, 0.0103, 0.0099, 0.0102], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 11:04:14,215 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 11:04:23,466 INFO [train.py:901] (0/2) Epoch 43, batch 1500, loss[loss=0.1528, simple_loss=0.2282, pruned_loss=0.03873, over 7345.00 frames. ], tot_loss[loss=0.1279, simple_loss=0.2092, pruned_loss=0.02329, over 1442252.27 frames. ], batch size: 75, lr: 3.83e-03, grad_scale: 8.0 +2023-03-21 11:04:31,003 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 11:04:34,808 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.23 vs. limit=2.0 +2023-03-21 11:04:48,522 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.152e+02 1.608e+02 1.838e+02 2.135e+02 4.788e+02, threshold=3.676e+02, percent-clipped=1.0 +2023-03-21 11:04:48,542 INFO [train.py:901] (0/2) Epoch 43, batch 1550, loss[loss=0.1344, simple_loss=0.2225, pruned_loss=0.02309, over 7276.00 frames. ], tot_loss[loss=0.1277, simple_loss=0.2091, pruned_loss=0.02314, over 1443259.79 frames. ], batch size: 70, lr: 3.83e-03, grad_scale: 8.0 +2023-03-21 11:04:54,089 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 11:04:54,173 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=120170.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:05:15,330 INFO [train.py:901] (0/2) Epoch 43, batch 1600, loss[loss=0.11, simple_loss=0.1925, pruned_loss=0.01373, over 7284.00 frames. ], tot_loss[loss=0.1275, simple_loss=0.2088, pruned_loss=0.02308, over 1441131.13 frames. ], batch size: 77, lr: 3.83e-03, grad_scale: 8.0 +2023-03-21 11:05:19,895 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=120218.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:05:26,408 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 11:05:26,894 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 11:05:30,343 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 11:05:39,536 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 11:05:40,489 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.046e+02 1.635e+02 1.960e+02 2.289e+02 3.677e+02, threshold=3.920e+02, percent-clipped=1.0 +2023-03-21 11:05:40,510 INFO [train.py:901] (0/2) Epoch 43, batch 1650, loss[loss=0.1322, simple_loss=0.2092, pruned_loss=0.0276, over 7263.00 frames. ], tot_loss[loss=0.1272, simple_loss=0.2087, pruned_loss=0.0228, over 1441883.68 frames. ], batch size: 47, lr: 3.83e-03, grad_scale: 8.0 +2023-03-21 11:05:44,626 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 11:05:53,347 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 11:06:06,869 INFO [train.py:901] (0/2) Epoch 43, batch 1700, loss[loss=0.1208, simple_loss=0.2073, pruned_loss=0.01718, over 7353.00 frames. ], tot_loss[loss=0.1267, simple_loss=0.2083, pruned_loss=0.02259, over 1443639.79 frames. ], batch size: 63, lr: 3.83e-03, grad_scale: 8.0 +2023-03-21 11:06:10,353 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 11:06:12,919 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=120321.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:06:14,403 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 11:06:16,250 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.80 vs. limit=2.0 +2023-03-21 11:06:25,076 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 11:06:26,185 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5382, 1.4890, 1.5989, 1.9001, 1.5683, 1.9129, 1.4603, 1.9243], + device='cuda:0'), covar=tensor([0.2247, 0.2671, 0.1521, 0.1602, 0.1708, 0.1277, 0.2591, 0.1459], + device='cuda:0'), in_proj_covar=tensor([0.0081, 0.0082, 0.0075, 0.0068, 0.0068, 0.0066, 0.0107, 0.0070], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 11:06:32,673 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.215e+02 1.697e+02 1.985e+02 2.246e+02 3.826e+02, threshold=3.970e+02, percent-clipped=0.0 +2023-03-21 11:06:32,693 INFO [train.py:901] (0/2) Epoch 43, batch 1750, loss[loss=0.1337, simple_loss=0.2162, pruned_loss=0.02556, over 7314.00 frames. ], tot_loss[loss=0.127, simple_loss=0.2084, pruned_loss=0.02276, over 1439815.88 frames. ], batch size: 49, lr: 3.83e-03, grad_scale: 8.0 +2023-03-21 11:06:38,440 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=120369.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:06:41,508 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2401, 3.5987, 3.1625, 3.3634, 3.4326, 2.9269, 3.4858, 2.8179], + device='cuda:0'), covar=tensor([0.0791, 0.0572, 0.0950, 0.0983, 0.1735, 0.0922, 0.0695, 0.1704], + device='cuda:0'), in_proj_covar=tensor([0.0060, 0.0060, 0.0069, 0.0061, 0.0057, 0.0064, 0.0057, 0.0055], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 11:06:49,396 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 11:06:50,398 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 11:06:58,740 INFO [train.py:901] (0/2) Epoch 43, batch 1800, loss[loss=0.1271, simple_loss=0.2159, pruned_loss=0.01916, over 7310.00 frames. ], tot_loss[loss=0.1275, simple_loss=0.2089, pruned_loss=0.02301, over 1440647.65 frames. ], batch size: 83, lr: 3.83e-03, grad_scale: 8.0 +2023-03-21 11:07:01,384 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1640, 3.7906, 3.8026, 3.7937, 3.7996, 3.6542, 4.0292, 3.6122], + device='cuda:0'), covar=tensor([0.0145, 0.0199, 0.0126, 0.0197, 0.0461, 0.0146, 0.0154, 0.0194], + device='cuda:0'), in_proj_covar=tensor([0.0099, 0.0106, 0.0107, 0.0092, 0.0184, 0.0112, 0.0110, 0.0118], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 11:07:11,671 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-03-21 11:07:11,781 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 11:07:14,054 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8274, 2.4617, 2.5982, 3.8760, 1.9845, 3.6892, 1.4731, 3.3499], + device='cuda:0'), covar=tensor([0.0182, 0.1430, 0.1771, 0.0212, 0.3635, 0.0312, 0.1284, 0.0315], + device='cuda:0'), in_proj_covar=tensor([0.0206, 0.0243, 0.0254, 0.0210, 0.0248, 0.0217, 0.0220, 0.0224], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:07:24,994 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.139e+02 1.732e+02 2.094e+02 2.439e+02 7.280e+02, threshold=4.189e+02, percent-clipped=3.0 +2023-03-21 11:07:25,015 INFO [train.py:901] (0/2) Epoch 43, batch 1850, loss[loss=0.1009, simple_loss=0.179, pruned_loss=0.01145, over 7319.00 frames. ], tot_loss[loss=0.1278, simple_loss=0.2095, pruned_loss=0.02311, over 1442274.35 frames. ], batch size: 42, lr: 3.83e-03, grad_scale: 8.0 +2023-03-21 11:07:25,115 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9593, 4.1695, 3.9685, 4.1357, 3.7667, 4.1213, 4.4035, 4.4590], + device='cuda:0'), covar=tensor([0.0226, 0.0149, 0.0238, 0.0172, 0.0370, 0.0265, 0.0231, 0.0167], + device='cuda:0'), in_proj_covar=tensor([0.0125, 0.0128, 0.0122, 0.0127, 0.0115, 0.0102, 0.0099, 0.0102], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 11:07:25,573 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 11:07:27,722 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2322, 3.3398, 2.3935, 3.6846, 2.9242, 3.1141, 1.5271, 2.5204], + device='cuda:0'), covar=tensor([0.0461, 0.0776, 0.2687, 0.0577, 0.0533, 0.0777, 0.4067, 0.1898], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0254, 0.0275, 0.0266, 0.0267, 0.0263, 0.0228, 0.0252], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:07:33,391 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.75 vs. limit=5.0 +2023-03-21 11:07:35,990 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 11:07:37,342 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.80 vs. limit=2.0 +2023-03-21 11:07:50,012 INFO [train.py:901] (0/2) Epoch 43, batch 1900, loss[loss=0.1385, simple_loss=0.2182, pruned_loss=0.02943, over 7341.00 frames. ], tot_loss[loss=0.1279, simple_loss=0.2094, pruned_loss=0.02321, over 1443251.17 frames. ], batch size: 51, lr: 3.83e-03, grad_scale: 8.0 +2023-03-21 11:07:53,037 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 11:07:57,675 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6592, 1.5287, 1.7874, 2.0449, 1.6120, 2.0270, 1.5412, 2.0686], + device='cuda:0'), covar=tensor([0.2787, 0.4237, 0.1498, 0.1600, 0.2666, 0.1373, 0.2404, 0.1924], + device='cuda:0'), in_proj_covar=tensor([0.0081, 0.0082, 0.0075, 0.0069, 0.0068, 0.0066, 0.0107, 0.0069], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 11:08:16,186 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.226e+02 1.737e+02 2.080e+02 2.428e+02 3.669e+02, threshold=4.159e+02, percent-clipped=0.0 +2023-03-21 11:08:16,206 INFO [train.py:901] (0/2) Epoch 43, batch 1950, loss[loss=0.09723, simple_loss=0.1602, pruned_loss=0.01714, over 6227.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2083, pruned_loss=0.02289, over 1443111.06 frames. ], batch size: 27, lr: 3.83e-03, grad_scale: 8.0 +2023-03-21 11:08:18,755 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 11:08:21,319 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8138, 4.0789, 3.8954, 4.0315, 3.6808, 4.1038, 4.3078, 4.3719], + device='cuda:0'), covar=tensor([0.0261, 0.0149, 0.0220, 0.0173, 0.0390, 0.0229, 0.0273, 0.0190], + device='cuda:0'), in_proj_covar=tensor([0.0124, 0.0127, 0.0121, 0.0126, 0.0113, 0.0101, 0.0098, 0.0102], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 11:08:30,339 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 11:08:35,391 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 11:08:35,897 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 11:08:41,818 INFO [train.py:901] (0/2) Epoch 43, batch 2000, loss[loss=0.1207, simple_loss=0.2024, pruned_loss=0.01949, over 7311.00 frames. ], tot_loss[loss=0.1276, simple_loss=0.2089, pruned_loss=0.02315, over 1441724.72 frames. ], batch size: 80, lr: 3.83e-03, grad_scale: 8.0 +2023-03-21 11:08:53,534 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 11:08:56,357 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0094, 2.8404, 2.9742, 3.0747, 2.7068, 2.6220, 3.0120, 2.1998], + device='cuda:0'), covar=tensor([0.0805, 0.0773, 0.0808, 0.0959, 0.0816, 0.1181, 0.0848, 0.2644], + device='cuda:0'), in_proj_covar=tensor([0.0330, 0.0337, 0.0272, 0.0354, 0.0282, 0.0286, 0.0347, 0.0241], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 11:09:05,241 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 11:09:08,199 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.300e+02 1.682e+02 2.036e+02 2.284e+02 3.564e+02, threshold=4.072e+02, percent-clipped=0.0 +2023-03-21 11:09:08,219 INFO [train.py:901] (0/2) Epoch 43, batch 2050, loss[loss=0.1206, simple_loss=0.209, pruned_loss=0.0161, over 7260.00 frames. ], tot_loss[loss=0.1277, simple_loss=0.2088, pruned_loss=0.02332, over 1442755.82 frames. ], batch size: 64, lr: 3.83e-03, grad_scale: 8.0 +2023-03-21 11:09:13,291 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 11:09:16,986 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=120676.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:09:32,824 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-03-21 11:09:33,481 INFO [train.py:901] (0/2) Epoch 43, batch 2100, loss[loss=0.1398, simple_loss=0.23, pruned_loss=0.02478, over 7231.00 frames. ], tot_loss[loss=0.1281, simple_loss=0.2093, pruned_loss=0.02348, over 1441288.80 frames. ], batch size: 93, lr: 3.83e-03, grad_scale: 8.0 +2023-03-21 11:09:46,695 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 11:09:48,887 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=120737.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:09:49,121 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-03-21 11:09:49,732 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 11:09:59,680 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.134e+02 1.702e+02 2.039e+02 2.311e+02 3.406e+02, threshold=4.078e+02, percent-clipped=0.0 +2023-03-21 11:09:59,700 INFO [train.py:901] (0/2) Epoch 43, batch 2150, loss[loss=0.1351, simple_loss=0.2235, pruned_loss=0.02336, over 7148.00 frames. ], tot_loss[loss=0.1279, simple_loss=0.2093, pruned_loss=0.02329, over 1441448.48 frames. ], batch size: 98, lr: 3.82e-03, grad_scale: 8.0 +2023-03-21 11:10:26,230 INFO [train.py:901] (0/2) Epoch 43, batch 2200, loss[loss=0.1274, simple_loss=0.2126, pruned_loss=0.02114, over 7280.00 frames. ], tot_loss[loss=0.1277, simple_loss=0.2092, pruned_loss=0.02309, over 1442414.05 frames. ], batch size: 70, lr: 3.82e-03, grad_scale: 8.0 +2023-03-21 11:10:34,302 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 11:10:51,387 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.129e+02 1.844e+02 2.112e+02 2.583e+02 4.485e+02, threshold=4.224e+02, percent-clipped=2.0 +2023-03-21 11:10:51,407 INFO [train.py:901] (0/2) Epoch 43, batch 2250, loss[loss=0.1466, simple_loss=0.2298, pruned_loss=0.03168, over 7248.00 frames. ], tot_loss[loss=0.1283, simple_loss=0.2099, pruned_loss=0.02341, over 1439687.18 frames. ], batch size: 93, lr: 3.82e-03, grad_scale: 8.0 +2023-03-21 11:10:55,775 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.55 vs. limit=2.0 +2023-03-21 11:11:07,651 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 11:11:08,692 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 11:11:17,805 INFO [train.py:901] (0/2) Epoch 43, batch 2300, loss[loss=0.1329, simple_loss=0.2185, pruned_loss=0.02363, over 7242.00 frames. ], tot_loss[loss=0.1276, simple_loss=0.2093, pruned_loss=0.023, over 1440169.39 frames. ], batch size: 93, lr: 3.82e-03, grad_scale: 8.0 +2023-03-21 11:11:20,301 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 11:11:30,820 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5060, 4.0485, 4.1089, 4.1162, 4.1528, 3.9132, 4.3297, 3.8863], + device='cuda:0'), covar=tensor([0.0151, 0.0170, 0.0121, 0.0185, 0.0453, 0.0157, 0.0142, 0.0204], + device='cuda:0'), in_proj_covar=tensor([0.0100, 0.0106, 0.0107, 0.0092, 0.0184, 0.0112, 0.0110, 0.0118], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 11:11:40,423 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.36 vs. limit=5.0 +2023-03-21 11:11:42,600 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.126e+02 1.667e+02 2.000e+02 2.396e+02 4.308e+02, threshold=4.000e+02, percent-clipped=1.0 +2023-03-21 11:11:42,620 INFO [train.py:901] (0/2) Epoch 43, batch 2350, loss[loss=0.1357, simple_loss=0.2219, pruned_loss=0.02472, over 7281.00 frames. ], tot_loss[loss=0.1281, simple_loss=0.2097, pruned_loss=0.02329, over 1441079.51 frames. ], batch size: 52, lr: 3.82e-03, grad_scale: 8.0 +2023-03-21 11:11:50,562 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.31 vs. limit=2.0 +2023-03-21 11:12:06,108 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=121003.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:12:06,533 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 11:12:09,063 INFO [train.py:901] (0/2) Epoch 43, batch 2400, loss[loss=0.1412, simple_loss=0.2301, pruned_loss=0.02615, over 7318.00 frames. ], tot_loss[loss=0.1278, simple_loss=0.2094, pruned_loss=0.02311, over 1442345.64 frames. ], batch size: 59, lr: 3.82e-03, grad_scale: 8.0 +2023-03-21 11:12:13,072 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 11:12:20,524 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=121032.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:12:23,024 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 11:12:25,041 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 11:12:35,174 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.205e+02 1.665e+02 2.063e+02 2.485e+02 4.404e+02, threshold=4.127e+02, percent-clipped=1.0 +2023-03-21 11:12:35,194 INFO [train.py:901] (0/2) Epoch 43, batch 2450, loss[loss=0.1128, simple_loss=0.1953, pruned_loss=0.01511, over 7342.00 frames. ], tot_loss[loss=0.1278, simple_loss=0.2095, pruned_loss=0.02306, over 1442161.27 frames. ], batch size: 44, lr: 3.82e-03, grad_scale: 8.0 +2023-03-21 11:12:37,854 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=121064.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:12:51,726 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 11:13:00,333 INFO [train.py:901] (0/2) Epoch 43, batch 2500, loss[loss=0.1215, simple_loss=0.2072, pruned_loss=0.01794, over 7270.00 frames. ], tot_loss[loss=0.1278, simple_loss=0.2095, pruned_loss=0.02308, over 1443098.56 frames. ], batch size: 77, lr: 3.82e-03, grad_scale: 4.0 +2023-03-21 11:13:18,050 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 11:13:26,644 INFO [train.py:901] (0/2) Epoch 43, batch 2550, loss[loss=0.1339, simple_loss=0.2062, pruned_loss=0.03078, over 7277.00 frames. ], tot_loss[loss=0.1278, simple_loss=0.2093, pruned_loss=0.02319, over 1444170.81 frames. ], batch size: 47, lr: 3.82e-03, grad_scale: 4.0 +2023-03-21 11:13:27,088 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.148e+02 1.681e+02 1.922e+02 2.296e+02 3.728e+02, threshold=3.843e+02, percent-clipped=0.0 +2023-03-21 11:13:29,367 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=121164.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:13:52,441 INFO [train.py:901] (0/2) Epoch 43, batch 2600, loss[loss=0.1281, simple_loss=0.1979, pruned_loss=0.02917, over 7227.00 frames. ], tot_loss[loss=0.1285, simple_loss=0.2101, pruned_loss=0.02352, over 1444927.27 frames. ], batch size: 45, lr: 3.82e-03, grad_scale: 4.0 +2023-03-21 11:14:00,572 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=121225.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:14:17,785 INFO [train.py:901] (0/2) Epoch 43, batch 2650, loss[loss=0.127, simple_loss=0.2074, pruned_loss=0.0233, over 7300.00 frames. ], tot_loss[loss=0.1284, simple_loss=0.2098, pruned_loss=0.02347, over 1446356.25 frames. ], batch size: 49, lr: 3.82e-03, grad_scale: 4.0 +2023-03-21 11:14:18,235 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.000e+02 1.753e+02 2.054e+02 2.477e+02 3.622e+02, threshold=4.107e+02, percent-clipped=0.0 +2023-03-21 11:14:42,690 INFO [train.py:901] (0/2) Epoch 43, batch 2700, loss[loss=0.1421, simple_loss=0.2246, pruned_loss=0.02982, over 7297.00 frames. ], tot_loss[loss=0.1288, simple_loss=0.2102, pruned_loss=0.02367, over 1445815.42 frames. ], batch size: 49, lr: 3.82e-03, grad_scale: 4.0 +2023-03-21 11:14:43,792 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=121311.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 11:14:45,906 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-03-21 11:14:54,032 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=121332.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:15:06,970 INFO [train.py:901] (0/2) Epoch 43, batch 2750, loss[loss=0.1312, simple_loss=0.2135, pruned_loss=0.02449, over 7279.00 frames. ], tot_loss[loss=0.1295, simple_loss=0.2108, pruned_loss=0.02414, over 1443518.92 frames. ], batch size: 52, lr: 3.81e-03, grad_scale: 4.0 +2023-03-21 11:15:07,047 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=121359.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:15:07,425 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.235e+02 1.784e+02 2.196e+02 2.525e+02 4.898e+02, threshold=4.392e+02, percent-clipped=1.0 +2023-03-21 11:15:13,540 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=121372.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 11:15:17,300 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=121380.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:15:29,860 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3959, 2.8776, 2.0779, 3.3065, 3.2041, 3.2574, 2.7464, 2.7445], + device='cuda:0'), covar=tensor([0.2278, 0.1115, 0.4261, 0.0580, 0.0330, 0.0292, 0.0332, 0.0446], + device='cuda:0'), in_proj_covar=tensor([0.0250, 0.0231, 0.0243, 0.0259, 0.0202, 0.0205, 0.0218, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:15:30,812 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=121407.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:15:31,681 INFO [train.py:901] (0/2) Epoch 43, batch 2800, loss[loss=0.1289, simple_loss=0.2109, pruned_loss=0.02347, over 7285.00 frames. ], tot_loss[loss=0.1289, simple_loss=0.2103, pruned_loss=0.02375, over 1441775.38 frames. ], batch size: 57, lr: 3.81e-03, grad_scale: 8.0 +2023-03-21 11:15:41,157 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-03-21 11:15:44,060 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-43.pt +2023-03-21 11:15:56,030 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 11:15:56,086 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 11:15:56,112 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 11:15:56,129 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 11:15:56,633 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 11:15:56,649 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 11:15:56,660 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 11:15:56,769 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 11:15:56,860 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 11:15:56,875 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 11:15:56,892 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 11:15:57,054 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 11:15:57,060 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 11:15:57,067 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 11:15:57,207 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 11:15:57,210 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 11:15:57,269 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 11:15:59,589 INFO [train.py:901] (0/2) Epoch 44, batch 0, loss[loss=0.1311, simple_loss=0.2184, pruned_loss=0.02189, over 7136.00 frames. ], tot_loss[loss=0.1311, simple_loss=0.2184, pruned_loss=0.02189, over 7136.00 frames. ], batch size: 98, lr: 3.77e-03, grad_scale: 8.0 +2023-03-21 11:15:59,591 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 11:16:06,978 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2296, 2.3568, 2.6139, 2.3816, 2.5028, 2.4110, 2.1684, 1.7804], + device='cuda:0'), covar=tensor([0.0317, 0.0383, 0.0304, 0.0181, 0.0248, 0.0375, 0.0394, 0.0387], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0040, 0.0041, 0.0040, 0.0037, 0.0037, 0.0044, 0.0043], + device='cuda:0'), out_proj_covar=tensor([1.0406e-04, 1.0404e-04, 1.0254e-04, 1.0218e-04, 9.8559e-05, 9.8310e-05, + 1.0859e-04, 1.0895e-04], device='cuda:0') +2023-03-21 11:16:17,526 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6337, 4.6300, 4.3773, 4.4459, 4.2762, 3.4902, 3.1437, 4.7530], + device='cuda:0'), covar=tensor([0.0032, 0.0034, 0.0063, 0.0047, 0.0068, 0.0402, 0.0431, 0.0029], + device='cuda:0'), in_proj_covar=tensor([0.0118, 0.0094, 0.0115, 0.0097, 0.0133, 0.0136, 0.0129, 0.0106], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 11:16:25,574 INFO [train.py:935] (0/2) Epoch 44, validation: loss=0.1656, simple_loss=0.2575, pruned_loss=0.03679, over 1622729.00 frames. +2023-03-21 11:16:25,575 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 11:16:32,087 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 11:16:39,140 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.245e+02 1.793e+02 2.069e+02 2.336e+02 4.118e+02, threshold=4.138e+02, percent-clipped=0.0 +2023-03-21 11:16:43,276 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 11:16:43,921 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=121468.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:16:50,233 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 11:16:51,201 INFO [train.py:901] (0/2) Epoch 44, batch 50, loss[loss=0.148, simple_loss=0.231, pruned_loss=0.03249, over 7335.00 frames. ], tot_loss[loss=0.129, simple_loss=0.211, pruned_loss=0.02352, over 326531.97 frames. ], batch size: 54, lr: 3.77e-03, grad_scale: 8.0 +2023-03-21 11:16:52,209 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 11:16:55,174 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 11:17:10,451 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=121520.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:17:12,880 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 11:17:13,369 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 11:17:16,522 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5833, 4.4046, 3.8481, 4.0673, 3.8127, 2.5484, 2.2235, 4.5973], + device='cuda:0'), covar=tensor([0.0042, 0.0064, 0.0105, 0.0070, 0.0123, 0.0569, 0.0591, 0.0044], + device='cuda:0'), in_proj_covar=tensor([0.0118, 0.0094, 0.0115, 0.0097, 0.0133, 0.0137, 0.0129, 0.0107], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 11:17:16,914 INFO [train.py:901] (0/2) Epoch 44, batch 100, loss[loss=0.1094, simple_loss=0.1881, pruned_loss=0.01535, over 7191.00 frames. ], tot_loss[loss=0.1284, simple_loss=0.21, pruned_loss=0.02335, over 573503.62 frames. ], batch size: 39, lr: 3.77e-03, grad_scale: 8.0 +2023-03-21 11:17:18,177 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=5.05 vs. limit=5.0 +2023-03-21 11:17:30,940 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.151e+02 1.757e+02 2.045e+02 2.523e+02 5.240e+02, threshold=4.089e+02, percent-clipped=2.0 +2023-03-21 11:17:38,459 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1992, 4.0576, 3.2658, 3.7679, 3.1040, 2.2066, 1.8869, 4.2427], + device='cuda:0'), covar=tensor([0.0045, 0.0070, 0.0155, 0.0069, 0.0186, 0.0627, 0.0661, 0.0042], + device='cuda:0'), in_proj_covar=tensor([0.0117, 0.0094, 0.0115, 0.0097, 0.0133, 0.0136, 0.0128, 0.0106], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 11:17:42,369 INFO [train.py:901] (0/2) Epoch 44, batch 150, loss[loss=0.109, simple_loss=0.1886, pruned_loss=0.0147, over 7200.00 frames. ], tot_loss[loss=0.1291, simple_loss=0.2109, pruned_loss=0.02365, over 767356.16 frames. ], batch size: 39, lr: 3.77e-03, grad_scale: 8.0 +2023-03-21 11:17:56,388 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.15 vs. limit=5.0 +2023-03-21 11:18:08,081 INFO [train.py:901] (0/2) Epoch 44, batch 200, loss[loss=0.1312, simple_loss=0.2136, pruned_loss=0.02438, over 7285.00 frames. ], tot_loss[loss=0.1282, simple_loss=0.2099, pruned_loss=0.02325, over 918511.32 frames. ], batch size: 57, lr: 3.77e-03, grad_scale: 8.0 +2023-03-21 11:18:13,176 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 11:18:18,158 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 11:18:18,284 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.3686, 1.6640, 1.3621, 1.6380, 1.7089, 1.6387, 1.4937, 1.4029], + device='cuda:0'), covar=tensor([0.0198, 0.0178, 0.0244, 0.0164, 0.0127, 0.0153, 0.0174, 0.0184], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0038, 0.0036, 0.0038, 0.0036, 0.0034, 0.0037, 0.0046], + device='cuda:0'), out_proj_covar=tensor([4.3497e-05, 4.2087e-05, 4.1049e-05, 4.1661e-05, 3.9710e-05, 3.8257e-05, + 4.1680e-05, 5.0860e-05], device='cuda:0') +2023-03-21 11:18:21,728 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=121659.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:18:22,110 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.191e+02 1.731e+02 1.966e+02 2.268e+02 4.894e+02, threshold=3.933e+02, percent-clipped=1.0 +2023-03-21 11:18:23,126 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 11:18:25,634 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=121667.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 11:18:34,222 INFO [train.py:901] (0/2) Epoch 44, batch 250, loss[loss=0.1401, simple_loss=0.2209, pruned_loss=0.02966, over 7364.00 frames. ], tot_loss[loss=0.128, simple_loss=0.2096, pruned_loss=0.0232, over 1033804.73 frames. ], batch size: 73, lr: 3.77e-03, grad_scale: 8.0 +2023-03-21 11:18:36,722 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 11:18:46,446 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=121707.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:18:52,646 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3521, 3.4424, 2.5999, 3.8203, 2.9264, 3.2671, 1.6531, 2.7070], + device='cuda:0'), covar=tensor([0.0514, 0.0790, 0.2563, 0.0699, 0.0498, 0.0633, 0.3759, 0.1683], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0255, 0.0275, 0.0266, 0.0265, 0.0262, 0.0227, 0.0253], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:18:57,521 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 11:18:59,995 INFO [train.py:901] (0/2) Epoch 44, batch 300, loss[loss=0.1304, simple_loss=0.2097, pruned_loss=0.02562, over 7338.00 frames. ], tot_loss[loss=0.1279, simple_loss=0.2097, pruned_loss=0.02309, over 1124822.58 frames. ], batch size: 61, lr: 3.77e-03, grad_scale: 8.0 +2023-03-21 11:19:04,953 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 11:19:11,702 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.62 vs. limit=5.0 +2023-03-21 11:19:13,966 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.123e+02 1.695e+02 1.999e+02 2.346e+02 5.920e+02, threshold=3.997e+02, percent-clipped=4.0 +2023-03-21 11:19:15,570 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=121763.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:19:18,143 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8593, 3.1556, 2.7307, 2.9841, 3.0492, 2.5424, 2.9664, 2.9938], + device='cuda:0'), covar=tensor([0.0548, 0.0427, 0.0993, 0.1270, 0.0826, 0.0930, 0.1113, 0.0768], + device='cuda:0'), in_proj_covar=tensor([0.0061, 0.0060, 0.0069, 0.0061, 0.0058, 0.0065, 0.0058, 0.0056], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:19:25,449 INFO [train.py:901] (0/2) Epoch 44, batch 350, loss[loss=0.1302, simple_loss=0.2097, pruned_loss=0.02532, over 7305.00 frames. ], tot_loss[loss=0.1279, simple_loss=0.2093, pruned_loss=0.02323, over 1193876.78 frames. ], batch size: 86, lr: 3.76e-03, grad_scale: 8.0 +2023-03-21 11:19:36,986 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7694, 4.6208, 4.4729, 4.2456, 3.9817, 3.1429, 2.8823, 4.7983], + device='cuda:0'), covar=tensor([0.0041, 0.0055, 0.0062, 0.0061, 0.0120, 0.0405, 0.0426, 0.0038], + device='cuda:0'), in_proj_covar=tensor([0.0118, 0.0094, 0.0115, 0.0098, 0.0134, 0.0136, 0.0129, 0.0107], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 11:19:39,771 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.50 vs. limit=5.0 +2023-03-21 11:19:40,904 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 11:19:44,962 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=121820.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:19:51,311 INFO [train.py:901] (0/2) Epoch 44, batch 400, loss[loss=0.1077, simple_loss=0.1775, pruned_loss=0.01899, over 5827.00 frames. ], tot_loss[loss=0.1272, simple_loss=0.2083, pruned_loss=0.02298, over 1245942.21 frames. ], batch size: 25, lr: 3.76e-03, grad_scale: 8.0 +2023-03-21 11:19:54,006 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9095, 2.7575, 2.0517, 2.9984, 2.3113, 2.5139, 1.2680, 2.1833], + device='cuda:0'), covar=tensor([0.0650, 0.1035, 0.2765, 0.0862, 0.0526, 0.0660, 0.3642, 0.1734], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0255, 0.0275, 0.0265, 0.0265, 0.0262, 0.0227, 0.0253], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:20:05,403 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.230e+02 1.734e+02 2.013e+02 2.477e+02 7.219e+02, threshold=4.025e+02, percent-clipped=2.0 +2023-03-21 11:20:09,533 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=121868.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:20:11,913 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.99 vs. limit=2.0 +2023-03-21 11:20:17,008 INFO [train.py:901] (0/2) Epoch 44, batch 450, loss[loss=0.1164, simple_loss=0.2031, pruned_loss=0.01481, over 7285.00 frames. ], tot_loss[loss=0.1266, simple_loss=0.208, pruned_loss=0.02258, over 1289535.87 frames. ], batch size: 66, lr: 3.76e-03, grad_scale: 8.0 +2023-03-21 11:20:22,001 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 11:20:22,499 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 11:20:36,441 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.79 vs. limit=2.0 +2023-03-21 11:20:37,773 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5542, 3.6674, 2.4971, 3.9022, 3.1587, 3.5419, 1.6976, 2.7394], + device='cuda:0'), covar=tensor([0.0578, 0.0947, 0.3033, 0.0585, 0.0498, 0.0616, 0.3992, 0.1740], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0255, 0.0275, 0.0266, 0.0265, 0.0262, 0.0227, 0.0252], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:20:43,172 INFO [train.py:901] (0/2) Epoch 44, batch 500, loss[loss=0.1286, simple_loss=0.2148, pruned_loss=0.02116, over 7263.00 frames. ], tot_loss[loss=0.1266, simple_loss=0.2078, pruned_loss=0.02273, over 1321931.68 frames. ], batch size: 52, lr: 3.76e-03, grad_scale: 8.0 +2023-03-21 11:20:44,481 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-03-21 11:20:48,520 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4551, 1.8310, 1.5985, 1.8477, 1.9640, 1.7637, 1.8076, 1.4196], + device='cuda:0'), covar=tensor([0.0191, 0.0249, 0.0261, 0.0155, 0.0114, 0.0162, 0.0198, 0.0252], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0038, 0.0036, 0.0038, 0.0036, 0.0035, 0.0037, 0.0046], + device='cuda:0'), out_proj_covar=tensor([4.3352e-05, 4.2205e-05, 4.1030e-05, 4.2077e-05, 3.9604e-05, 3.8427e-05, + 4.2012e-05, 5.0714e-05], device='cuda:0') +2023-03-21 11:20:51,973 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=121950.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:20:54,390 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 11:20:55,837 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 11:20:56,391 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 11:20:56,847 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.208e+02 1.686e+02 1.981e+02 2.363e+02 5.254e+02, threshold=3.961e+02, percent-clipped=1.0 +2023-03-21 11:20:58,416 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 11:21:00,566 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=121967.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 11:21:02,948 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 11:21:04,041 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4986, 1.7758, 1.5501, 1.7853, 1.8573, 1.7617, 1.6791, 1.3929], + device='cuda:0'), covar=tensor([0.0177, 0.0145, 0.0217, 0.0145, 0.0121, 0.0134, 0.0162, 0.0200], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0038, 0.0036, 0.0038, 0.0036, 0.0035, 0.0037, 0.0046], + device='cuda:0'), out_proj_covar=tensor([4.3227e-05, 4.2060e-05, 4.0865e-05, 4.2038e-05, 3.9452e-05, 3.8357e-05, + 4.1888e-05, 5.0451e-05], device='cuda:0') +2023-03-21 11:21:09,142 INFO [train.py:901] (0/2) Epoch 44, batch 550, loss[loss=0.1228, simple_loss=0.2011, pruned_loss=0.02225, over 7213.00 frames. ], tot_loss[loss=0.127, simple_loss=0.2082, pruned_loss=0.02286, over 1344338.38 frames. ], batch size: 50, lr: 3.76e-03, grad_scale: 8.0 +2023-03-21 11:21:15,209 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 11:21:23,665 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 11:21:23,789 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=122011.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:21:26,301 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=122015.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 11:21:27,745 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 11:21:34,157 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 11:21:35,136 INFO [train.py:901] (0/2) Epoch 44, batch 600, loss[loss=0.09768, simple_loss=0.1771, pruned_loss=0.009148, over 7166.00 frames. ], tot_loss[loss=0.1273, simple_loss=0.2082, pruned_loss=0.02314, over 1367167.62 frames. ], batch size: 39, lr: 3.76e-03, grad_scale: 8.0 +2023-03-21 11:21:41,868 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9958, 2.6442, 3.1112, 3.0731, 3.1387, 3.0790, 2.7857, 3.0349], + device='cuda:0'), covar=tensor([0.1381, 0.0734, 0.1034, 0.0901, 0.0800, 0.0817, 0.1472, 0.1393], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0070, 0.0053, 0.0052, 0.0052, 0.0051, 0.0070, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 11:21:47,473 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6205, 4.1331, 4.0206, 4.5588, 4.4173, 4.5081, 4.0683, 4.1171], + device='cuda:0'), covar=tensor([0.0750, 0.2412, 0.2294, 0.1130, 0.0936, 0.1154, 0.0815, 0.1224], + device='cuda:0'), in_proj_covar=tensor([0.0271, 0.0399, 0.0301, 0.0317, 0.0235, 0.0373, 0.0234, 0.0283], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 11:21:48,863 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.220e+02 1.747e+02 1.962e+02 2.510e+02 3.500e+02, threshold=3.924e+02, percent-clipped=0.0 +2023-03-21 11:21:50,426 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 11:21:50,515 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=122063.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:21:50,793 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.60 vs. limit=5.0 +2023-03-21 11:22:00,096 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 11:22:01,114 INFO [train.py:901] (0/2) Epoch 44, batch 650, loss[loss=0.1267, simple_loss=0.2073, pruned_loss=0.023, over 7319.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2081, pruned_loss=0.02303, over 1384324.70 frames. ], batch size: 80, lr: 3.76e-03, grad_scale: 8.0 +2023-03-21 11:22:15,964 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=122111.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:22:18,034 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 11:22:19,127 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=122117.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:22:26,078 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 11:22:27,037 INFO [train.py:901] (0/2) Epoch 44, batch 700, loss[loss=0.133, simple_loss=0.2179, pruned_loss=0.02406, over 7301.00 frames. ], tot_loss[loss=0.1274, simple_loss=0.2083, pruned_loss=0.02324, over 1396778.79 frames. ], batch size: 68, lr: 3.76e-03, grad_scale: 8.0 +2023-03-21 11:22:27,242 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4012, 1.5473, 1.3486, 1.5208, 1.6158, 1.5996, 1.5050, 1.4012], + device='cuda:0'), covar=tensor([0.0164, 0.0217, 0.0220, 0.0147, 0.0164, 0.0117, 0.0149, 0.0169], + device='cuda:0'), in_proj_covar=tensor([0.0038, 0.0038, 0.0036, 0.0038, 0.0035, 0.0034, 0.0037, 0.0045], + device='cuda:0'), out_proj_covar=tensor([4.2913e-05, 4.1913e-05, 4.0485e-05, 4.1833e-05, 3.9283e-05, 3.7837e-05, + 4.1653e-05, 4.9911e-05], device='cuda:0') +2023-03-21 11:22:30,233 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0532, 3.9984, 3.1532, 3.5880, 3.0439, 2.2006, 1.7646, 4.0351], + device='cuda:0'), covar=tensor([0.0051, 0.0052, 0.0152, 0.0081, 0.0169, 0.0563, 0.0659, 0.0052], + device='cuda:0'), in_proj_covar=tensor([0.0117, 0.0093, 0.0114, 0.0097, 0.0132, 0.0135, 0.0128, 0.0105], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 11:22:41,313 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.249e+02 1.853e+02 2.110e+02 2.490e+02 4.389e+02, threshold=4.220e+02, percent-clipped=2.0 +2023-03-21 11:22:50,416 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 11:22:50,554 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=122178.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 11:22:50,919 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 11:22:52,897 INFO [train.py:901] (0/2) Epoch 44, batch 750, loss[loss=0.1265, simple_loss=0.2057, pruned_loss=0.02363, over 7322.00 frames. ], tot_loss[loss=0.1275, simple_loss=0.2086, pruned_loss=0.02323, over 1407460.17 frames. ], batch size: 49, lr: 3.76e-03, grad_scale: 8.0 +2023-03-21 11:22:59,444 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=122195.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:23:05,676 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 11:23:06,417 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-03-21 11:23:10,187 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 11:23:16,202 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 11:23:17,220 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 11:23:18,710 INFO [train.py:901] (0/2) Epoch 44, batch 800, loss[loss=0.131, simple_loss=0.2138, pruned_loss=0.02404, over 7323.00 frames. ], tot_loss[loss=0.1274, simple_loss=0.2087, pruned_loss=0.02303, over 1414677.06 frames. ], batch size: 59, lr: 3.76e-03, grad_scale: 8.0 +2023-03-21 11:23:28,383 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 11:23:31,170 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=122256.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:23:33,116 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.265e+02 1.730e+02 1.943e+02 2.255e+02 3.893e+02, threshold=3.886e+02, percent-clipped=0.0 +2023-03-21 11:23:45,354 INFO [train.py:901] (0/2) Epoch 44, batch 850, loss[loss=0.1308, simple_loss=0.2099, pruned_loss=0.02589, over 7283.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2082, pruned_loss=0.02295, over 1420766.56 frames. ], batch size: 70, lr: 3.76e-03, grad_scale: 8.0 +2023-03-21 11:23:48,766 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.58 vs. limit=2.0 +2023-03-21 11:23:48,849 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 11:23:48,861 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 11:23:54,398 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 11:23:55,094 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1476, 3.3454, 2.4072, 3.6518, 2.7569, 3.1233, 1.5867, 2.4820], + device='cuda:0'), covar=tensor([0.0432, 0.0847, 0.2685, 0.0696, 0.0618, 0.0658, 0.3782, 0.1785], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0256, 0.0277, 0.0268, 0.0267, 0.0264, 0.0228, 0.0255], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:23:56,998 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=122306.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:23:57,953 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 11:24:03,577 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=122319.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:24:10,463 INFO [train.py:901] (0/2) Epoch 44, batch 900, loss[loss=0.1208, simple_loss=0.2104, pruned_loss=0.01562, over 7291.00 frames. ], tot_loss[loss=0.1267, simple_loss=0.2077, pruned_loss=0.02287, over 1425261.45 frames. ], batch size: 77, lr: 3.76e-03, grad_scale: 8.0 +2023-03-21 11:24:24,636 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.155e+02 1.690e+02 1.979e+02 2.336e+02 1.057e+03, threshold=3.958e+02, percent-clipped=2.0 +2023-03-21 11:24:35,480 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=122380.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:24:36,824 INFO [train.py:901] (0/2) Epoch 44, batch 950, loss[loss=0.109, simple_loss=0.1802, pruned_loss=0.01892, over 7201.00 frames. ], tot_loss[loss=0.1262, simple_loss=0.2075, pruned_loss=0.0225, over 1427788.25 frames. ], batch size: 39, lr: 3.76e-03, grad_scale: 8.0 +2023-03-21 11:24:37,874 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 11:24:52,789 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7334, 2.7997, 3.6632, 3.7220, 3.7148, 3.8327, 3.6831, 3.6115], + device='cuda:0'), covar=tensor([0.0039, 0.0159, 0.0040, 0.0038, 0.0040, 0.0032, 0.0055, 0.0056], + device='cuda:0'), in_proj_covar=tensor([0.0050, 0.0073, 0.0060, 0.0058, 0.0056, 0.0061, 0.0050, 0.0081], + device='cuda:0'), out_proj_covar=tensor([8.3597e-05, 1.4496e-04, 1.0622e-04, 9.8740e-05, 9.3795e-05, 1.0473e-04, + 9.2478e-05, 1.4671e-04], device='cuda:0') +2023-03-21 11:25:01,729 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 11:25:02,659 INFO [train.py:901] (0/2) Epoch 44, batch 1000, loss[loss=0.1074, simple_loss=0.189, pruned_loss=0.01293, over 7292.00 frames. ], tot_loss[loss=0.1264, simple_loss=0.2076, pruned_loss=0.02259, over 1431655.94 frames. ], batch size: 42, lr: 3.75e-03, grad_scale: 8.0 +2023-03-21 11:25:11,951 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1373, 3.4324, 2.5650, 3.7213, 2.9164, 3.1150, 1.6489, 2.7596], + device='cuda:0'), covar=tensor([0.0426, 0.0824, 0.2694, 0.0604, 0.0476, 0.0751, 0.4025, 0.1735], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0255, 0.0276, 0.0268, 0.0266, 0.0264, 0.0228, 0.0254], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:25:15,040 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7861, 3.7977, 2.8335, 3.3927, 2.5391, 2.1991, 1.7154, 3.8119], + device='cuda:0'), covar=tensor([0.0060, 0.0054, 0.0212, 0.0089, 0.0262, 0.0629, 0.0687, 0.0057], + device='cuda:0'), in_proj_covar=tensor([0.0117, 0.0093, 0.0115, 0.0097, 0.0132, 0.0135, 0.0128, 0.0105], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 11:25:16,933 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.044e+02 1.750e+02 2.006e+02 2.317e+02 3.184e+02, threshold=4.012e+02, percent-clipped=0.0 +2023-03-21 11:25:21,131 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=122468.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:25:23,519 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 11:25:23,577 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=122473.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 11:25:28,436 INFO [train.py:901] (0/2) Epoch 44, batch 1050, loss[loss=0.1285, simple_loss=0.2173, pruned_loss=0.01986, over 7342.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2085, pruned_loss=0.02288, over 1436204.50 frames. ], batch size: 63, lr: 3.75e-03, grad_scale: 8.0 +2023-03-21 11:25:28,702 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.22 vs. limit=2.0 +2023-03-21 11:25:45,278 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 11:25:50,566 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 11:25:52,767 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=122529.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:25:54,637 INFO [train.py:901] (0/2) Epoch 44, batch 1100, loss[loss=0.1172, simple_loss=0.1936, pruned_loss=0.02042, over 7139.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2083, pruned_loss=0.02289, over 1437407.96 frames. ], batch size: 41, lr: 3.75e-03, grad_scale: 8.0 +2023-03-21 11:26:04,401 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=122551.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:26:08,861 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.256e+02 1.775e+02 1.996e+02 2.337e+02 4.122e+02, threshold=3.992e+02, percent-clipped=0.0 +2023-03-21 11:26:10,969 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7119, 4.2615, 4.0940, 4.7260, 4.4887, 4.6072, 4.1664, 4.2647], + device='cuda:0'), covar=tensor([0.0873, 0.2349, 0.2188, 0.0869, 0.0914, 0.1112, 0.0839, 0.1336], + device='cuda:0'), in_proj_covar=tensor([0.0271, 0.0399, 0.0302, 0.0314, 0.0234, 0.0372, 0.0234, 0.0283], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 11:26:11,528 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3320, 3.6542, 2.5936, 3.9128, 3.0761, 3.4628, 1.7364, 2.4983], + device='cuda:0'), covar=tensor([0.0515, 0.1061, 0.2926, 0.0655, 0.0643, 0.0823, 0.4234, 0.2065], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0257, 0.0277, 0.0269, 0.0267, 0.0265, 0.0229, 0.0257], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:26:18,406 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 11:26:18,934 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 11:26:20,442 INFO [train.py:901] (0/2) Epoch 44, batch 1150, loss[loss=0.1363, simple_loss=0.2164, pruned_loss=0.02805, over 7295.00 frames. ], tot_loss[loss=0.127, simple_loss=0.2083, pruned_loss=0.02282, over 1436192.64 frames. ], batch size: 80, lr: 3.75e-03, grad_scale: 8.0 +2023-03-21 11:26:24,856 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.59 vs. limit=2.0 +2023-03-21 11:26:32,547 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 11:26:33,120 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=122606.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:26:33,530 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 11:26:47,174 INFO [train.py:901] (0/2) Epoch 44, batch 1200, loss[loss=0.1007, simple_loss=0.1782, pruned_loss=0.01162, over 7167.00 frames. ], tot_loss[loss=0.1273, simple_loss=0.2087, pruned_loss=0.02294, over 1437555.48 frames. ], batch size: 41, lr: 3.75e-03, grad_scale: 8.0 +2023-03-21 11:26:56,496 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.59 vs. limit=2.0 +2023-03-21 11:26:57,733 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=122654.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:27:00,708 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.286e+02 1.642e+02 1.822e+02 2.159e+02 3.974e+02, threshold=3.644e+02, percent-clipped=1.0 +2023-03-21 11:27:05,242 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 11:27:05,831 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9954, 3.7245, 3.6993, 3.6884, 3.4690, 3.5235, 3.9205, 3.5316], + device='cuda:0'), covar=tensor([0.0200, 0.0233, 0.0173, 0.0277, 0.0627, 0.0191, 0.0215, 0.0230], + device='cuda:0'), in_proj_covar=tensor([0.0103, 0.0110, 0.0110, 0.0095, 0.0189, 0.0115, 0.0113, 0.0122], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 11:27:08,249 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=122675.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:27:12,260 INFO [train.py:901] (0/2) Epoch 44, batch 1250, loss[loss=0.1223, simple_loss=0.2075, pruned_loss=0.01856, over 7301.00 frames. ], tot_loss[loss=0.1269, simple_loss=0.2082, pruned_loss=0.02276, over 1438034.68 frames. ], batch size: 83, lr: 3.75e-03, grad_scale: 8.0 +2023-03-21 11:27:28,451 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 11:27:33,675 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 11:27:34,715 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 11:27:37,393 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4845, 1.7430, 1.5411, 1.6619, 1.7885, 1.6595, 1.6880, 1.3043], + device='cuda:0'), covar=tensor([0.0207, 0.0200, 0.0261, 0.0165, 0.0168, 0.0160, 0.0154, 0.0217], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0039, 0.0037, 0.0039, 0.0037, 0.0035, 0.0038, 0.0047], + device='cuda:0'), out_proj_covar=tensor([4.4168e-05, 4.3496e-05, 4.1571e-05, 4.3259e-05, 4.0736e-05, 3.9356e-05, + 4.2757e-05, 5.1646e-05], device='cuda:0') +2023-03-21 11:27:38,717 INFO [train.py:901] (0/2) Epoch 44, batch 1300, loss[loss=0.1039, simple_loss=0.1842, pruned_loss=0.01176, over 7173.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2087, pruned_loss=0.02271, over 1439052.41 frames. ], batch size: 39, lr: 3.75e-03, grad_scale: 8.0 +2023-03-21 11:27:44,321 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7036, 1.4928, 1.6529, 1.9050, 1.7835, 2.0174, 1.4518, 2.0475], + device='cuda:0'), covar=tensor([0.1882, 0.4320, 0.1253, 0.1408, 0.1033, 0.1779, 0.2830, 0.1826], + device='cuda:0'), in_proj_covar=tensor([0.0082, 0.0082, 0.0075, 0.0068, 0.0067, 0.0066, 0.0107, 0.0070], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 11:27:52,318 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.097e+02 1.777e+02 1.977e+02 2.285e+02 3.721e+02, threshold=3.954e+02, percent-clipped=1.0 +2023-03-21 11:27:57,962 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 11:27:59,002 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=122773.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 11:28:00,493 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 11:28:03,469 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 11:28:04,458 INFO [train.py:901] (0/2) Epoch 44, batch 1350, loss[loss=0.1363, simple_loss=0.221, pruned_loss=0.02584, over 7110.00 frames. ], tot_loss[loss=0.1277, simple_loss=0.2093, pruned_loss=0.023, over 1439833.43 frames. ], batch size: 98, lr: 3.75e-03, grad_scale: 8.0 +2023-03-21 11:28:05,264 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.34 vs. limit=2.0 +2023-03-21 11:28:14,012 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 11:28:24,608 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=122821.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:28:26,165 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=122824.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:28:30,647 INFO [train.py:901] (0/2) Epoch 44, batch 1400, loss[loss=0.1397, simple_loss=0.2217, pruned_loss=0.02889, over 7307.00 frames. ], tot_loss[loss=0.1276, simple_loss=0.2096, pruned_loss=0.02279, over 1442526.99 frames. ], batch size: 83, lr: 3.75e-03, grad_scale: 8.0 +2023-03-21 11:28:39,879 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=122851.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:28:44,867 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.199e+02 1.709e+02 2.025e+02 2.415e+02 3.679e+02, threshold=4.051e+02, percent-clipped=0.0 +2023-03-21 11:28:46,369 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 11:28:51,040 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=122872.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:28:56,304 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.39 vs. limit=5.0 +2023-03-21 11:28:56,514 INFO [train.py:901] (0/2) Epoch 44, batch 1450, loss[loss=0.1429, simple_loss=0.2266, pruned_loss=0.02963, over 7325.00 frames. ], tot_loss[loss=0.1277, simple_loss=0.2098, pruned_loss=0.02285, over 1444821.02 frames. ], batch size: 61, lr: 3.75e-03, grad_scale: 8.0 +2023-03-21 11:29:05,325 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=122899.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:29:09,428 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7094, 2.2511, 2.9081, 2.7738, 2.8928, 2.6784, 2.3315, 2.9075], + device='cuda:0'), covar=tensor([0.1425, 0.1013, 0.0844, 0.1364, 0.0851, 0.0935, 0.2040, 0.0766], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0071, 0.0053, 0.0052, 0.0052, 0.0051, 0.0071, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 11:29:09,462 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5764, 1.7360, 1.5179, 1.6350, 1.7666, 1.6587, 1.6059, 1.3615], + device='cuda:0'), covar=tensor([0.0168, 0.0186, 0.0212, 0.0183, 0.0174, 0.0155, 0.0193, 0.0219], + device='cuda:0'), in_proj_covar=tensor([0.0039, 0.0039, 0.0036, 0.0039, 0.0036, 0.0035, 0.0038, 0.0046], + device='cuda:0'), out_proj_covar=tensor([4.3760e-05, 4.2865e-05, 4.1205e-05, 4.2913e-05, 4.0453e-05, 3.9196e-05, + 4.2299e-05, 5.1151e-05], device='cuda:0') +2023-03-21 11:29:09,873 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4746, 3.5671, 3.5143, 3.4979, 3.4956, 3.1456, 3.6166, 3.6464], + device='cuda:0'), covar=tensor([0.0355, 0.0293, 0.0368, 0.0382, 0.0417, 0.0593, 0.0485, 0.0420], + device='cuda:0'), in_proj_covar=tensor([0.0127, 0.0130, 0.0124, 0.0129, 0.0117, 0.0103, 0.0100, 0.0104], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 11:29:10,787 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 11:29:16,629 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.00 vs. limit=2.0 +2023-03-21 11:29:22,308 INFO [train.py:901] (0/2) Epoch 44, batch 1500, loss[loss=0.1385, simple_loss=0.2147, pruned_loss=0.03115, over 7279.00 frames. ], tot_loss[loss=0.1278, simple_loss=0.2095, pruned_loss=0.023, over 1443682.61 frames. ], batch size: 47, lr: 3.75e-03, grad_scale: 8.0 +2023-03-21 11:29:22,445 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=122933.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:29:26,334 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 11:29:36,565 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.123e+02 1.737e+02 1.988e+02 2.376e+02 3.622e+02, threshold=3.976e+02, percent-clipped=0.0 +2023-03-21 11:29:44,745 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=122975.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:29:48,676 INFO [train.py:901] (0/2) Epoch 44, batch 1550, loss[loss=0.1185, simple_loss=0.2019, pruned_loss=0.01753, over 7291.00 frames. ], tot_loss[loss=0.128, simple_loss=0.2095, pruned_loss=0.02327, over 1442720.08 frames. ], batch size: 77, lr: 3.75e-03, grad_scale: 8.0 +2023-03-21 11:29:50,239 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 11:29:51,554 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.22 vs. limit=2.0 +2023-03-21 11:30:09,332 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=123023.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:30:14,957 INFO [train.py:901] (0/2) Epoch 44, batch 1600, loss[loss=0.1305, simple_loss=0.2181, pruned_loss=0.02145, over 7306.00 frames. ], tot_loss[loss=0.1278, simple_loss=0.2092, pruned_loss=0.02317, over 1442758.21 frames. ], batch size: 83, lr: 3.75e-03, grad_scale: 8.0 +2023-03-21 11:30:22,699 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 11:30:23,699 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 11:30:26,746 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 11:30:29,320 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.296e+02 1.717e+02 1.945e+02 2.316e+02 3.893e+02, threshold=3.890e+02, percent-clipped=0.0 +2023-03-21 11:30:36,877 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 11:30:41,003 INFO [train.py:901] (0/2) Epoch 44, batch 1650, loss[loss=0.1274, simple_loss=0.207, pruned_loss=0.02385, over 7256.00 frames. ], tot_loss[loss=0.128, simple_loss=0.2092, pruned_loss=0.02335, over 1443891.43 frames. ], batch size: 47, lr: 3.74e-03, grad_scale: 16.0 +2023-03-21 11:30:41,516 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 11:30:46,802 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.16 vs. limit=2.0 +2023-03-21 11:30:50,088 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 11:31:02,440 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=123124.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:31:06,784 INFO [train.py:901] (0/2) Epoch 44, batch 1700, loss[loss=0.1417, simple_loss=0.2183, pruned_loss=0.03251, over 7265.00 frames. ], tot_loss[loss=0.1274, simple_loss=0.2087, pruned_loss=0.02309, over 1443130.70 frames. ], batch size: 52, lr: 3.74e-03, grad_scale: 16.0 +2023-03-21 11:31:07,771 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 11:31:12,263 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 11:31:21,063 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.227e+02 1.700e+02 1.935e+02 2.300e+02 4.091e+02, threshold=3.869e+02, percent-clipped=2.0 +2023-03-21 11:31:23,536 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 11:31:27,121 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=123172.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:31:32,599 INFO [train.py:901] (0/2) Epoch 44, batch 1750, loss[loss=0.1044, simple_loss=0.1703, pruned_loss=0.01923, over 6013.00 frames. ], tot_loss[loss=0.1279, simple_loss=0.209, pruned_loss=0.02345, over 1439231.40 frames. ], batch size: 26, lr: 3.74e-03, grad_scale: 16.0 +2023-03-21 11:31:43,321 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.0783, 2.5325, 1.9625, 2.8682, 2.8646, 3.0660, 2.6833, 2.4089], + device='cuda:0'), covar=tensor([0.2355, 0.1246, 0.4145, 0.0777, 0.0397, 0.0374, 0.0436, 0.0440], + device='cuda:0'), in_proj_covar=tensor([0.0252, 0.0234, 0.0247, 0.0261, 0.0204, 0.0208, 0.0221, 0.0232], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:31:45,762 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2288, 4.6383, 4.7156, 4.6971, 4.5718, 4.1949, 4.7446, 4.5514], + device='cuda:0'), covar=tensor([0.0447, 0.0423, 0.0373, 0.0427, 0.0369, 0.0447, 0.0323, 0.0474], + device='cuda:0'), in_proj_covar=tensor([0.0220, 0.0268, 0.0206, 0.0204, 0.0159, 0.0233, 0.0214, 0.0152], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 11:31:48,863 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 11:31:49,872 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 11:31:56,438 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=123228.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:31:58,861 INFO [train.py:901] (0/2) Epoch 44, batch 1800, loss[loss=0.1265, simple_loss=0.2117, pruned_loss=0.0206, over 7263.00 frames. ], tot_loss[loss=0.1276, simple_loss=0.2089, pruned_loss=0.02322, over 1439168.64 frames. ], batch size: 64, lr: 3.74e-03, grad_scale: 8.0 +2023-03-21 11:32:11,112 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 11:32:13,580 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.238e+02 1.759e+02 1.926e+02 2.309e+02 3.312e+02, threshold=3.851e+02, percent-clipped=0.0 +2023-03-21 11:32:24,077 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 11:32:24,494 INFO [train.py:901] (0/2) Epoch 44, batch 1850, loss[loss=0.1422, simple_loss=0.209, pruned_loss=0.03765, over 7218.00 frames. ], tot_loss[loss=0.1273, simple_loss=0.2083, pruned_loss=0.0232, over 1439992.38 frames. ], batch size: 45, lr: 3.74e-03, grad_scale: 8.0 +2023-03-21 11:32:35,347 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 11:32:41,013 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=123314.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:32:51,188 INFO [train.py:901] (0/2) Epoch 44, batch 1900, loss[loss=0.1436, simple_loss=0.2289, pruned_loss=0.02917, over 7369.00 frames. ], tot_loss[loss=0.1272, simple_loss=0.2083, pruned_loss=0.0231, over 1440820.84 frames. ], batch size: 65, lr: 3.74e-03, grad_scale: 8.0 +2023-03-21 11:32:52,740 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 11:33:05,243 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.291e+02 1.842e+02 2.137e+02 2.447e+02 5.205e+02, threshold=4.273e+02, percent-clipped=1.0 +2023-03-21 11:33:12,489 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=123375.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:33:15,708 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 11:33:16,215 INFO [train.py:901] (0/2) Epoch 44, batch 1950, loss[loss=0.1178, simple_loss=0.1985, pruned_loss=0.0186, over 7352.00 frames. ], tot_loss[loss=0.1276, simple_loss=0.2089, pruned_loss=0.02317, over 1443936.84 frames. ], batch size: 44, lr: 3.74e-03, grad_scale: 8.0 +2023-03-21 11:33:26,039 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.94 vs. limit=5.0 +2023-03-21 11:33:27,726 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 11:33:33,349 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 11:33:34,358 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 11:33:40,523 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=123428.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:33:42,919 INFO [train.py:901] (0/2) Epoch 44, batch 2000, loss[loss=0.108, simple_loss=0.1767, pruned_loss=0.01962, over 5952.00 frames. ], tot_loss[loss=0.1274, simple_loss=0.2086, pruned_loss=0.02309, over 1441770.90 frames. ], batch size: 26, lr: 3.74e-03, grad_scale: 8.0 +2023-03-21 11:33:50,460 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 11:33:56,969 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.089e+02 1.719e+02 1.918e+02 2.343e+02 3.686e+02, threshold=3.836e+02, percent-clipped=0.0 +2023-03-21 11:34:01,604 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 11:34:06,056 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.82 vs. limit=2.0 +2023-03-21 11:34:08,695 INFO [train.py:901] (0/2) Epoch 44, batch 2050, loss[loss=0.1081, simple_loss=0.1909, pruned_loss=0.01265, over 7316.00 frames. ], tot_loss[loss=0.127, simple_loss=0.2083, pruned_loss=0.02283, over 1442569.74 frames. ], batch size: 44, lr: 3.74e-03, grad_scale: 8.0 +2023-03-21 11:34:10,733 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 11:34:11,876 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=123489.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:34:32,057 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=123528.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:34:34,500 INFO [train.py:901] (0/2) Epoch 44, batch 2100, loss[loss=0.1405, simple_loss=0.2247, pruned_loss=0.02815, over 7313.00 frames. ], tot_loss[loss=0.1275, simple_loss=0.2088, pruned_loss=0.02311, over 1441788.76 frames. ], batch size: 59, lr: 3.74e-03, grad_scale: 8.0 +2023-03-21 11:34:42,622 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 11:34:45,131 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 11:34:48,571 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.323e+02 1.796e+02 2.189e+02 2.494e+02 4.694e+02, threshold=4.378e+02, percent-clipped=1.0 +2023-03-21 11:34:56,949 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=123576.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:35:01,051 INFO [train.py:901] (0/2) Epoch 44, batch 2150, loss[loss=0.1131, simple_loss=0.1972, pruned_loss=0.0145, over 7345.00 frames. ], tot_loss[loss=0.1282, simple_loss=0.2097, pruned_loss=0.02333, over 1444017.73 frames. ], batch size: 61, lr: 3.74e-03, grad_scale: 8.0 +2023-03-21 11:35:11,616 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8191, 3.2112, 2.7841, 3.1352, 3.0244, 2.6954, 3.0291, 2.8797], + device='cuda:0'), covar=tensor([0.1115, 0.0663, 0.0769, 0.0691, 0.1246, 0.0699, 0.1135, 0.1153], + device='cuda:0'), in_proj_covar=tensor([0.0062, 0.0061, 0.0069, 0.0061, 0.0058, 0.0065, 0.0058, 0.0056], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:35:15,123 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9531, 3.8987, 2.9398, 3.5007, 2.8412, 2.1767, 1.6087, 3.9637], + device='cuda:0'), covar=tensor([0.0055, 0.0067, 0.0200, 0.0092, 0.0218, 0.0706, 0.0793, 0.0061], + device='cuda:0'), in_proj_covar=tensor([0.0118, 0.0094, 0.0116, 0.0098, 0.0133, 0.0136, 0.0130, 0.0107], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 11:35:26,689 INFO [train.py:901] (0/2) Epoch 44, batch 2200, loss[loss=0.1421, simple_loss=0.2179, pruned_loss=0.03314, over 7226.00 frames. ], tot_loss[loss=0.1282, simple_loss=0.2095, pruned_loss=0.02346, over 1443027.04 frames. ], batch size: 45, lr: 3.74e-03, grad_scale: 8.0 +2023-03-21 11:35:33,363 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 11:35:34,649 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.14 vs. limit=2.0 +2023-03-21 11:35:39,116 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6013, 1.8317, 1.4600, 1.8058, 1.8014, 1.6922, 1.8357, 1.3361], + device='cuda:0'), covar=tensor([0.0163, 0.0168, 0.0345, 0.0199, 0.0144, 0.0160, 0.0144, 0.0235], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0039, 0.0037, 0.0040, 0.0037, 0.0036, 0.0038, 0.0047], + device='cuda:0'), out_proj_covar=tensor([4.4473e-05, 4.3402e-05, 4.2144e-05, 4.3945e-05, 4.1195e-05, 3.9943e-05, + 4.2832e-05, 5.1759e-05], device='cuda:0') +2023-03-21 11:35:40,747 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-03-21 11:35:41,438 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.097e+02 1.824e+02 2.081e+02 2.587e+02 3.989e+02, threshold=4.161e+02, percent-clipped=0.0 +2023-03-21 11:35:46,038 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=123670.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:35:51,204 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5841, 3.8411, 3.5935, 3.8159, 3.4482, 3.7154, 4.0765, 4.1090], + device='cuda:0'), covar=tensor([0.0243, 0.0149, 0.0251, 0.0177, 0.0304, 0.0364, 0.0187, 0.0155], + device='cuda:0'), in_proj_covar=tensor([0.0124, 0.0127, 0.0121, 0.0127, 0.0114, 0.0101, 0.0098, 0.0101], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 11:35:53,141 INFO [train.py:901] (0/2) Epoch 44, batch 2250, loss[loss=0.1274, simple_loss=0.2129, pruned_loss=0.02091, over 7368.00 frames. ], tot_loss[loss=0.1281, simple_loss=0.2096, pruned_loss=0.02333, over 1444205.12 frames. ], batch size: 73, lr: 3.74e-03, grad_scale: 8.0 +2023-03-21 11:36:06,801 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 11:36:07,295 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 11:36:18,196 INFO [train.py:901] (0/2) Epoch 44, batch 2300, loss[loss=0.1135, simple_loss=0.2037, pruned_loss=0.01167, over 7292.00 frames. ], tot_loss[loss=0.1282, simple_loss=0.2096, pruned_loss=0.0234, over 1443724.99 frames. ], batch size: 68, lr: 3.73e-03, grad_scale: 8.0 +2023-03-21 11:36:19,350 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 11:36:30,041 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5461, 4.3431, 3.9052, 3.9818, 3.5912, 2.5618, 1.8274, 4.4947], + device='cuda:0'), covar=tensor([0.0040, 0.0055, 0.0090, 0.0070, 0.0138, 0.0523, 0.0653, 0.0045], + device='cuda:0'), in_proj_covar=tensor([0.0118, 0.0094, 0.0116, 0.0098, 0.0133, 0.0136, 0.0130, 0.0107], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 11:36:33,460 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.213e+02 1.843e+02 2.130e+02 2.540e+02 4.098e+02, threshold=4.259e+02, percent-clipped=0.0 +2023-03-21 11:36:44,418 INFO [train.py:901] (0/2) Epoch 44, batch 2350, loss[loss=0.1513, simple_loss=0.2324, pruned_loss=0.03506, over 7283.00 frames. ], tot_loss[loss=0.1285, simple_loss=0.2099, pruned_loss=0.0235, over 1446432.12 frames. ], batch size: 52, lr: 3.73e-03, grad_scale: 8.0 +2023-03-21 11:36:45,009 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=123784.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:36:59,482 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0705, 3.7968, 3.7256, 3.7111, 3.7730, 3.6475, 3.9246, 3.6087], + device='cuda:0'), covar=tensor([0.0149, 0.0174, 0.0134, 0.0193, 0.0413, 0.0132, 0.0164, 0.0193], + device='cuda:0'), in_proj_covar=tensor([0.0102, 0.0108, 0.0109, 0.0094, 0.0187, 0.0114, 0.0112, 0.0121], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 11:37:00,968 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9412, 4.4731, 4.3493, 4.9046, 4.7321, 4.8273, 4.1218, 4.5360], + device='cuda:0'), covar=tensor([0.0874, 0.2491, 0.2278, 0.0949, 0.0891, 0.1146, 0.0828, 0.0973], + device='cuda:0'), in_proj_covar=tensor([0.0271, 0.0398, 0.0297, 0.0314, 0.0231, 0.0369, 0.0234, 0.0280], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 11:37:06,538 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 11:37:10,574 INFO [train.py:901] (0/2) Epoch 44, batch 2400, loss[loss=0.1396, simple_loss=0.2294, pruned_loss=0.02489, over 7284.00 frames. ], tot_loss[loss=0.1281, simple_loss=0.2094, pruned_loss=0.02333, over 1442894.60 frames. ], batch size: 66, lr: 3.73e-03, grad_scale: 8.0 +2023-03-21 11:37:13,017 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 11:37:24,327 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 11:37:25,317 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.028e+02 1.694e+02 1.923e+02 2.237e+02 4.292e+02, threshold=3.846e+02, percent-clipped=1.0 +2023-03-21 11:37:27,387 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 11:37:31,126 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=123872.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:37:36,352 INFO [train.py:901] (0/2) Epoch 44, batch 2450, loss[loss=0.1331, simple_loss=0.2168, pruned_loss=0.02472, over 7284.00 frames. ], tot_loss[loss=0.1281, simple_loss=0.2096, pruned_loss=0.02327, over 1445289.30 frames. ], batch size: 66, lr: 3.73e-03, grad_scale: 8.0 +2023-03-21 11:37:54,159 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 11:38:02,812 INFO [train.py:901] (0/2) Epoch 44, batch 2500, loss[loss=0.1316, simple_loss=0.2126, pruned_loss=0.0253, over 7353.00 frames. ], tot_loss[loss=0.1283, simple_loss=0.2097, pruned_loss=0.02343, over 1443835.12 frames. ], batch size: 63, lr: 3.73e-03, grad_scale: 8.0 +2023-03-21 11:38:02,957 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=123933.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:38:03,927 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1790, 3.0649, 3.4695, 3.2605, 3.4837, 3.1314, 2.7715, 3.3192], + device='cuda:0'), covar=tensor([0.1697, 0.0592, 0.0791, 0.1040, 0.0651, 0.0889, 0.1962, 0.1258], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0072, 0.0053, 0.0053, 0.0053, 0.0051, 0.0071, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 11:38:13,510 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.93 vs. limit=5.0 +2023-03-21 11:38:16,734 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.179e+02 1.717e+02 1.944e+02 2.434e+02 4.068e+02, threshold=3.888e+02, percent-clipped=1.0 +2023-03-21 11:38:19,729 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 11:38:21,341 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=123970.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:38:27,698 INFO [train.py:901] (0/2) Epoch 44, batch 2550, loss[loss=0.1304, simple_loss=0.2126, pruned_loss=0.02411, over 7200.00 frames. ], tot_loss[loss=0.1281, simple_loss=0.2093, pruned_loss=0.02342, over 1443529.43 frames. ], batch size: 50, lr: 3.73e-03, grad_scale: 8.0 +2023-03-21 11:38:30,871 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1914, 2.5976, 3.1456, 3.0870, 3.2624, 2.8604, 2.7804, 3.0908], + device='cuda:0'), covar=tensor([0.1044, 0.0730, 0.1134, 0.0973, 0.0622, 0.1188, 0.1517, 0.1231], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0072, 0.0054, 0.0053, 0.0053, 0.0052, 0.0071, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 11:38:37,232 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-124000.pt +2023-03-21 11:38:42,873 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1895, 3.3190, 2.4386, 3.7587, 2.8289, 3.0538, 1.7057, 2.6082], + device='cuda:0'), covar=tensor([0.0449, 0.0799, 0.2701, 0.0504, 0.0470, 0.0675, 0.4030, 0.1736], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0257, 0.0275, 0.0269, 0.0266, 0.0266, 0.0229, 0.0257], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:38:50,871 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=124018.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:38:50,917 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1082, 3.8324, 3.8101, 3.8810, 3.3294, 3.7101, 4.0096, 3.6805], + device='cuda:0'), covar=tensor([0.0267, 0.0243, 0.0184, 0.0215, 0.0870, 0.0205, 0.0283, 0.0258], + device='cuda:0'), in_proj_covar=tensor([0.0102, 0.0108, 0.0109, 0.0094, 0.0187, 0.0114, 0.0112, 0.0120], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 11:38:58,387 INFO [train.py:901] (0/2) Epoch 44, batch 2600, loss[loss=0.09412, simple_loss=0.1518, pruned_loss=0.01824, over 6378.00 frames. ], tot_loss[loss=0.1278, simple_loss=0.2088, pruned_loss=0.02334, over 1442661.37 frames. ], batch size: 27, lr: 3.73e-03, grad_scale: 8.0 +2023-03-21 11:39:00,104 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=124036.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:39:12,321 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.198e+02 1.686e+02 1.984e+02 2.417e+02 5.230e+02, threshold=3.969e+02, percent-clipped=2.0 +2023-03-21 11:39:12,913 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=124062.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:39:22,977 INFO [train.py:901] (0/2) Epoch 44, batch 2650, loss[loss=0.102, simple_loss=0.181, pruned_loss=0.01147, over 7159.00 frames. ], tot_loss[loss=0.128, simple_loss=0.209, pruned_loss=0.02351, over 1441606.43 frames. ], batch size: 39, lr: 3.73e-03, grad_scale: 8.0 +2023-03-21 11:39:23,554 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=124084.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:39:30,144 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=124097.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:39:34,704 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9815, 2.6780, 2.9473, 3.0043, 2.7683, 2.7137, 2.9864, 2.2265], + device='cuda:0'), covar=tensor([0.0653, 0.0666, 0.0841, 0.0761, 0.0760, 0.1158, 0.0826, 0.2637], + device='cuda:0'), in_proj_covar=tensor([0.0330, 0.0336, 0.0271, 0.0353, 0.0281, 0.0286, 0.0348, 0.0239], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 11:39:43,212 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=124123.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:39:47,587 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=124132.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:39:48,035 INFO [train.py:901] (0/2) Epoch 44, batch 2700, loss[loss=0.1206, simple_loss=0.2052, pruned_loss=0.018, over 7311.00 frames. ], tot_loss[loss=0.1273, simple_loss=0.2083, pruned_loss=0.02315, over 1442637.70 frames. ], batch size: 59, lr: 3.73e-03, grad_scale: 8.0 +2023-03-21 11:40:01,925 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.189e+02 1.704e+02 1.876e+02 2.261e+02 3.990e+02, threshold=3.752e+02, percent-clipped=1.0 +2023-03-21 11:40:12,784 INFO [train.py:901] (0/2) Epoch 44, batch 2750, loss[loss=0.1148, simple_loss=0.1948, pruned_loss=0.01738, over 7131.00 frames. ], tot_loss[loss=0.1274, simple_loss=0.2084, pruned_loss=0.02324, over 1440764.11 frames. ], batch size: 41, lr: 3.73e-03, grad_scale: 8.0 +2023-03-21 11:40:27,573 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8202, 3.1378, 2.8077, 3.0712, 3.0747, 2.6693, 3.0339, 2.9785], + device='cuda:0'), covar=tensor([0.0583, 0.0336, 0.0860, 0.0716, 0.0679, 0.0613, 0.0769, 0.0795], + device='cuda:0'), in_proj_covar=tensor([0.0061, 0.0060, 0.0069, 0.0061, 0.0058, 0.0064, 0.0057, 0.0055], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 11:40:35,773 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=124228.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:40:38,225 INFO [train.py:901] (0/2) Epoch 44, batch 2800, loss[loss=0.1319, simple_loss=0.2154, pruned_loss=0.02419, over 7274.00 frames. ], tot_loss[loss=0.1277, simple_loss=0.2089, pruned_loss=0.02324, over 1442552.01 frames. ], batch size: 57, lr: 3.73e-03, grad_scale: 8.0 +2023-03-21 11:40:50,741 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-44.pt +2023-03-21 11:41:06,822 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 11:41:06,877 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 11:41:06,903 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 11:41:06,920 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 11:41:07,095 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 11:41:07,111 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 11:41:07,122 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 11:41:07,220 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 11:41:07,296 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 11:41:07,312 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 11:41:07,328 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 11:41:07,842 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 11:41:07,848 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 11:41:07,856 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 11:41:07,998 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 11:41:08,001 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 11:41:08,060 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 11:41:10,380 INFO [train.py:901] (0/2) Epoch 45, batch 0, loss[loss=0.1344, simple_loss=0.2204, pruned_loss=0.02424, over 7365.00 frames. ], tot_loss[loss=0.1344, simple_loss=0.2204, pruned_loss=0.02424, over 7365.00 frames. ], batch size: 65, lr: 3.69e-03, grad_scale: 8.0 +2023-03-21 11:41:10,381 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 11:41:36,462 INFO [train.py:935] (0/2) Epoch 45, validation: loss=0.1662, simple_loss=0.2586, pruned_loss=0.03694, over 1622729.00 frames. +2023-03-21 11:41:36,463 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 11:41:39,079 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.245e+02 1.746e+02 2.054e+02 2.421e+02 4.265e+02, threshold=4.108e+02, percent-clipped=2.0 +2023-03-21 11:41:43,536 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 11:41:44,372 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.54 vs. limit=5.0 +2023-03-21 11:41:44,691 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=124272.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:41:53,516 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 11:41:57,790 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.38 vs. limit=2.0 +2023-03-21 11:42:00,542 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 11:42:01,983 INFO [train.py:901] (0/2) Epoch 45, batch 50, loss[loss=0.1388, simple_loss=0.2305, pruned_loss=0.02359, over 6632.00 frames. ], tot_loss[loss=0.1291, simple_loss=0.2116, pruned_loss=0.02336, over 326423.20 frames. ], batch size: 107, lr: 3.68e-03, grad_scale: 8.0 +2023-03-21 11:42:03,030 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 11:42:05,483 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 11:42:05,786 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-03-21 11:42:15,160 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=124333.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:42:22,759 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 11:42:23,238 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 11:42:24,046 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=124348.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 11:42:28,453 INFO [train.py:901] (0/2) Epoch 45, batch 100, loss[loss=0.1153, simple_loss=0.1893, pruned_loss=0.02064, over 7228.00 frames. ], tot_loss[loss=0.1266, simple_loss=0.209, pruned_loss=0.0221, over 573970.43 frames. ], batch size: 45, lr: 3.68e-03, grad_scale: 8.0 +2023-03-21 11:42:30,319 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.176e+02 1.695e+02 1.936e+02 2.313e+02 3.507e+02, threshold=3.872e+02, percent-clipped=0.0 +2023-03-21 11:42:42,105 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2620, 3.2873, 2.2901, 3.5679, 2.8565, 3.1526, 1.5731, 2.4350], + device='cuda:0'), covar=tensor([0.0572, 0.0873, 0.3093, 0.0812, 0.0646, 0.0904, 0.4079, 0.2047], + device='cuda:0'), in_proj_covar=tensor([0.0277, 0.0258, 0.0276, 0.0271, 0.0269, 0.0265, 0.0229, 0.0258], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:42:46,003 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=124392.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:42:48,695 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.25 vs. limit=2.0 +2023-03-21 11:42:51,581 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-03-21 11:42:52,930 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7627, 1.4992, 1.8719, 2.0846, 1.8933, 2.1317, 1.5739, 2.2211], + device='cuda:0'), covar=tensor([0.2012, 0.3604, 0.1253, 0.1020, 0.1033, 0.1655, 0.1749, 0.1575], + device='cuda:0'), in_proj_covar=tensor([0.0082, 0.0082, 0.0075, 0.0068, 0.0068, 0.0067, 0.0107, 0.0069], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 11:42:53,742 INFO [train.py:901] (0/2) Epoch 45, batch 150, loss[loss=0.1324, simple_loss=0.2132, pruned_loss=0.0258, over 7310.00 frames. ], tot_loss[loss=0.1283, simple_loss=0.2106, pruned_loss=0.02295, over 768679.35 frames. ], batch size: 80, lr: 3.68e-03, grad_scale: 8.0 +2023-03-21 11:42:54,905 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=124409.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 11:42:59,356 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=124418.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:43:07,819 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=124432.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:43:20,195 INFO [train.py:901] (0/2) Epoch 45, batch 200, loss[loss=0.1236, simple_loss=0.2109, pruned_loss=0.01814, over 7312.00 frames. ], tot_loss[loss=0.1279, simple_loss=0.2103, pruned_loss=0.02275, over 917634.64 frames. ], batch size: 83, lr: 3.68e-03, grad_scale: 8.0 +2023-03-21 11:43:22,218 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.227e+02 1.748e+02 2.057e+02 2.417e+02 4.468e+02, threshold=4.114e+02, percent-clipped=4.0 +2023-03-21 11:43:22,770 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 11:43:26,634 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 11:43:32,682 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 11:43:38,384 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=124493.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:43:45,281 INFO [train.py:901] (0/2) Epoch 45, batch 250, loss[loss=0.1176, simple_loss=0.2037, pruned_loss=0.01573, over 7382.00 frames. ], tot_loss[loss=0.1277, simple_loss=0.2098, pruned_loss=0.0228, over 1033196.47 frames. ], batch size: 65, lr: 3.68e-03, grad_scale: 8.0 +2023-03-21 11:43:45,791 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 11:43:56,916 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=124528.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:44:06,275 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 11:44:11,389 INFO [train.py:901] (0/2) Epoch 45, batch 300, loss[loss=0.09841, simple_loss=0.1604, pruned_loss=0.01821, over 5937.00 frames. ], tot_loss[loss=0.1281, simple_loss=0.2099, pruned_loss=0.02311, over 1124057.22 frames. ], batch size: 25, lr: 3.68e-03, grad_scale: 8.0 +2023-03-21 11:44:13,403 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.259e+02 1.695e+02 1.987e+02 2.294e+02 4.140e+02, threshold=3.974e+02, percent-clipped=0.0 +2023-03-21 11:44:15,448 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 11:44:17,086 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9635, 2.5225, 3.0732, 2.7697, 2.9994, 2.8814, 2.4989, 3.0646], + device='cuda:0'), covar=tensor([0.1469, 0.0812, 0.0886, 0.1374, 0.0784, 0.1000, 0.1985, 0.0943], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0071, 0.0053, 0.0053, 0.0052, 0.0051, 0.0070, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 11:44:18,294 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.66 vs. limit=2.0 +2023-03-21 11:44:21,063 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=124576.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:44:38,058 INFO [train.py:901] (0/2) Epoch 45, batch 350, loss[loss=0.1243, simple_loss=0.2122, pruned_loss=0.01819, over 7112.00 frames. ], tot_loss[loss=0.1272, simple_loss=0.2088, pruned_loss=0.02279, over 1192177.24 frames. ], batch size: 98, lr: 3.68e-03, grad_scale: 8.0 +2023-03-21 11:44:47,251 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1759, 2.5494, 2.0479, 2.9250, 2.6584, 3.0761, 2.5975, 2.4166], + device='cuda:0'), covar=tensor([0.2434, 0.1177, 0.4050, 0.0856, 0.0318, 0.0324, 0.0424, 0.0405], + device='cuda:0'), in_proj_covar=tensor([0.0252, 0.0233, 0.0246, 0.0261, 0.0202, 0.0207, 0.0222, 0.0231], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:44:48,217 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2058, 2.3170, 2.6134, 2.3765, 2.1624, 2.3574, 2.3798, 1.8823], + device='cuda:0'), covar=tensor([0.0578, 0.0570, 0.0390, 0.0272, 0.0880, 0.0625, 0.0424, 0.0380], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0041, 0.0042, 0.0041, 0.0038, 0.0039, 0.0045, 0.0044], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 11:44:48,633 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 11:44:48,685 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=124628.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:44:53,756 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=124638.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:45:03,206 INFO [train.py:901] (0/2) Epoch 45, batch 400, loss[loss=0.1286, simple_loss=0.2088, pruned_loss=0.02419, over 7209.00 frames. ], tot_loss[loss=0.1273, simple_loss=0.2089, pruned_loss=0.02289, over 1248400.54 frames. ], batch size: 93, lr: 3.68e-03, grad_scale: 8.0 +2023-03-21 11:45:05,234 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.191e+02 1.772e+02 2.009e+02 2.324e+02 6.365e+02, threshold=4.017e+02, percent-clipped=3.0 +2023-03-21 11:45:06,390 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=124663.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:45:22,296 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=124692.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:45:25,864 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=124699.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:45:28,307 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=124704.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 11:45:28,867 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6411, 3.3182, 3.4270, 3.4982, 3.0886, 3.0897, 3.6369, 2.5229], + device='cuda:0'), covar=tensor([0.0627, 0.0623, 0.0788, 0.0770, 0.0951, 0.1220, 0.0751, 0.2945], + device='cuda:0'), in_proj_covar=tensor([0.0328, 0.0338, 0.0270, 0.0353, 0.0283, 0.0285, 0.0348, 0.0239], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 11:45:29,689 INFO [train.py:901] (0/2) Epoch 45, batch 450, loss[loss=0.113, simple_loss=0.1956, pruned_loss=0.01521, over 7340.00 frames. ], tot_loss[loss=0.1269, simple_loss=0.2085, pruned_loss=0.02269, over 1292563.63 frames. ], batch size: 44, lr: 3.68e-03, grad_scale: 8.0 +2023-03-21 11:45:31,689 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 11:45:32,158 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 11:45:35,273 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=124718.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:45:38,356 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=124724.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:45:46,278 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=124740.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:45:54,793 INFO [train.py:901] (0/2) Epoch 45, batch 500, loss[loss=0.1229, simple_loss=0.2084, pruned_loss=0.01872, over 7360.00 frames. ], tot_loss[loss=0.1266, simple_loss=0.2081, pruned_loss=0.02252, over 1324803.73 frames. ], batch size: 63, lr: 3.68e-03, grad_scale: 8.0 +2023-03-21 11:45:56,800 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.234e+02 1.743e+02 1.901e+02 2.274e+02 3.978e+02, threshold=3.801e+02, percent-clipped=0.0 +2023-03-21 11:46:00,029 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=124766.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:46:03,651 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1683, 3.0096, 3.1939, 2.9602, 3.1463, 3.1157, 2.8205, 2.9304], + device='cuda:0'), covar=tensor([0.1632, 0.0693, 0.1191, 0.2162, 0.0951, 0.1041, 0.1718, 0.1920], + device='cuda:0'), in_proj_covar=tensor([0.0055, 0.0071, 0.0053, 0.0054, 0.0053, 0.0051, 0.0071, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 11:46:06,188 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 11:46:07,761 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 11:46:08,769 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 11:46:10,795 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 11:46:11,899 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=124788.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:46:12,578 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-03-21 11:46:12,997 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0128, 2.8482, 2.0120, 3.0857, 2.5720, 2.8739, 1.4142, 2.1218], + device='cuda:0'), covar=tensor([0.0795, 0.1574, 0.3436, 0.1016, 0.0696, 0.0842, 0.4228, 0.2081], + device='cuda:0'), in_proj_covar=tensor([0.0277, 0.0259, 0.0276, 0.0270, 0.0269, 0.0265, 0.0230, 0.0258], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:46:15,274 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 11:46:21,540 INFO [train.py:901] (0/2) Epoch 45, batch 550, loss[loss=0.1412, simple_loss=0.2245, pruned_loss=0.02891, over 7342.00 frames. ], tot_loss[loss=0.1269, simple_loss=0.2086, pruned_loss=0.02259, over 1351244.22 frames. ], batch size: 73, lr: 3.68e-03, grad_scale: 8.0 +2023-03-21 11:46:26,605 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 11:46:28,263 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=124820.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:46:29,896 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.17 vs. limit=2.0 +2023-03-21 11:46:34,835 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 11:46:38,258 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 11:46:44,974 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 11:46:47,465 INFO [train.py:901] (0/2) Epoch 45, batch 600, loss[loss=0.1268, simple_loss=0.2108, pruned_loss=0.02142, over 7236.00 frames. ], tot_loss[loss=0.1274, simple_loss=0.2091, pruned_loss=0.02289, over 1371919.14 frames. ], batch size: 93, lr: 3.68e-03, grad_scale: 8.0 +2023-03-21 11:46:50,047 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.309e+02 1.678e+02 1.954e+02 2.245e+02 3.670e+02, threshold=3.907e+02, percent-clipped=0.0 +2023-03-21 11:46:54,257 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5619, 1.8759, 1.5129, 1.7660, 1.8843, 1.7925, 1.7826, 1.4702], + device='cuda:0'), covar=tensor([0.0158, 0.0190, 0.0391, 0.0193, 0.0159, 0.0145, 0.0172, 0.0254], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0039, 0.0037, 0.0039, 0.0037, 0.0036, 0.0039, 0.0047], + device='cuda:0'), out_proj_covar=tensor([4.4498e-05, 4.3589e-05, 4.1776e-05, 4.3465e-05, 4.0993e-05, 4.0135e-05, + 4.3080e-05, 5.1541e-05], device='cuda:0') +2023-03-21 11:47:00,332 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=124881.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:47:01,766 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 11:47:10,731 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 11:47:13,268 INFO [train.py:901] (0/2) Epoch 45, batch 650, loss[loss=0.1298, simple_loss=0.2156, pruned_loss=0.02202, over 7219.00 frames. ], tot_loss[loss=0.127, simple_loss=0.2086, pruned_loss=0.0227, over 1389282.61 frames. ], batch size: 50, lr: 3.68e-03, grad_scale: 8.0 +2023-03-21 11:47:19,712 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.92 vs. limit=2.0 +2023-03-21 11:47:23,967 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=124928.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:47:27,843 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 11:47:37,627 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 11:47:39,632 INFO [train.py:901] (0/2) Epoch 45, batch 700, loss[loss=0.1275, simple_loss=0.2118, pruned_loss=0.02154, over 7273.00 frames. ], tot_loss[loss=0.1267, simple_loss=0.2082, pruned_loss=0.02265, over 1397451.06 frames. ], batch size: 89, lr: 3.67e-03, grad_scale: 8.0 +2023-03-21 11:47:41,670 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.255e+02 1.758e+02 2.104e+02 2.521e+02 4.438e+02, threshold=4.207e+02, percent-clipped=2.0 +2023-03-21 11:47:42,270 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6037, 4.1570, 4.0361, 4.5084, 4.4109, 4.4848, 3.8725, 4.1804], + device='cuda:0'), covar=tensor([0.0956, 0.2880, 0.2950, 0.1299, 0.1025, 0.1318, 0.1013, 0.1372], + device='cuda:0'), in_proj_covar=tensor([0.0272, 0.0405, 0.0305, 0.0321, 0.0237, 0.0376, 0.0237, 0.0284], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 11:47:49,170 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=124976.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:47:55,350 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2938, 4.0291, 3.4727, 3.7526, 3.1190, 2.2966, 1.8308, 4.2407], + device='cuda:0'), covar=tensor([0.0043, 0.0083, 0.0133, 0.0078, 0.0172, 0.0558, 0.0665, 0.0048], + device='cuda:0'), in_proj_covar=tensor([0.0117, 0.0094, 0.0115, 0.0098, 0.0132, 0.0134, 0.0128, 0.0106], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 11:47:58,306 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=124994.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:48:01,673 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 11:48:02,162 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 11:48:03,735 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=125004.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 11:48:05,254 INFO [train.py:901] (0/2) Epoch 45, batch 750, loss[loss=0.1268, simple_loss=0.2086, pruned_loss=0.02252, over 7284.00 frames. ], tot_loss[loss=0.127, simple_loss=0.2086, pruned_loss=0.02275, over 1410358.37 frames. ], batch size: 66, lr: 3.67e-03, grad_scale: 8.0 +2023-03-21 11:48:11,399 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=125019.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:48:16,012 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 11:48:21,594 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 11:48:21,875 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.65 vs. limit=2.0 +2023-03-21 11:48:28,156 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 11:48:29,178 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 11:48:29,212 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=125052.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 11:48:31,655 INFO [train.py:901] (0/2) Epoch 45, batch 800, loss[loss=0.1367, simple_loss=0.211, pruned_loss=0.03121, over 7243.00 frames. ], tot_loss[loss=0.1274, simple_loss=0.2087, pruned_loss=0.02303, over 1417384.20 frames. ], batch size: 55, lr: 3.67e-03, grad_scale: 8.0 +2023-03-21 11:48:33,690 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.161e+02 1.703e+02 1.941e+02 2.250e+02 3.887e+02, threshold=3.881e+02, percent-clipped=0.0 +2023-03-21 11:48:40,389 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 11:48:47,466 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=125088.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:48:56,889 INFO [train.py:901] (0/2) Epoch 45, batch 850, loss[loss=0.1341, simple_loss=0.2192, pruned_loss=0.02451, over 7374.00 frames. ], tot_loss[loss=0.127, simple_loss=0.2086, pruned_loss=0.02271, over 1425036.42 frames. ], batch size: 65, lr: 3.67e-03, grad_scale: 8.0 +2023-03-21 11:48:58,969 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 11:48:58,979 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 11:49:04,606 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 11:49:08,178 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 11:49:08,779 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=125128.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:49:10,316 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3184, 2.3223, 2.5482, 2.2681, 2.3446, 2.3277, 2.2465, 1.8601], + device='cuda:0'), covar=tensor([0.0664, 0.0372, 0.0405, 0.0361, 0.0519, 0.0506, 0.0348, 0.0300], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0041, 0.0042, 0.0041, 0.0038, 0.0039, 0.0045, 0.0044], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 11:49:12,735 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=125136.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:49:13,252 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.0662, 4.5998, 4.3666, 4.9705, 4.7935, 4.9655, 4.4307, 4.6248], + device='cuda:0'), covar=tensor([0.0841, 0.2605, 0.2823, 0.1078, 0.1078, 0.1205, 0.0708, 0.1183], + device='cuda:0'), in_proj_covar=tensor([0.0270, 0.0402, 0.0302, 0.0319, 0.0236, 0.0375, 0.0236, 0.0283], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 11:49:23,162 INFO [train.py:901] (0/2) Epoch 45, batch 900, loss[loss=0.1336, simple_loss=0.2152, pruned_loss=0.026, over 7297.00 frames. ], tot_loss[loss=0.1274, simple_loss=0.2091, pruned_loss=0.02284, over 1431672.55 frames. ], batch size: 66, lr: 3.67e-03, grad_scale: 8.0 +2023-03-21 11:49:25,151 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.155e+02 1.791e+02 2.042e+02 2.328e+02 6.599e+02, threshold=4.084e+02, percent-clipped=1.0 +2023-03-21 11:49:26,255 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3859, 4.8893, 5.0042, 4.9242, 4.8287, 4.4189, 5.0176, 4.8368], + device='cuda:0'), covar=tensor([0.0467, 0.0364, 0.0309, 0.0476, 0.0316, 0.0379, 0.0299, 0.0437], + device='cuda:0'), in_proj_covar=tensor([0.0225, 0.0274, 0.0210, 0.0210, 0.0162, 0.0238, 0.0218, 0.0154], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 11:49:32,745 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=125176.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:49:39,357 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=125189.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:49:39,965 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-03-21 11:49:47,732 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 11:49:47,799 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3879, 4.8725, 4.9391, 4.8944, 4.8065, 4.4065, 4.9698, 4.8027], + device='cuda:0'), covar=tensor([0.0496, 0.0411, 0.0401, 0.0482, 0.0398, 0.0435, 0.0340, 0.0527], + device='cuda:0'), in_proj_covar=tensor([0.0225, 0.0274, 0.0210, 0.0210, 0.0162, 0.0238, 0.0218, 0.0154], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 11:49:49,225 INFO [train.py:901] (0/2) Epoch 45, batch 950, loss[loss=0.1267, simple_loss=0.2064, pruned_loss=0.02352, over 7210.00 frames. ], tot_loss[loss=0.1267, simple_loss=0.2083, pruned_loss=0.02251, over 1435743.20 frames. ], batch size: 50, lr: 3.67e-03, grad_scale: 8.0 +2023-03-21 11:49:54,011 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5216, 3.3913, 3.5438, 3.5451, 3.1915, 3.0977, 3.7639, 2.5664], + device='cuda:0'), covar=tensor([0.0468, 0.0633, 0.0761, 0.0726, 0.0794, 0.1060, 0.0886, 0.2861], + device='cuda:0'), in_proj_covar=tensor([0.0327, 0.0334, 0.0270, 0.0350, 0.0282, 0.0284, 0.0345, 0.0238], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 11:49:55,475 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6677, 1.5236, 1.8428, 2.1030, 1.8486, 2.1241, 1.5737, 2.1903], + device='cuda:0'), covar=tensor([0.2233, 0.3804, 0.1707, 0.1172, 0.1476, 0.1125, 0.1945, 0.1388], + device='cuda:0'), in_proj_covar=tensor([0.0083, 0.0082, 0.0076, 0.0069, 0.0069, 0.0066, 0.0108, 0.0069], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 11:50:02,014 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7393, 2.1090, 1.6397, 2.1099, 2.1566, 1.9857, 1.8868, 1.5870], + device='cuda:0'), covar=tensor([0.0176, 0.0259, 0.0240, 0.0179, 0.0132, 0.0176, 0.0269, 0.0245], + device='cuda:0'), in_proj_covar=tensor([0.0040, 0.0039, 0.0037, 0.0040, 0.0037, 0.0036, 0.0039, 0.0047], + device='cuda:0'), out_proj_covar=tensor([4.5026e-05, 4.3169e-05, 4.1990e-05, 4.3600e-05, 4.0986e-05, 3.9963e-05, + 4.3030e-05, 5.1229e-05], device='cuda:0') +2023-03-21 11:50:11,379 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 11:50:14,958 INFO [train.py:901] (0/2) Epoch 45, batch 1000, loss[loss=0.1209, simple_loss=0.2129, pruned_loss=0.01448, over 7308.00 frames. ], tot_loss[loss=0.1262, simple_loss=0.2077, pruned_loss=0.02236, over 1437726.90 frames. ], batch size: 80, lr: 3.67e-03, grad_scale: 16.0 +2023-03-21 11:50:16,956 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.326e+02 1.650e+02 2.037e+02 2.329e+02 4.165e+02, threshold=4.073e+02, percent-clipped=1.0 +2023-03-21 11:50:32,606 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 11:50:34,246 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=125294.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:50:36,327 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9834, 3.2493, 2.9555, 3.1508, 3.1371, 2.9877, 3.2391, 3.1494], + device='cuda:0'), covar=tensor([0.0896, 0.0810, 0.0957, 0.1269, 0.1161, 0.0553, 0.0852, 0.0826], + device='cuda:0'), in_proj_covar=tensor([0.0061, 0.0060, 0.0069, 0.0061, 0.0057, 0.0065, 0.0058, 0.0055], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 11:50:41,358 INFO [train.py:901] (0/2) Epoch 45, batch 1050, loss[loss=0.1296, simple_loss=0.2141, pruned_loss=0.02261, over 7154.00 frames. ], tot_loss[loss=0.1264, simple_loss=0.2077, pruned_loss=0.02256, over 1438785.68 frames. ], batch size: 98, lr: 3.67e-03, grad_scale: 16.0 +2023-03-21 11:50:47,377 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=125319.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:50:54,319 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 11:50:58,336 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 11:50:58,881 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=125342.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:51:06,067 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=125356.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:51:06,439 INFO [train.py:901] (0/2) Epoch 45, batch 1100, loss[loss=0.1232, simple_loss=0.2121, pruned_loss=0.01714, over 7285.00 frames. ], tot_loss[loss=0.1267, simple_loss=0.2084, pruned_loss=0.02248, over 1440998.08 frames. ], batch size: 66, lr: 3.67e-03, grad_scale: 16.0 +2023-03-21 11:51:08,423 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.160e+02 1.764e+02 2.014e+02 2.431e+02 3.712e+02, threshold=4.027e+02, percent-clipped=0.0 +2023-03-21 11:51:11,480 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=125367.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:51:16,812 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=125376.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:51:28,863 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 11:51:29,723 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 11:51:33,195 INFO [train.py:901] (0/2) Epoch 45, batch 1150, loss[loss=0.1351, simple_loss=0.2253, pruned_loss=0.02243, over 6765.00 frames. ], tot_loss[loss=0.1266, simple_loss=0.2082, pruned_loss=0.02246, over 1436727.01 frames. ], batch size: 107, lr: 3.67e-03, grad_scale: 16.0 +2023-03-21 11:51:38,461 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=125417.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 11:51:40,424 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=125421.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:51:42,318 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 11:51:42,788 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 11:51:48,377 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=125437.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:51:58,988 INFO [train.py:901] (0/2) Epoch 45, batch 1200, loss[loss=0.1324, simple_loss=0.2192, pruned_loss=0.02284, over 7285.00 frames. ], tot_loss[loss=0.1265, simple_loss=0.2081, pruned_loss=0.02246, over 1437924.54 frames. ], batch size: 68, lr: 3.67e-03, grad_scale: 16.0 +2023-03-21 11:52:00,948 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.253e+02 1.727e+02 2.078e+02 2.554e+02 4.178e+02, threshold=4.157e+02, percent-clipped=1.0 +2023-03-21 11:52:08,688 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=125476.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:52:12,304 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=125482.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:52:13,250 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=125484.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:52:15,199 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 11:52:24,734 INFO [train.py:901] (0/2) Epoch 45, batch 1250, loss[loss=0.1261, simple_loss=0.2042, pruned_loss=0.02393, over 7337.00 frames. ], tot_loss[loss=0.1261, simple_loss=0.2078, pruned_loss=0.02217, over 1438457.30 frames. ], batch size: 54, lr: 3.67e-03, grad_scale: 16.0 +2023-03-21 11:52:31,801 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2112, 4.3528, 4.1090, 4.3222, 3.8957, 4.2960, 4.6139, 4.6008], + device='cuda:0'), covar=tensor([0.0204, 0.0144, 0.0214, 0.0166, 0.0400, 0.0246, 0.0231, 0.0185], + device='cuda:0'), in_proj_covar=tensor([0.0127, 0.0130, 0.0123, 0.0129, 0.0117, 0.0104, 0.0101, 0.0105], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 11:52:33,316 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=125524.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:52:38,871 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 11:52:43,424 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 11:52:44,461 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 11:52:50,561 INFO [train.py:901] (0/2) Epoch 45, batch 1300, loss[loss=0.1211, simple_loss=0.2037, pruned_loss=0.01927, over 7229.00 frames. ], tot_loss[loss=0.1263, simple_loss=0.208, pruned_loss=0.02233, over 1439152.28 frames. ], batch size: 50, lr: 3.67e-03, grad_scale: 16.0 +2023-03-21 11:52:52,515 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.223e+02 1.582e+02 2.005e+02 2.378e+02 3.739e+02, threshold=4.009e+02, percent-clipped=0.0 +2023-03-21 11:53:06,655 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 11:53:08,648 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 11:53:11,266 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5900, 4.1268, 4.1808, 4.2544, 4.2104, 4.1941, 4.5013, 4.0091], + device='cuda:0'), covar=tensor([0.0173, 0.0176, 0.0139, 0.0147, 0.0454, 0.0123, 0.0133, 0.0159], + device='cuda:0'), in_proj_covar=tensor([0.0099, 0.0105, 0.0107, 0.0092, 0.0182, 0.0112, 0.0110, 0.0118], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 11:53:12,176 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 11:53:14,612 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=125603.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:53:16,432 INFO [train.py:901] (0/2) Epoch 45, batch 1350, loss[loss=0.1318, simple_loss=0.2114, pruned_loss=0.02611, over 7281.00 frames. ], tot_loss[loss=0.1267, simple_loss=0.2085, pruned_loss=0.0224, over 1441014.38 frames. ], batch size: 70, lr: 3.67e-03, grad_scale: 8.0 +2023-03-21 11:53:22,962 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 11:53:42,808 INFO [train.py:901] (0/2) Epoch 45, batch 1400, loss[loss=0.1407, simple_loss=0.2183, pruned_loss=0.03156, over 7203.00 frames. ], tot_loss[loss=0.1273, simple_loss=0.2088, pruned_loss=0.02288, over 1440751.12 frames. ], batch size: 50, lr: 3.66e-03, grad_scale: 8.0 +2023-03-21 11:53:45,262 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.279e+02 1.672e+02 2.023e+02 2.311e+02 3.601e+02, threshold=4.045e+02, percent-clipped=0.0 +2023-03-21 11:53:46,459 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=125664.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:53:55,398 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 11:54:07,909 INFO [train.py:901] (0/2) Epoch 45, batch 1450, loss[loss=0.121, simple_loss=0.2059, pruned_loss=0.01801, over 7281.00 frames. ], tot_loss[loss=0.127, simple_loss=0.2085, pruned_loss=0.02272, over 1439442.36 frames. ], batch size: 68, lr: 3.66e-03, grad_scale: 8.0 +2023-03-21 11:54:11,184 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=125712.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 11:54:17,677 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=125725.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:54:20,040 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 11:54:21,126 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=125732.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:54:34,095 INFO [train.py:901] (0/2) Epoch 45, batch 1500, loss[loss=0.1338, simple_loss=0.2123, pruned_loss=0.02769, over 7333.00 frames. ], tot_loss[loss=0.1274, simple_loss=0.2089, pruned_loss=0.02292, over 1442788.73 frames. ], batch size: 49, lr: 3.66e-03, grad_scale: 8.0 +2023-03-21 11:54:36,667 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.334e+02 1.679e+02 1.970e+02 2.341e+02 4.100e+02, threshold=3.939e+02, percent-clipped=1.0 +2023-03-21 11:54:36,719 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 11:54:38,995 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-03-21 11:54:44,241 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=125777.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:54:47,804 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=125784.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:54:48,828 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=125786.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:54:51,857 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2825, 2.8199, 2.1066, 3.1250, 3.1511, 3.1953, 2.7401, 2.8315], + device='cuda:0'), covar=tensor([0.2293, 0.1086, 0.3783, 0.0729, 0.0338, 0.0295, 0.0451, 0.0488], + device='cuda:0'), in_proj_covar=tensor([0.0249, 0.0230, 0.0243, 0.0258, 0.0200, 0.0205, 0.0220, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:55:00,111 INFO [train.py:901] (0/2) Epoch 45, batch 1550, loss[loss=0.1198, simple_loss=0.206, pruned_loss=0.01676, over 7342.00 frames. ], tot_loss[loss=0.1272, simple_loss=0.2086, pruned_loss=0.02291, over 1440728.96 frames. ], batch size: 61, lr: 3.66e-03, grad_scale: 8.0 +2023-03-21 11:55:01,661 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 11:55:02,237 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.1611, 4.6181, 4.5026, 4.9877, 4.9221, 4.9717, 4.3582, 4.6026], + device='cuda:0'), covar=tensor([0.0754, 0.2287, 0.2066, 0.1148, 0.0845, 0.1094, 0.0805, 0.1136], + device='cuda:0'), in_proj_covar=tensor([0.0268, 0.0396, 0.0299, 0.0316, 0.0235, 0.0372, 0.0235, 0.0280], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 11:55:13,187 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=125832.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:55:21,607 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-03-21 11:55:25,719 INFO [train.py:901] (0/2) Epoch 45, batch 1600, loss[loss=0.1324, simple_loss=0.2039, pruned_loss=0.0305, over 7213.00 frames. ], tot_loss[loss=0.1276, simple_loss=0.2091, pruned_loss=0.02309, over 1441828.39 frames. ], batch size: 45, lr: 3.66e-03, grad_scale: 8.0 +2023-03-21 11:55:28,124 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.320e+02 1.720e+02 2.021e+02 2.344e+02 4.247e+02, threshold=4.043e+02, percent-clipped=1.0 +2023-03-21 11:55:32,296 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 11:55:32,897 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 11:55:36,450 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 11:55:39,534 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3011, 3.9420, 3.8980, 3.9274, 3.9135, 3.8601, 4.1891, 3.7181], + device='cuda:0'), covar=tensor([0.0138, 0.0163, 0.0131, 0.0172, 0.0427, 0.0122, 0.0138, 0.0188], + device='cuda:0'), in_proj_covar=tensor([0.0099, 0.0105, 0.0107, 0.0091, 0.0181, 0.0112, 0.0110, 0.0118], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 11:55:46,625 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 11:55:50,623 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 11:55:51,571 INFO [train.py:901] (0/2) Epoch 45, batch 1650, loss[loss=0.1345, simple_loss=0.2169, pruned_loss=0.02604, over 7277.00 frames. ], tot_loss[loss=0.1272, simple_loss=0.2086, pruned_loss=0.02287, over 1442599.60 frames. ], batch size: 70, lr: 3.66e-03, grad_scale: 8.0 +2023-03-21 11:55:59,915 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 11:56:16,526 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 11:56:17,522 INFO [train.py:901] (0/2) Epoch 45, batch 1700, loss[loss=0.117, simple_loss=0.1991, pruned_loss=0.01741, over 7307.00 frames. ], tot_loss[loss=0.1269, simple_loss=0.2082, pruned_loss=0.02278, over 1441964.37 frames. ], batch size: 49, lr: 3.66e-03, grad_scale: 8.0 +2023-03-21 11:56:18,616 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=125959.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:56:18,718 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5860, 3.8525, 2.6144, 4.1739, 3.4003, 3.8056, 1.9289, 2.6875], + device='cuda:0'), covar=tensor([0.0488, 0.0748, 0.3109, 0.0473, 0.0435, 0.0776, 0.4118, 0.1880], + device='cuda:0'), in_proj_covar=tensor([0.0276, 0.0257, 0.0275, 0.0268, 0.0266, 0.0262, 0.0229, 0.0257], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:56:20,110 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.100e+02 1.725e+02 1.970e+02 2.452e+02 4.185e+02, threshold=3.941e+02, percent-clipped=1.0 +2023-03-21 11:56:20,688 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 11:56:31,424 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 11:56:35,793 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2816, 3.4792, 2.5126, 3.8400, 3.0114, 3.4483, 1.7060, 2.6694], + device='cuda:0'), covar=tensor([0.0485, 0.0757, 0.2652, 0.0559, 0.0514, 0.0737, 0.4113, 0.1889], + device='cuda:0'), in_proj_covar=tensor([0.0276, 0.0256, 0.0274, 0.0267, 0.0265, 0.0262, 0.0228, 0.0256], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 11:56:39,351 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=125998.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:56:44,675 INFO [train.py:901] (0/2) Epoch 45, batch 1750, loss[loss=0.107, simple_loss=0.1905, pruned_loss=0.01168, over 7152.00 frames. ], tot_loss[loss=0.1276, simple_loss=0.2089, pruned_loss=0.02318, over 1442461.30 frames. ], batch size: 41, lr: 3.66e-03, grad_scale: 8.0 +2023-03-21 11:56:46,343 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6717, 3.7074, 2.4490, 3.3384, 2.5556, 2.1198, 1.6923, 3.7508], + device='cuda:0'), covar=tensor([0.0066, 0.0070, 0.0303, 0.0107, 0.0276, 0.0723, 0.0816, 0.0070], + device='cuda:0'), in_proj_covar=tensor([0.0121, 0.0097, 0.0119, 0.0102, 0.0137, 0.0138, 0.0131, 0.0108], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 11:56:47,313 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=126012.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 11:56:55,898 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 11:56:56,924 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 11:56:57,512 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=126032.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:57:10,369 INFO [train.py:901] (0/2) Epoch 45, batch 1800, loss[loss=0.1301, simple_loss=0.2177, pruned_loss=0.02129, over 7249.00 frames. ], tot_loss[loss=0.1276, simple_loss=0.2088, pruned_loss=0.02322, over 1443045.91 frames. ], batch size: 89, lr: 3.66e-03, grad_scale: 8.0 +2023-03-21 11:57:12,162 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=126059.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:57:12,585 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=126060.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:57:13,514 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.242e+02 1.649e+02 1.912e+02 2.281e+02 4.540e+02, threshold=3.824e+02, percent-clipped=2.0 +2023-03-21 11:57:19,162 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 11:57:21,342 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=126077.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:57:22,825 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=126080.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:57:23,353 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=126081.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:57:34,594 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 11:57:37,138 INFO [train.py:901] (0/2) Epoch 45, batch 1850, loss[loss=0.1236, simple_loss=0.1965, pruned_loss=0.02538, over 7207.00 frames. ], tot_loss[loss=0.1273, simple_loss=0.2087, pruned_loss=0.02297, over 1444289.47 frames. ], batch size: 45, lr: 3.66e-03, grad_scale: 8.0 +2023-03-21 11:57:38,309 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2638, 3.1234, 3.2217, 3.1917, 3.4280, 3.1310, 2.8054, 3.2177], + device='cuda:0'), covar=tensor([0.1348, 0.0652, 0.1591, 0.1179, 0.0708, 0.1023, 0.1750, 0.2446], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0070, 0.0053, 0.0052, 0.0051, 0.0050, 0.0070, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 11:57:44,257 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 11:57:46,343 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=126125.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:57:55,103 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9257, 3.2044, 3.8663, 3.8653, 3.9595, 3.8983, 4.0048, 3.8726], + device='cuda:0'), covar=tensor([0.0035, 0.0128, 0.0037, 0.0036, 0.0033, 0.0035, 0.0038, 0.0051], + device='cuda:0'), in_proj_covar=tensor([0.0052, 0.0074, 0.0061, 0.0060, 0.0057, 0.0063, 0.0051, 0.0083], + device='cuda:0'), out_proj_covar=tensor([8.6937e-05, 1.4741e-04, 1.0757e-04, 1.0090e-04, 9.4926e-05, 1.0678e-04, + 9.5045e-05, 1.5013e-04], device='cuda:0') +2023-03-21 11:58:00,761 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 11:58:03,208 INFO [train.py:901] (0/2) Epoch 45, batch 1900, loss[loss=0.1251, simple_loss=0.2092, pruned_loss=0.02048, over 7278.00 frames. ], tot_loss[loss=0.1272, simple_loss=0.2087, pruned_loss=0.0229, over 1441454.58 frames. ], batch size: 52, lr: 3.66e-03, grad_scale: 8.0 +2023-03-21 11:58:05,650 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.195e+02 1.674e+02 1.922e+02 2.434e+02 3.525e+02, threshold=3.845e+02, percent-clipped=0.0 +2023-03-21 11:58:13,561 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6984, 5.2071, 5.3267, 5.2402, 5.0727, 4.8205, 5.3457, 5.1807], + device='cuda:0'), covar=tensor([0.0473, 0.0338, 0.0339, 0.0434, 0.0313, 0.0365, 0.0271, 0.0363], + device='cuda:0'), in_proj_covar=tensor([0.0220, 0.0268, 0.0207, 0.0205, 0.0159, 0.0233, 0.0215, 0.0151], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 11:58:25,175 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 11:58:29,525 INFO [train.py:901] (0/2) Epoch 45, batch 1950, loss[loss=0.1434, simple_loss=0.2189, pruned_loss=0.03391, over 7299.00 frames. ], tot_loss[loss=0.1278, simple_loss=0.2095, pruned_loss=0.0231, over 1441764.70 frames. ], batch size: 68, lr: 3.66e-03, grad_scale: 8.0 +2023-03-21 11:58:36,575 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 11:58:41,780 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 11:58:42,794 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 11:58:56,037 INFO [train.py:901] (0/2) Epoch 45, batch 2000, loss[loss=0.1026, simple_loss=0.1826, pruned_loss=0.01127, over 7184.00 frames. ], tot_loss[loss=0.1279, simple_loss=0.2096, pruned_loss=0.02313, over 1442045.55 frames. ], batch size: 39, lr: 3.66e-03, grad_scale: 8.0 +2023-03-21 11:58:57,167 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=126259.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:58:57,197 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7353, 1.4283, 1.9639, 2.0848, 2.0490, 2.2251, 1.7246, 2.1878], + device='cuda:0'), covar=tensor([0.2438, 0.3390, 0.2121, 0.0911, 0.1353, 0.1264, 0.1565, 0.1736], + device='cuda:0'), in_proj_covar=tensor([0.0083, 0.0082, 0.0076, 0.0069, 0.0069, 0.0066, 0.0108, 0.0069], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 11:58:58,548 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.249e+02 1.755e+02 1.955e+02 2.409e+02 3.875e+02, threshold=3.909e+02, percent-clipped=1.0 +2023-03-21 11:58:58,598 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 11:59:09,355 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 11:59:17,183 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 11:59:21,731 INFO [train.py:901] (0/2) Epoch 45, batch 2050, loss[loss=0.1325, simple_loss=0.2203, pruned_loss=0.02234, over 7208.00 frames. ], tot_loss[loss=0.1276, simple_loss=0.2093, pruned_loss=0.02299, over 1440914.02 frames. ], batch size: 93, lr: 3.66e-03, grad_scale: 8.0 +2023-03-21 11:59:21,785 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=126307.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:59:42,928 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.16 vs. limit=2.0 +2023-03-21 11:59:46,769 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=126354.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 11:59:48,241 INFO [train.py:901] (0/2) Epoch 45, batch 2100, loss[loss=0.1332, simple_loss=0.2112, pruned_loss=0.02763, over 7233.00 frames. ], tot_loss[loss=0.1281, simple_loss=0.2098, pruned_loss=0.02319, over 1440962.83 frames. ], batch size: 55, lr: 3.65e-03, grad_scale: 8.0 +2023-03-21 11:59:50,655 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.203e+02 1.670e+02 2.047e+02 2.347e+02 4.553e+02, threshold=4.095e+02, percent-clipped=3.0 +2023-03-21 11:59:51,193 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 11:59:54,703 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 11:59:59,974 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=126380.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:00:00,470 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=126381.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:00:14,577 INFO [train.py:901] (0/2) Epoch 45, batch 2150, loss[loss=0.1213, simple_loss=0.2078, pruned_loss=0.01741, over 7305.00 frames. ], tot_loss[loss=0.1281, simple_loss=0.2097, pruned_loss=0.02318, over 1441511.65 frames. ], batch size: 83, lr: 3.65e-03, grad_scale: 8.0 +2023-03-21 12:00:22,508 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.85 vs. limit=2.0 +2023-03-21 12:00:26,288 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=126429.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:00:32,507 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=126441.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:00:37,012 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=126450.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:00:40,385 INFO [train.py:901] (0/2) Epoch 45, batch 2200, loss[loss=0.1445, simple_loss=0.2224, pruned_loss=0.03331, over 7288.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2085, pruned_loss=0.02284, over 1439705.30 frames. ], batch size: 57, lr: 3.65e-03, grad_scale: 8.0 +2023-03-21 12:00:41,926 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 12:00:42,886 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 9.798e+01 1.679e+02 2.078e+02 2.432e+02 4.559e+02, threshold=4.157e+02, percent-clipped=2.0 +2023-03-21 12:00:45,531 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6300, 3.7677, 3.5836, 3.7103, 3.5110, 3.7405, 4.0856, 4.0946], + device='cuda:0'), covar=tensor([0.0273, 0.0193, 0.0268, 0.0220, 0.0464, 0.0433, 0.0246, 0.0205], + device='cuda:0'), in_proj_covar=tensor([0.0128, 0.0130, 0.0124, 0.0129, 0.0118, 0.0104, 0.0100, 0.0105], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 12:01:06,381 INFO [train.py:901] (0/2) Epoch 45, batch 2250, loss[loss=0.1458, simple_loss=0.2223, pruned_loss=0.03462, over 7261.00 frames. ], tot_loss[loss=0.1273, simple_loss=0.2084, pruned_loss=0.02306, over 1438388.19 frames. ], batch size: 47, lr: 3.65e-03, grad_scale: 8.0 +2023-03-21 12:01:08,550 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=126511.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:01:16,703 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 12:01:17,220 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 12:01:26,160 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.90 vs. limit=2.0 +2023-03-21 12:01:29,469 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 12:01:32,492 INFO [train.py:901] (0/2) Epoch 45, batch 2300, loss[loss=0.1318, simple_loss=0.2119, pruned_loss=0.02584, over 7312.00 frames. ], tot_loss[loss=0.1274, simple_loss=0.2087, pruned_loss=0.02298, over 1438999.93 frames. ], batch size: 83, lr: 3.65e-03, grad_scale: 8.0 +2023-03-21 12:01:34,907 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.261e+02 1.701e+02 2.024e+02 2.404e+02 3.706e+02, threshold=4.049e+02, percent-clipped=0.0 +2023-03-21 12:01:51,323 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7222, 1.4066, 1.7752, 2.0737, 1.9804, 2.1553, 1.5397, 2.0987], + device='cuda:0'), covar=tensor([0.2625, 0.2923, 0.1273, 0.1520, 0.1234, 0.1556, 0.2013, 0.2096], + device='cuda:0'), in_proj_covar=tensor([0.0082, 0.0081, 0.0075, 0.0068, 0.0068, 0.0067, 0.0107, 0.0068], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 12:01:59,156 INFO [train.py:901] (0/2) Epoch 45, batch 2350, loss[loss=0.127, simple_loss=0.2013, pruned_loss=0.02633, over 7297.00 frames. ], tot_loss[loss=0.1276, simple_loss=0.2091, pruned_loss=0.02307, over 1439858.65 frames. ], batch size: 49, lr: 3.65e-03, grad_scale: 8.0 +2023-03-21 12:02:16,253 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 12:02:22,844 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 12:02:22,927 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=126654.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:02:24,366 INFO [train.py:901] (0/2) Epoch 45, batch 2400, loss[loss=0.1091, simple_loss=0.1908, pruned_loss=0.01372, over 7160.00 frames. ], tot_loss[loss=0.128, simple_loss=0.2094, pruned_loss=0.02333, over 1439144.72 frames. ], batch size: 41, lr: 3.65e-03, grad_scale: 8.0 +2023-03-21 12:02:24,504 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3558, 2.2920, 2.5865, 2.3002, 2.4680, 2.2383, 2.3155, 1.8133], + device='cuda:0'), covar=tensor([0.0433, 0.0413, 0.0328, 0.0288, 0.0481, 0.0616, 0.0373, 0.0408], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0042, 0.0042, 0.0042, 0.0039, 0.0039, 0.0045, 0.0045], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 12:02:27,507 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.239e+02 1.819e+02 2.131e+02 2.403e+02 3.692e+02, threshold=4.262e+02, percent-clipped=0.0 +2023-03-21 12:02:34,164 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 12:02:37,165 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 12:02:48,447 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=126702.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:02:50,913 INFO [train.py:901] (0/2) Epoch 45, batch 2450, loss[loss=0.1292, simple_loss=0.2103, pruned_loss=0.02407, over 7300.00 frames. ], tot_loss[loss=0.1277, simple_loss=0.209, pruned_loss=0.02317, over 1438333.21 frames. ], batch size: 80, lr: 3.65e-03, grad_scale: 8.0 +2023-03-21 12:02:54,511 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8747, 4.1148, 3.8835, 4.0056, 3.6912, 4.0495, 4.4073, 4.3977], + device='cuda:0'), covar=tensor([0.0264, 0.0169, 0.0243, 0.0222, 0.0374, 0.0413, 0.0244, 0.0205], + device='cuda:0'), in_proj_covar=tensor([0.0128, 0.0130, 0.0124, 0.0130, 0.0119, 0.0105, 0.0101, 0.0106], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 12:03:03,204 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 12:03:05,857 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=126736.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:03:08,382 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.0073, 4.4939, 4.3027, 4.9671, 4.8176, 4.8683, 4.3858, 4.6227], + device='cuda:0'), covar=tensor([0.0906, 0.2422, 0.2414, 0.1014, 0.0957, 0.1211, 0.0831, 0.1047], + device='cuda:0'), in_proj_covar=tensor([0.0272, 0.0404, 0.0306, 0.0322, 0.0242, 0.0378, 0.0239, 0.0286], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 12:03:10,543 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3504, 3.5613, 2.5781, 3.9036, 2.8411, 3.4219, 1.7982, 2.8114], + device='cuda:0'), covar=tensor([0.0463, 0.0751, 0.2875, 0.0496, 0.0570, 0.0522, 0.4098, 0.1877], + device='cuda:0'), in_proj_covar=tensor([0.0276, 0.0255, 0.0276, 0.0267, 0.0266, 0.0261, 0.0227, 0.0257], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 12:03:17,010 INFO [train.py:901] (0/2) Epoch 45, batch 2500, loss[loss=0.1201, simple_loss=0.1965, pruned_loss=0.02181, over 7314.00 frames. ], tot_loss[loss=0.1273, simple_loss=0.2087, pruned_loss=0.02299, over 1440405.92 frames. ], batch size: 49, lr: 3.65e-03, grad_scale: 8.0 +2023-03-21 12:03:19,525 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.195e+02 1.802e+02 2.146e+02 2.550e+02 5.501e+02, threshold=4.291e+02, percent-clipped=1.0 +2023-03-21 12:03:26,711 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2983, 2.4536, 2.6291, 2.3706, 2.5813, 2.4954, 2.5025, 1.9274], + device='cuda:0'), covar=tensor([0.0536, 0.0502, 0.0437, 0.0288, 0.0430, 0.0667, 0.0318, 0.0399], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0042, 0.0042, 0.0041, 0.0038, 0.0039, 0.0045, 0.0045], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 12:03:29,296 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 12:03:42,785 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=126806.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:03:43,243 INFO [train.py:901] (0/2) Epoch 45, batch 2550, loss[loss=0.1069, simple_loss=0.1796, pruned_loss=0.01715, over 6933.00 frames. ], tot_loss[loss=0.1275, simple_loss=0.209, pruned_loss=0.023, over 1440525.11 frames. ], batch size: 35, lr: 3.65e-03, grad_scale: 8.0 +2023-03-21 12:03:44,924 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9594, 3.1468, 2.1615, 3.5117, 2.3730, 3.1020, 1.4604, 2.3345], + device='cuda:0'), covar=tensor([0.0613, 0.1170, 0.3364, 0.0710, 0.0562, 0.0559, 0.4419, 0.2100], + device='cuda:0'), in_proj_covar=tensor([0.0276, 0.0256, 0.0276, 0.0268, 0.0266, 0.0262, 0.0228, 0.0257], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 12:03:47,945 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=126816.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:04:07,702 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6429, 3.6973, 2.7878, 4.0929, 3.1827, 3.6391, 1.8266, 2.8835], + device='cuda:0'), covar=tensor([0.0532, 0.0697, 0.2758, 0.0685, 0.0486, 0.0726, 0.4334, 0.2094], + device='cuda:0'), in_proj_covar=tensor([0.0278, 0.0257, 0.0277, 0.0269, 0.0267, 0.0263, 0.0229, 0.0258], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 12:04:09,082 INFO [train.py:901] (0/2) Epoch 45, batch 2600, loss[loss=0.1203, simple_loss=0.2021, pruned_loss=0.01926, over 7340.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2085, pruned_loss=0.02288, over 1440429.45 frames. ], batch size: 75, lr: 3.65e-03, grad_scale: 8.0 +2023-03-21 12:04:09,717 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4382, 2.5532, 2.7103, 2.3019, 2.4731, 2.4045, 2.3985, 1.9312], + device='cuda:0'), covar=tensor([0.0602, 0.0369, 0.0336, 0.0327, 0.0677, 0.0657, 0.0332, 0.0366], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0041, 0.0042, 0.0041, 0.0038, 0.0039, 0.0044, 0.0044], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 12:04:11,549 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.143e+02 1.684e+02 1.885e+02 2.330e+02 6.166e+02, threshold=3.769e+02, percent-clipped=1.0 +2023-03-21 12:04:19,292 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=126877.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:04:27,689 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4172, 2.4002, 2.5574, 2.2698, 2.6451, 2.4523, 2.3654, 1.9257], + device='cuda:0'), covar=tensor([0.0482, 0.0466, 0.0296, 0.0349, 0.0511, 0.0509, 0.0323, 0.0342], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0041, 0.0041, 0.0041, 0.0038, 0.0038, 0.0044, 0.0044], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 12:04:33,299 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.32 vs. limit=5.0 +2023-03-21 12:04:34,046 INFO [train.py:901] (0/2) Epoch 45, batch 2650, loss[loss=0.1286, simple_loss=0.2122, pruned_loss=0.02254, over 7265.00 frames. ], tot_loss[loss=0.1272, simple_loss=0.2084, pruned_loss=0.02296, over 1440769.16 frames. ], batch size: 64, lr: 3.65e-03, grad_scale: 8.0 +2023-03-21 12:04:59,423 INFO [train.py:901] (0/2) Epoch 45, batch 2700, loss[loss=0.1327, simple_loss=0.2177, pruned_loss=0.02386, over 7335.00 frames. ], tot_loss[loss=0.1272, simple_loss=0.2084, pruned_loss=0.02298, over 1441651.03 frames. ], batch size: 61, lr: 3.65e-03, grad_scale: 8.0 +2023-03-21 12:05:01,894 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.267e+02 1.714e+02 2.002e+02 2.426e+02 5.440e+02, threshold=4.003e+02, percent-clipped=3.0 +2023-03-21 12:05:24,669 INFO [train.py:901] (0/2) Epoch 45, batch 2750, loss[loss=0.1411, simple_loss=0.2219, pruned_loss=0.03015, over 7258.00 frames. ], tot_loss[loss=0.127, simple_loss=0.2085, pruned_loss=0.02278, over 1441501.66 frames. ], batch size: 89, lr: 3.65e-03, grad_scale: 8.0 +2023-03-21 12:05:38,867 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=127036.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:05:42,760 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4147, 3.9962, 4.0537, 4.0943, 4.0354, 4.0000, 4.3388, 3.8470], + device='cuda:0'), covar=tensor([0.0135, 0.0183, 0.0118, 0.0152, 0.0462, 0.0114, 0.0122, 0.0171], + device='cuda:0'), in_proj_covar=tensor([0.0100, 0.0107, 0.0109, 0.0093, 0.0185, 0.0114, 0.0111, 0.0119], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 12:05:49,143 INFO [train.py:901] (0/2) Epoch 45, batch 2800, loss[loss=0.1344, simple_loss=0.2156, pruned_loss=0.02663, over 7284.00 frames. ], tot_loss[loss=0.127, simple_loss=0.2082, pruned_loss=0.02287, over 1443668.92 frames. ], batch size: 52, lr: 3.64e-03, grad_scale: 8.0 +2023-03-21 12:05:52,005 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.152e+02 1.659e+02 1.865e+02 2.207e+02 4.110e+02, threshold=3.730e+02, percent-clipped=1.0 +2023-03-21 12:06:02,182 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-45.pt +2023-03-21 12:06:17,563 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 12:06:17,616 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 12:06:17,648 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 12:06:17,665 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 12:06:17,841 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 12:06:17,857 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 12:06:17,868 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 12:06:17,968 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 12:06:18,422 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 12:06:18,438 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 12:06:18,454 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 12:06:18,618 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 12:06:18,624 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 12:06:18,631 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 12:06:18,775 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 12:06:18,778 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 12:06:18,837 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 12:06:21,169 INFO [train.py:901] (0/2) Epoch 46, batch 0, loss[loss=0.1276, simple_loss=0.216, pruned_loss=0.01956, over 7327.00 frames. ], tot_loss[loss=0.1276, simple_loss=0.216, pruned_loss=0.01956, over 7327.00 frames. ], batch size: 75, lr: 3.60e-03, grad_scale: 8.0 +2023-03-21 12:06:21,171 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 12:06:38,012 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0969, 3.7636, 3.7344, 3.7371, 3.8177, 3.6626, 3.9807, 3.5335], + device='cuda:0'), covar=tensor([0.0116, 0.0183, 0.0141, 0.0163, 0.0406, 0.0123, 0.0147, 0.0202], + device='cuda:0'), in_proj_covar=tensor([0.0100, 0.0107, 0.0108, 0.0092, 0.0184, 0.0114, 0.0110, 0.0119], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 12:06:39,526 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4229, 2.2659, 2.5589, 2.3555, 2.5906, 2.4284, 2.3124, 1.8122], + device='cuda:0'), covar=tensor([0.0358, 0.0456, 0.0278, 0.0273, 0.0288, 0.0450, 0.0397, 0.0371], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0041, 0.0042, 0.0041, 0.0038, 0.0038, 0.0044, 0.0044], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 12:06:47,485 INFO [train.py:935] (0/2) Epoch 46, validation: loss=0.1654, simple_loss=0.2576, pruned_loss=0.03655, over 1622729.00 frames. +2023-03-21 12:06:47,486 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 12:06:49,033 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=127084.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:06:52,605 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0776, 4.5056, 4.6112, 4.5329, 4.5201, 4.0943, 4.6301, 4.4535], + device='cuda:0'), covar=tensor([0.0535, 0.0509, 0.0392, 0.0535, 0.0392, 0.0503, 0.0375, 0.0482], + device='cuda:0'), in_proj_covar=tensor([0.0227, 0.0277, 0.0213, 0.0211, 0.0165, 0.0240, 0.0221, 0.0156], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 12:06:53,529 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 12:07:00,269 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=127106.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:07:04,710 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 12:07:11,209 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 12:07:12,657 INFO [train.py:901] (0/2) Epoch 46, batch 50, loss[loss=0.1496, simple_loss=0.2248, pruned_loss=0.03719, over 7277.00 frames. ], tot_loss[loss=0.1286, simple_loss=0.2097, pruned_loss=0.02379, over 325706.94 frames. ], batch size: 57, lr: 3.60e-03, grad_scale: 8.0 +2023-03-21 12:07:13,640 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 12:07:16,167 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 12:07:24,996 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=127154.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:07:29,000 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.146e+02 1.681e+02 1.912e+02 2.297e+02 4.386e+02, threshold=3.823e+02, percent-clipped=1.0 +2023-03-21 12:07:34,612 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 12:07:34,624 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 12:07:34,678 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=127172.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:07:39,145 INFO [train.py:901] (0/2) Epoch 46, batch 100, loss[loss=0.1202, simple_loss=0.2051, pruned_loss=0.01764, over 7329.00 frames. ], tot_loss[loss=0.129, simple_loss=0.2108, pruned_loss=0.0236, over 572393.32 frames. ], batch size: 75, lr: 3.60e-03, grad_scale: 8.0 +2023-03-21 12:07:41,225 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4859, 2.1652, 1.6594, 1.8503, 2.0205, 1.7948, 1.7827, 1.2352], + device='cuda:0'), covar=tensor([0.0301, 0.0138, 0.0269, 0.0264, 0.0132, 0.0162, 0.0275, 0.0314], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0040, 0.0038, 0.0040, 0.0038, 0.0037, 0.0039, 0.0048], + device='cuda:0'), out_proj_covar=tensor([4.6188e-05, 4.4664e-05, 4.2799e-05, 4.4241e-05, 4.2467e-05, 4.0535e-05, + 4.3700e-05, 5.3050e-05], device='cuda:0') +2023-03-21 12:07:49,445 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.83 vs. limit=5.0 +2023-03-21 12:08:03,147 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9912, 3.1591, 3.5040, 3.2641, 3.4346, 3.2536, 2.8095, 3.3563], + device='cuda:0'), covar=tensor([0.1616, 0.0631, 0.0805, 0.1278, 0.0697, 0.0808, 0.1934, 0.1189], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0071, 0.0052, 0.0052, 0.0052, 0.0050, 0.0069, 0.0052], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 12:08:04,558 INFO [train.py:901] (0/2) Epoch 46, batch 150, loss[loss=0.1198, simple_loss=0.2036, pruned_loss=0.01798, over 7369.00 frames. ], tot_loss[loss=0.1272, simple_loss=0.209, pruned_loss=0.02271, over 766297.03 frames. ], batch size: 44, lr: 3.60e-03, grad_scale: 8.0 +2023-03-21 12:08:14,365 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=127249.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:08:21,345 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.127e+02 1.712e+02 2.119e+02 2.555e+02 4.679e+02, threshold=4.238e+02, percent-clipped=2.0 +2023-03-21 12:08:30,978 INFO [train.py:901] (0/2) Epoch 46, batch 200, loss[loss=0.1293, simple_loss=0.2048, pruned_loss=0.02686, over 7228.00 frames. ], tot_loss[loss=0.1268, simple_loss=0.2081, pruned_loss=0.02268, over 916290.87 frames. ], batch size: 45, lr: 3.60e-03, grad_scale: 8.0 +2023-03-21 12:08:36,550 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 12:08:41,648 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 12:08:45,910 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=127310.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:08:47,947 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 12:08:56,983 INFO [train.py:901] (0/2) Epoch 46, batch 250, loss[loss=0.1348, simple_loss=0.2143, pruned_loss=0.02769, over 7286.00 frames. ], tot_loss[loss=0.1264, simple_loss=0.208, pruned_loss=0.02235, over 1033618.20 frames. ], batch size: 57, lr: 3.60e-03, grad_scale: 8.0 +2023-03-21 12:09:00,525 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 12:09:13,225 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.252e+02 1.735e+02 1.962e+02 2.246e+02 8.468e+02, threshold=3.924e+02, percent-clipped=3.0 +2023-03-21 12:09:21,341 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 12:09:22,820 INFO [train.py:901] (0/2) Epoch 46, batch 300, loss[loss=0.1179, simple_loss=0.2007, pruned_loss=0.01756, over 7332.00 frames. ], tot_loss[loss=0.126, simple_loss=0.2073, pruned_loss=0.02233, over 1124439.66 frames. ], batch size: 75, lr: 3.60e-03, grad_scale: 8.0 +2023-03-21 12:09:30,985 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 12:09:32,121 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1249, 4.0541, 3.2135, 3.7482, 3.1951, 2.2085, 1.7913, 4.1466], + device='cuda:0'), covar=tensor([0.0069, 0.0065, 0.0251, 0.0087, 0.0235, 0.0809, 0.0785, 0.0075], + device='cuda:0'), in_proj_covar=tensor([0.0119, 0.0096, 0.0118, 0.0101, 0.0135, 0.0138, 0.0130, 0.0108], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 12:09:47,396 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=127427.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 12:09:49,292 INFO [train.py:901] (0/2) Epoch 46, batch 350, loss[loss=0.1149, simple_loss=0.1867, pruned_loss=0.02154, over 7050.00 frames. ], tot_loss[loss=0.1263, simple_loss=0.2073, pruned_loss=0.02263, over 1194194.90 frames. ], batch size: 35, lr: 3.60e-03, grad_scale: 8.0 +2023-03-21 12:10:05,546 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.180e+02 1.753e+02 1.971e+02 2.330e+02 3.656e+02, threshold=3.942e+02, percent-clipped=0.0 +2023-03-21 12:10:07,612 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 12:10:10,645 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=127472.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:10:15,187 INFO [train.py:901] (0/2) Epoch 46, batch 400, loss[loss=0.1348, simple_loss=0.2175, pruned_loss=0.02602, over 7272.00 frames. ], tot_loss[loss=0.126, simple_loss=0.2071, pruned_loss=0.02243, over 1248719.15 frames. ], batch size: 57, lr: 3.60e-03, grad_scale: 8.0 +2023-03-21 12:10:18,884 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=127488.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 12:10:29,740 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8011, 1.6791, 2.3490, 2.2895, 2.2147, 2.4699, 2.1423, 2.4451], + device='cuda:0'), covar=tensor([0.4039, 0.3687, 0.1873, 0.1296, 0.1684, 0.1455, 0.1857, 0.2971], + device='cuda:0'), in_proj_covar=tensor([0.0082, 0.0081, 0.0076, 0.0068, 0.0068, 0.0067, 0.0107, 0.0070], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 12:10:35,726 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=127520.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:10:41,015 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=127529.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:10:41,892 INFO [train.py:901] (0/2) Epoch 46, batch 450, loss[loss=0.1353, simple_loss=0.2166, pruned_loss=0.02702, over 7290.00 frames. ], tot_loss[loss=0.1261, simple_loss=0.2071, pruned_loss=0.02256, over 1291215.78 frames. ], batch size: 66, lr: 3.60e-03, grad_scale: 8.0 +2023-03-21 12:10:51,131 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 12:10:51,145 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 12:10:57,668 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.232e+02 1.678e+02 2.040e+02 2.414e+02 3.760e+02, threshold=4.080e+02, percent-clipped=0.0 +2023-03-21 12:11:04,213 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-03-21 12:11:07,476 INFO [train.py:901] (0/2) Epoch 46, batch 500, loss[loss=0.1208, simple_loss=0.2088, pruned_loss=0.01642, over 7252.00 frames. ], tot_loss[loss=0.1263, simple_loss=0.2075, pruned_loss=0.02255, over 1324639.03 frames. ], batch size: 89, lr: 3.60e-03, grad_scale: 8.0 +2023-03-21 12:11:12,659 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=127590.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:11:20,649 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=127605.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:11:23,249 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 12:11:24,781 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 12:11:25,282 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 12:11:27,344 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 12:11:32,372 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 12:11:34,344 INFO [train.py:901] (0/2) Epoch 46, batch 550, loss[loss=0.1282, simple_loss=0.2109, pruned_loss=0.02276, over 7162.00 frames. ], tot_loss[loss=0.1265, simple_loss=0.2079, pruned_loss=0.02249, over 1351292.48 frames. ], batch size: 41, lr: 3.60e-03, grad_scale: 8.0 +2023-03-21 12:11:43,516 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 12:11:50,401 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=127662.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:11:50,750 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.225e+02 1.755e+02 2.100e+02 2.513e+02 5.423e+02, threshold=4.201e+02, percent-clipped=2.0 +2023-03-21 12:11:50,816 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 12:11:54,905 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 12:12:00,439 INFO [train.py:901] (0/2) Epoch 46, batch 600, loss[loss=0.1441, simple_loss=0.225, pruned_loss=0.03153, over 7224.00 frames. ], tot_loss[loss=0.1263, simple_loss=0.2081, pruned_loss=0.02231, over 1369830.12 frames. ], batch size: 93, lr: 3.60e-03, grad_scale: 8.0 +2023-03-21 12:12:01,466 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 12:12:15,083 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0511, 3.3355, 2.3635, 3.6215, 2.7701, 3.2041, 1.6230, 2.5417], + device='cuda:0'), covar=tensor([0.0454, 0.0773, 0.2840, 0.0707, 0.0466, 0.0786, 0.4207, 0.1847], + device='cuda:0'), in_proj_covar=tensor([0.0273, 0.0254, 0.0274, 0.0266, 0.0264, 0.0260, 0.0227, 0.0254], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 12:12:17,954 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 12:12:22,082 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=127723.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:12:26,002 INFO [train.py:901] (0/2) Epoch 46, batch 650, loss[loss=0.1432, simple_loss=0.216, pruned_loss=0.03522, over 7338.00 frames. ], tot_loss[loss=0.1269, simple_loss=0.2086, pruned_loss=0.02263, over 1388528.14 frames. ], batch size: 54, lr: 3.59e-03, grad_scale: 8.0 +2023-03-21 12:12:26,517 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 12:12:38,632 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.2695, 4.7570, 4.5869, 5.2172, 5.0045, 5.1421, 4.5964, 4.8553], + device='cuda:0'), covar=tensor([0.0826, 0.2584, 0.2275, 0.0838, 0.0948, 0.1168, 0.0745, 0.1094], + device='cuda:0'), in_proj_covar=tensor([0.0271, 0.0403, 0.0304, 0.0321, 0.0240, 0.0376, 0.0237, 0.0286], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 12:12:42,527 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.153e+02 1.739e+02 1.949e+02 2.309e+02 6.874e+02, threshold=3.898e+02, percent-clipped=2.0 +2023-03-21 12:12:43,590 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 12:12:51,292 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=127779.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:12:52,164 INFO [train.py:901] (0/2) Epoch 46, batch 700, loss[loss=0.1284, simple_loss=0.211, pruned_loss=0.02294, over 7284.00 frames. ], tot_loss[loss=0.127, simple_loss=0.2086, pruned_loss=0.02274, over 1397645.75 frames. ], batch size: 77, lr: 3.59e-03, grad_scale: 8.0 +2023-03-21 12:12:52,723 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 12:12:53,262 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=127783.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 12:12:56,899 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1770, 2.5377, 1.9030, 3.2077, 3.0294, 3.1158, 2.8190, 2.7978], + device='cuda:0'), covar=tensor([0.2638, 0.1254, 0.4417, 0.0856, 0.0430, 0.0388, 0.0573, 0.0562], + device='cuda:0'), in_proj_covar=tensor([0.0247, 0.0230, 0.0244, 0.0258, 0.0202, 0.0204, 0.0219, 0.0228], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 12:13:16,752 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 12:13:17,241 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 12:13:17,726 INFO [train.py:901] (0/2) Epoch 46, batch 750, loss[loss=0.1285, simple_loss=0.2098, pruned_loss=0.02356, over 7232.00 frames. ], tot_loss[loss=0.1268, simple_loss=0.2084, pruned_loss=0.02259, over 1409560.16 frames. ], batch size: 50, lr: 3.59e-03, grad_scale: 8.0 +2023-03-21 12:13:23,060 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=127840.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:13:28,175 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=127850.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:13:29,873 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.18 vs. limit=5.0 +2023-03-21 12:13:31,637 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 12:13:34,689 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.337e+02 1.689e+02 1.958e+02 2.466e+02 6.234e+02, threshold=3.916e+02, percent-clipped=1.0 +2023-03-21 12:13:36,875 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 12:13:39,925 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=127872.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:13:42,843 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 12:13:43,128 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.57 vs. limit=2.0 +2023-03-21 12:13:43,848 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 12:13:44,326 INFO [train.py:901] (0/2) Epoch 46, batch 800, loss[loss=0.1455, simple_loss=0.2286, pruned_loss=0.03122, over 7226.00 frames. ], tot_loss[loss=0.1272, simple_loss=0.2088, pruned_loss=0.02286, over 1414278.77 frames. ], batch size: 93, lr: 3.59e-03, grad_scale: 8.0 +2023-03-21 12:13:46,487 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=127885.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:13:47,259 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.47 vs. limit=2.0 +2023-03-21 12:13:54,456 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 12:13:56,525 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=127905.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:13:59,501 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=127911.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:14:09,945 INFO [train.py:901] (0/2) Epoch 46, batch 850, loss[loss=0.1327, simple_loss=0.215, pruned_loss=0.02522, over 7335.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2085, pruned_loss=0.02284, over 1420293.20 frames. ], batch size: 54, lr: 3.59e-03, grad_scale: 8.0 +2023-03-21 12:14:11,142 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=127933.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:14:13,062 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 12:14:13,537 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 12:14:18,154 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 12:14:21,933 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=127953.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:14:22,883 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 12:14:26,863 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.179e+02 1.746e+02 1.944e+02 2.264e+02 3.850e+02, threshold=3.887e+02, percent-clipped=0.0 +2023-03-21 12:14:34,237 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.61 vs. limit=2.0 +2023-03-21 12:14:35,893 INFO [train.py:901] (0/2) Epoch 46, batch 900, loss[loss=0.12, simple_loss=0.2034, pruned_loss=0.01825, over 7278.00 frames. ], tot_loss[loss=0.1266, simple_loss=0.2081, pruned_loss=0.02254, over 1426468.91 frames. ], batch size: 77, lr: 3.59e-03, grad_scale: 8.0 +2023-03-21 12:14:46,004 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-128000.pt +2023-03-21 12:14:59,362 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=128018.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:15:03,376 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 12:15:06,545 INFO [train.py:901] (0/2) Epoch 46, batch 950, loss[loss=0.1176, simple_loss=0.2037, pruned_loss=0.01579, over 7278.00 frames. ], tot_loss[loss=0.1262, simple_loss=0.2077, pruned_loss=0.02237, over 1429712.13 frames. ], batch size: 70, lr: 3.59e-03, grad_scale: 8.0 +2023-03-21 12:15:08,581 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7665, 2.3135, 2.8608, 2.8299, 2.8150, 2.6159, 2.3900, 2.6832], + device='cuda:0'), covar=tensor([0.1261, 0.1157, 0.1211, 0.1218, 0.1001, 0.1168, 0.1834, 0.1594], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0071, 0.0053, 0.0052, 0.0052, 0.0051, 0.0070, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 12:15:22,629 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.200e+02 1.644e+02 1.867e+02 2.236e+02 5.162e+02, threshold=3.734e+02, percent-clipped=3.0 +2023-03-21 12:15:26,713 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 12:15:27,374 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1342, 3.3535, 2.3951, 3.7058, 2.7191, 3.2256, 1.6015, 2.5971], + device='cuda:0'), covar=tensor([0.0478, 0.0919, 0.2879, 0.0691, 0.0486, 0.0633, 0.4062, 0.1822], + device='cuda:0'), in_proj_covar=tensor([0.0272, 0.0254, 0.0273, 0.0266, 0.0264, 0.0261, 0.0226, 0.0254], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 12:15:31,752 INFO [train.py:901] (0/2) Epoch 46, batch 1000, loss[loss=0.1318, simple_loss=0.2178, pruned_loss=0.02289, over 7259.00 frames. ], tot_loss[loss=0.1267, simple_loss=0.2084, pruned_loss=0.02251, over 1434725.17 frames. ], batch size: 64, lr: 3.59e-03, grad_scale: 8.0 +2023-03-21 12:15:32,816 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=128083.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 12:15:47,844 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 12:15:58,211 INFO [train.py:901] (0/2) Epoch 46, batch 1050, loss[loss=0.1374, simple_loss=0.219, pruned_loss=0.02792, over 7327.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2087, pruned_loss=0.02272, over 1432866.70 frames. ], batch size: 59, lr: 3.59e-03, grad_scale: 8.0 +2023-03-21 12:15:58,261 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=128131.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 12:16:00,270 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=128135.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:16:08,869 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 12:16:12,338 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 12:16:14,378 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.236e+02 1.762e+02 2.099e+02 2.466e+02 4.105e+02, threshold=4.197e+02, percent-clipped=1.0 +2023-03-21 12:16:24,063 INFO [train.py:901] (0/2) Epoch 46, batch 1100, loss[loss=0.12, simple_loss=0.2137, pruned_loss=0.01318, over 7281.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2087, pruned_loss=0.02273, over 1433552.44 frames. ], batch size: 68, lr: 3.59e-03, grad_scale: 8.0 +2023-03-21 12:16:26,216 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=128185.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:16:30,203 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5772, 1.8003, 1.5649, 1.7580, 1.8760, 1.7348, 1.7804, 1.3138], + device='cuda:0'), covar=tensor([0.0182, 0.0192, 0.0235, 0.0240, 0.0143, 0.0147, 0.0184, 0.0215], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0041, 0.0038, 0.0040, 0.0038, 0.0037, 0.0039, 0.0048], + device='cuda:0'), out_proj_covar=tensor([4.5899e-05, 4.5007e-05, 4.2941e-05, 4.4478e-05, 4.2240e-05, 4.0783e-05, + 4.3587e-05, 5.2787e-05], device='cuda:0') +2023-03-21 12:16:37,716 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=128206.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:16:42,743 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 12:16:43,248 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 12:16:48,817 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=128228.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:16:50,332 INFO [train.py:901] (0/2) Epoch 46, batch 1150, loss[loss=0.1391, simple_loss=0.2181, pruned_loss=0.03005, over 7286.00 frames. ], tot_loss[loss=0.1274, simple_loss=0.209, pruned_loss=0.02288, over 1437500.54 frames. ], batch size: 57, lr: 3.59e-03, grad_scale: 8.0 +2023-03-21 12:16:51,389 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=128233.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:16:56,349 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 12:16:56,845 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 12:17:06,366 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.224e+02 1.776e+02 2.012e+02 2.405e+02 4.369e+02, threshold=4.024e+02, percent-clipped=1.0 +2023-03-21 12:17:16,093 INFO [train.py:901] (0/2) Epoch 46, batch 1200, loss[loss=0.1316, simple_loss=0.2119, pruned_loss=0.02568, over 7352.00 frames. ], tot_loss[loss=0.1272, simple_loss=0.2087, pruned_loss=0.02286, over 1437074.25 frames. ], batch size: 63, lr: 3.59e-03, grad_scale: 8.0 +2023-03-21 12:17:17,287 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2946, 3.5575, 3.0084, 3.3890, 3.3974, 3.0005, 3.3650, 3.1088], + device='cuda:0'), covar=tensor([0.0895, 0.0787, 0.1572, 0.0986, 0.1124, 0.0674, 0.0735, 0.0951], + device='cuda:0'), in_proj_covar=tensor([0.0061, 0.0061, 0.0071, 0.0062, 0.0058, 0.0065, 0.0059, 0.0056], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 12:17:26,007 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1741, 4.4077, 4.1263, 4.3415, 3.9822, 4.4165, 4.6605, 4.6342], + device='cuda:0'), covar=tensor([0.0235, 0.0135, 0.0214, 0.0152, 0.0371, 0.0219, 0.0212, 0.0202], + device='cuda:0'), in_proj_covar=tensor([0.0128, 0.0130, 0.0124, 0.0130, 0.0117, 0.0104, 0.0101, 0.0106], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 12:17:27,099 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2373, 4.4940, 4.2119, 4.4348, 4.0144, 4.4550, 4.7547, 4.7280], + device='cuda:0'), covar=tensor([0.0214, 0.0141, 0.0233, 0.0141, 0.0351, 0.0217, 0.0184, 0.0179], + device='cuda:0'), in_proj_covar=tensor([0.0128, 0.0130, 0.0124, 0.0130, 0.0117, 0.0104, 0.0101, 0.0106], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 12:17:31,505 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 12:17:35,653 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=128318.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:17:36,632 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=128320.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:17:42,001 INFO [train.py:901] (0/2) Epoch 46, batch 1250, loss[loss=0.1295, simple_loss=0.208, pruned_loss=0.0255, over 7320.00 frames. ], tot_loss[loss=0.1269, simple_loss=0.2082, pruned_loss=0.02276, over 1436088.84 frames. ], batch size: 80, lr: 3.59e-03, grad_scale: 8.0 +2023-03-21 12:17:55,478 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 12:17:59,031 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.390e+02 1.796e+02 1.977e+02 2.380e+02 4.062e+02, threshold=3.955e+02, percent-clipped=1.0 +2023-03-21 12:18:00,563 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 12:18:00,607 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=128366.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:18:01,421 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.96 vs. limit=2.0 +2023-03-21 12:18:01,591 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 12:18:08,156 INFO [train.py:901] (0/2) Epoch 46, batch 1300, loss[loss=0.1303, simple_loss=0.2181, pruned_loss=0.02128, over 7213.00 frames. ], tot_loss[loss=0.1267, simple_loss=0.208, pruned_loss=0.02276, over 1435802.08 frames. ], batch size: 93, lr: 3.59e-03, grad_scale: 8.0 +2023-03-21 12:18:08,280 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=128381.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:18:25,343 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=128413.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:18:25,759 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 12:18:27,944 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9082, 2.3656, 2.9976, 2.7510, 2.9306, 2.8016, 2.4051, 2.9235], + device='cuda:0'), covar=tensor([0.1043, 0.0941, 0.0967, 0.1278, 0.0734, 0.0708, 0.2141, 0.0982], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0071, 0.0053, 0.0052, 0.0053, 0.0051, 0.0070, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 12:18:28,348 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 12:18:31,831 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 12:18:34,370 INFO [train.py:901] (0/2) Epoch 46, batch 1350, loss[loss=0.1238, simple_loss=0.2053, pruned_loss=0.02118, over 7344.00 frames. ], tot_loss[loss=0.1261, simple_loss=0.2074, pruned_loss=0.02239, over 1438572.15 frames. ], batch size: 63, lr: 3.59e-03, grad_scale: 8.0 +2023-03-21 12:18:36,557 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=128435.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:18:40,601 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 12:18:51,192 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.207e+02 1.598e+02 1.958e+02 2.374e+02 5.805e+02, threshold=3.916e+02, percent-clipped=3.0 +2023-03-21 12:18:51,322 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7870, 2.8458, 3.8861, 3.6394, 3.9969, 3.8134, 3.9069, 3.5544], + device='cuda:0'), covar=tensor([0.0053, 0.0255, 0.0046, 0.0066, 0.0045, 0.0056, 0.0060, 0.0088], + device='cuda:0'), in_proj_covar=tensor([0.0051, 0.0073, 0.0060, 0.0060, 0.0057, 0.0062, 0.0049, 0.0081], + device='cuda:0'), out_proj_covar=tensor([8.5029e-05, 1.4546e-04, 1.0570e-04, 1.0028e-04, 9.3535e-05, 1.0567e-04, + 9.1870e-05, 1.4715e-04], device='cuda:0') +2023-03-21 12:18:57,562 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=128474.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:19:00,961 INFO [train.py:901] (0/2) Epoch 46, batch 1400, loss[loss=0.1414, simple_loss=0.2166, pruned_loss=0.03308, over 7343.00 frames. ], tot_loss[loss=0.1263, simple_loss=0.2075, pruned_loss=0.02257, over 1440042.68 frames. ], batch size: 51, lr: 3.58e-03, grad_scale: 8.0 +2023-03-21 12:19:02,041 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=128483.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:19:13,827 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=128506.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:19:14,224 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 12:19:21,616 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.65 vs. limit=2.0 +2023-03-21 12:19:24,916 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=128528.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:19:26,277 INFO [train.py:901] (0/2) Epoch 46, batch 1450, loss[loss=0.1287, simple_loss=0.2154, pruned_loss=0.02096, over 7281.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2085, pruned_loss=0.0229, over 1440931.77 frames. ], batch size: 66, lr: 3.58e-03, grad_scale: 8.0 +2023-03-21 12:19:38,608 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 12:19:39,166 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=128554.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:19:43,534 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.251e+02 1.693e+02 1.942e+02 2.319e+02 7.127e+02, threshold=3.884e+02, percent-clipped=1.0 +2023-03-21 12:19:50,123 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=128576.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:19:52,614 INFO [train.py:901] (0/2) Epoch 46, batch 1500, loss[loss=0.1419, simple_loss=0.2289, pruned_loss=0.02747, over 6497.00 frames. ], tot_loss[loss=0.1264, simple_loss=0.2079, pruned_loss=0.02249, over 1441277.24 frames. ], batch size: 106, lr: 3.58e-03, grad_scale: 8.0 +2023-03-21 12:19:54,686 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 12:20:05,405 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5820, 1.4843, 1.6941, 2.0466, 1.7737, 2.0340, 1.5049, 2.0005], + device='cuda:0'), covar=tensor([0.2493, 0.3233, 0.1531, 0.1182, 0.1226, 0.0878, 0.2059, 0.2004], + device='cuda:0'), in_proj_covar=tensor([0.0086, 0.0084, 0.0078, 0.0071, 0.0069, 0.0068, 0.0110, 0.0071], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 12:20:18,686 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 12:20:19,188 INFO [train.py:901] (0/2) Epoch 46, batch 1550, loss[loss=0.1464, simple_loss=0.2285, pruned_loss=0.03213, over 7253.00 frames. ], tot_loss[loss=0.127, simple_loss=0.2085, pruned_loss=0.02273, over 1441349.58 frames. ], batch size: 55, lr: 3.58e-03, grad_scale: 8.0 +2023-03-21 12:20:29,467 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2266, 2.1265, 2.5251, 2.2012, 2.3958, 2.4898, 2.3419, 1.8442], + device='cuda:0'), covar=tensor([0.0376, 0.0634, 0.0343, 0.0323, 0.0513, 0.0391, 0.0365, 0.0340], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0042, 0.0042, 0.0041, 0.0039, 0.0040, 0.0046, 0.0045], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 12:20:35,800 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.290e+02 1.729e+02 1.986e+02 2.320e+02 3.643e+02, threshold=3.971e+02, percent-clipped=0.0 +2023-03-21 12:20:42,477 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=128676.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:20:44,920 INFO [train.py:901] (0/2) Epoch 46, batch 1600, loss[loss=0.1276, simple_loss=0.2128, pruned_loss=0.0212, over 7340.00 frames. ], tot_loss[loss=0.1267, simple_loss=0.2084, pruned_loss=0.02252, over 1442886.01 frames. ], batch size: 59, lr: 3.58e-03, grad_scale: 8.0 +2023-03-21 12:20:49,929 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 12:20:50,945 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 12:20:53,455 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 12:20:55,098 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2092, 3.4286, 2.3808, 3.6603, 2.8506, 3.1735, 1.6464, 2.5786], + device='cuda:0'), covar=tensor([0.0531, 0.0956, 0.2951, 0.0853, 0.0579, 0.0910, 0.4081, 0.1874], + device='cuda:0'), in_proj_covar=tensor([0.0270, 0.0253, 0.0272, 0.0264, 0.0262, 0.0259, 0.0224, 0.0253], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 12:21:00,357 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5380, 3.3061, 3.3944, 3.5763, 3.1866, 3.0580, 3.6863, 2.5059], + device='cuda:0'), covar=tensor([0.0542, 0.0609, 0.0804, 0.0789, 0.0875, 0.1170, 0.0839, 0.2990], + device='cuda:0'), in_proj_covar=tensor([0.0327, 0.0336, 0.0268, 0.0348, 0.0279, 0.0282, 0.0343, 0.0236], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 12:21:01,367 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0858, 3.4970, 2.9993, 3.2817, 3.3854, 2.9633, 3.1321, 3.1240], + device='cuda:0'), covar=tensor([0.0802, 0.0553, 0.0862, 0.1124, 0.0829, 0.0562, 0.1550, 0.1087], + device='cuda:0'), in_proj_covar=tensor([0.0061, 0.0061, 0.0070, 0.0062, 0.0058, 0.0066, 0.0059, 0.0056], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 12:21:03,839 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 12:21:07,975 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 12:21:11,654 INFO [train.py:901] (0/2) Epoch 46, batch 1650, loss[loss=0.1615, simple_loss=0.2391, pruned_loss=0.04191, over 6643.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2087, pruned_loss=0.02271, over 1443385.76 frames. ], batch size: 107, lr: 3.58e-03, grad_scale: 8.0 +2023-03-21 12:21:17,731 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 12:21:27,885 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.187e+02 1.643e+02 1.877e+02 2.168e+02 3.378e+02, threshold=3.754e+02, percent-clipped=0.0 +2023-03-21 12:21:30,983 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=128769.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:21:33,466 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 12:21:36,914 INFO [train.py:901] (0/2) Epoch 46, batch 1700, loss[loss=0.1001, simple_loss=0.1731, pruned_loss=0.01359, over 6993.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2087, pruned_loss=0.02277, over 1444000.11 frames. ], batch size: 35, lr: 3.58e-03, grad_scale: 8.0 +2023-03-21 12:21:37,056 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=128781.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:21:37,947 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 12:21:39,628 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7211, 3.4788, 3.4711, 3.3808, 3.4496, 3.3087, 3.6282, 3.3037], + device='cuda:0'), covar=tensor([0.0141, 0.0214, 0.0142, 0.0246, 0.0406, 0.0140, 0.0146, 0.0198], + device='cuda:0'), in_proj_covar=tensor([0.0099, 0.0106, 0.0108, 0.0092, 0.0182, 0.0113, 0.0109, 0.0118], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 12:21:50,007 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 12:22:04,343 INFO [train.py:901] (0/2) Epoch 46, batch 1750, loss[loss=0.1368, simple_loss=0.2104, pruned_loss=0.03159, over 7300.00 frames. ], tot_loss[loss=0.1274, simple_loss=0.209, pruned_loss=0.02297, over 1445058.05 frames. ], batch size: 47, lr: 3.58e-03, grad_scale: 8.0 +2023-03-21 12:22:10,007 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=128842.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:22:15,965 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 12:22:16,989 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 12:22:20,530 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.121e+02 1.769e+02 2.041e+02 2.508e+02 4.392e+02, threshold=4.083e+02, percent-clipped=1.0 +2023-03-21 12:22:29,563 INFO [train.py:901] (0/2) Epoch 46, batch 1800, loss[loss=0.1419, simple_loss=0.2192, pruned_loss=0.03234, over 7261.00 frames. ], tot_loss[loss=0.1275, simple_loss=0.2091, pruned_loss=0.02299, over 1445464.84 frames. ], batch size: 52, lr: 3.58e-03, grad_scale: 8.0 +2023-03-21 12:22:34,948 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2046, 3.8458, 3.8582, 3.8611, 3.9162, 3.7961, 4.1232, 3.6171], + device='cuda:0'), covar=tensor([0.0143, 0.0179, 0.0126, 0.0177, 0.0376, 0.0129, 0.0136, 0.0189], + device='cuda:0'), in_proj_covar=tensor([0.0100, 0.0107, 0.0108, 0.0093, 0.0183, 0.0114, 0.0110, 0.0119], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 12:22:34,989 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=128890.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 12:22:38,414 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 12:22:46,368 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8623, 2.1939, 1.7007, 2.0265, 2.1841, 1.9291, 1.8122, 1.5929], + device='cuda:0'), covar=tensor([0.0156, 0.0242, 0.0326, 0.0312, 0.0130, 0.0150, 0.0320, 0.0265], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0041, 0.0038, 0.0040, 0.0039, 0.0037, 0.0039, 0.0048], + device='cuda:0'), out_proj_covar=tensor([4.5954e-05, 4.5157e-05, 4.2964e-05, 4.4410e-05, 4.2677e-05, 4.0698e-05, + 4.3576e-05, 5.2915e-05], device='cuda:0') +2023-03-21 12:22:51,850 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 12:22:56,410 INFO [train.py:901] (0/2) Epoch 46, batch 1850, loss[loss=0.08841, simple_loss=0.1526, pruned_loss=0.01212, over 6563.00 frames. ], tot_loss[loss=0.1275, simple_loss=0.2091, pruned_loss=0.02293, over 1444520.18 frames. ], batch size: 29, lr: 3.58e-03, grad_scale: 8.0 +2023-03-21 12:23:02,501 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 12:23:06,771 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=128951.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 12:23:12,630 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.131e+02 1.665e+02 1.878e+02 2.204e+02 4.262e+02, threshold=3.756e+02, percent-clipped=1.0 +2023-03-21 12:23:19,340 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 12:23:19,945 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=128976.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:23:22,338 INFO [train.py:901] (0/2) Epoch 46, batch 1900, loss[loss=0.1175, simple_loss=0.2019, pruned_loss=0.01651, over 7352.00 frames. ], tot_loss[loss=0.127, simple_loss=0.2085, pruned_loss=0.02281, over 1438902.03 frames. ], batch size: 61, lr: 3.58e-03, grad_scale: 8.0 +2023-03-21 12:23:35,159 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9159, 4.2697, 4.6022, 4.5823, 4.6815, 4.5502, 4.8602, 4.3441], + device='cuda:0'), covar=tensor([0.0170, 0.0173, 0.0115, 0.0132, 0.0322, 0.0112, 0.0101, 0.0197], + device='cuda:0'), in_proj_covar=tensor([0.0100, 0.0107, 0.0108, 0.0093, 0.0183, 0.0114, 0.0110, 0.0119], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 12:23:41,496 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5335, 1.7185, 1.4491, 1.6362, 1.7441, 1.6970, 1.7206, 1.4031], + device='cuda:0'), covar=tensor([0.0238, 0.0337, 0.0386, 0.0268, 0.0237, 0.0292, 0.0222, 0.0221], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0040, 0.0038, 0.0040, 0.0038, 0.0037, 0.0039, 0.0048], + device='cuda:0'), out_proj_covar=tensor([4.5479e-05, 4.4805e-05, 4.2736e-05, 4.3988e-05, 4.2302e-05, 4.0474e-05, + 4.3261e-05, 5.2491e-05], device='cuda:0') +2023-03-21 12:23:44,383 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 12:23:45,435 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=129024.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:23:49,013 INFO [train.py:901] (0/2) Epoch 46, batch 1950, loss[loss=0.1334, simple_loss=0.2085, pruned_loss=0.02918, over 7321.00 frames. ], tot_loss[loss=0.1263, simple_loss=0.2075, pruned_loss=0.02251, over 1436105.35 frames. ], batch size: 49, lr: 3.58e-03, grad_scale: 8.0 +2023-03-21 12:23:56,727 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 12:23:58,365 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8549, 3.7009, 2.8594, 3.4762, 2.8120, 2.0723, 1.7264, 3.8416], + device='cuda:0'), covar=tensor([0.0070, 0.0080, 0.0199, 0.0106, 0.0237, 0.0724, 0.0810, 0.0081], + device='cuda:0'), in_proj_covar=tensor([0.0120, 0.0096, 0.0118, 0.0101, 0.0134, 0.0137, 0.0130, 0.0107], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 12:24:01,390 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 12:24:02,399 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 12:24:05,975 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.250e+02 1.809e+02 2.110e+02 2.690e+02 9.152e+02, threshold=4.221e+02, percent-clipped=6.0 +2023-03-21 12:24:09,088 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=129069.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:24:15,734 INFO [train.py:901] (0/2) Epoch 46, batch 2000, loss[loss=0.1058, simple_loss=0.1904, pruned_loss=0.01061, over 7137.00 frames. ], tot_loss[loss=0.1259, simple_loss=0.2068, pruned_loss=0.02247, over 1435284.70 frames. ], batch size: 41, lr: 3.58e-03, grad_scale: 8.0 +2023-03-21 12:24:19,884 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 12:24:30,203 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.21 vs. limit=2.0 +2023-03-21 12:24:30,957 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 12:24:32,893 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.75 vs. limit=5.0 +2023-03-21 12:24:34,070 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=129117.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:24:38,627 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 12:24:41,085 INFO [train.py:901] (0/2) Epoch 46, batch 2050, loss[loss=0.1172, simple_loss=0.2006, pruned_loss=0.01687, over 7349.00 frames. ], tot_loss[loss=0.1267, simple_loss=0.2078, pruned_loss=0.02282, over 1438953.14 frames. ], batch size: 63, lr: 3.58e-03, grad_scale: 8.0 +2023-03-21 12:24:44,161 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=129137.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:24:57,815 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.071e+02 1.744e+02 1.999e+02 2.402e+02 4.134e+02, threshold=3.998e+02, percent-clipped=0.0 +2023-03-21 12:25:05,362 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.89 vs. limit=2.0 +2023-03-21 12:25:07,486 INFO [train.py:901] (0/2) Epoch 46, batch 2100, loss[loss=0.1234, simple_loss=0.2055, pruned_loss=0.02072, over 7254.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2082, pruned_loss=0.02295, over 1440003.34 frames. ], batch size: 89, lr: 3.57e-03, grad_scale: 8.0 +2023-03-21 12:25:12,584 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 12:25:15,611 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 12:25:24,923 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6452, 5.1126, 5.2323, 5.1470, 4.9566, 4.6798, 5.2326, 5.0726], + device='cuda:0'), covar=tensor([0.0405, 0.0370, 0.0346, 0.0399, 0.0373, 0.0392, 0.0325, 0.0422], + device='cuda:0'), in_proj_covar=tensor([0.0217, 0.0266, 0.0208, 0.0203, 0.0160, 0.0233, 0.0215, 0.0149], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 12:25:33,589 INFO [train.py:901] (0/2) Epoch 46, batch 2150, loss[loss=0.1169, simple_loss=0.2043, pruned_loss=0.01476, over 7288.00 frames. ], tot_loss[loss=0.1267, simple_loss=0.2081, pruned_loss=0.02272, over 1439823.74 frames. ], batch size: 68, lr: 3.57e-03, grad_scale: 8.0 +2023-03-21 12:25:42,520 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=129246.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 12:25:50,978 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.299e+02 1.747e+02 2.060e+02 2.463e+02 4.425e+02, threshold=4.120e+02, percent-clipped=1.0 +2023-03-21 12:25:59,670 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1024, 2.6104, 3.2144, 2.9634, 3.2629, 2.9199, 2.6916, 3.0481], + device='cuda:0'), covar=tensor([0.1267, 0.0888, 0.1222, 0.1635, 0.0747, 0.1191, 0.1573, 0.1508], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0072, 0.0055, 0.0053, 0.0054, 0.0053, 0.0071, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 12:26:00,054 INFO [train.py:901] (0/2) Epoch 46, batch 2200, loss[loss=0.1276, simple_loss=0.2108, pruned_loss=0.02217, over 7304.00 frames. ], tot_loss[loss=0.1267, simple_loss=0.2079, pruned_loss=0.02275, over 1439731.17 frames. ], batch size: 59, lr: 3.57e-03, grad_scale: 8.0 +2023-03-21 12:26:02,535 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 12:26:26,031 INFO [train.py:901] (0/2) Epoch 46, batch 2250, loss[loss=0.1421, simple_loss=0.2163, pruned_loss=0.0339, over 7260.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2085, pruned_loss=0.02287, over 1442760.20 frames. ], batch size: 47, lr: 3.57e-03, grad_scale: 8.0 +2023-03-21 12:26:29,523 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.02 vs. limit=5.0 +2023-03-21 12:26:37,747 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 12:26:37,759 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 12:26:42,750 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.311e+02 1.693e+02 1.987e+02 2.200e+02 3.291e+02, threshold=3.973e+02, percent-clipped=0.0 +2023-03-21 12:26:50,357 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 12:26:51,829 INFO [train.py:901] (0/2) Epoch 46, batch 2300, loss[loss=0.1105, simple_loss=0.1884, pruned_loss=0.01631, over 7167.00 frames. ], tot_loss[loss=0.1272, simple_loss=0.2089, pruned_loss=0.0228, over 1443199.15 frames. ], batch size: 39, lr: 3.57e-03, grad_scale: 8.0 +2023-03-21 12:26:56,345 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4336, 3.9226, 3.8231, 4.3624, 4.2128, 4.2631, 3.7487, 3.9014], + device='cuda:0'), covar=tensor([0.0846, 0.2535, 0.2483, 0.1103, 0.1007, 0.1279, 0.1075, 0.1163], + device='cuda:0'), in_proj_covar=tensor([0.0268, 0.0399, 0.0302, 0.0319, 0.0234, 0.0376, 0.0233, 0.0281], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 12:27:09,049 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=129413.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 12:27:14,751 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6114, 1.6237, 2.0605, 2.2787, 2.0290, 2.2849, 1.8771, 2.1991], + device='cuda:0'), covar=tensor([0.3298, 0.2870, 0.2627, 0.1424, 0.1993, 0.1871, 0.1972, 0.2143], + device='cuda:0'), in_proj_covar=tensor([0.0083, 0.0083, 0.0077, 0.0068, 0.0068, 0.0067, 0.0106, 0.0069], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 12:27:18,753 INFO [train.py:901] (0/2) Epoch 46, batch 2350, loss[loss=0.1257, simple_loss=0.2084, pruned_loss=0.02154, over 7274.00 frames. ], tot_loss[loss=0.1272, simple_loss=0.2088, pruned_loss=0.02274, over 1444129.80 frames. ], batch size: 57, lr: 3.57e-03, grad_scale: 8.0 +2023-03-21 12:27:21,898 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=129437.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:27:34,996 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.209e+02 1.731e+02 2.025e+02 2.422e+02 3.708e+02, threshold=4.049e+02, percent-clipped=0.0 +2023-03-21 12:27:37,105 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 12:27:40,786 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=129474.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 12:27:43,718 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 12:27:44,209 INFO [train.py:901] (0/2) Epoch 46, batch 2400, loss[loss=0.1316, simple_loss=0.2048, pruned_loss=0.02917, over 7284.00 frames. ], tot_loss[loss=0.1274, simple_loss=0.2092, pruned_loss=0.02285, over 1445666.51 frames. ], batch size: 47, lr: 3.57e-03, grad_scale: 8.0 +2023-03-21 12:27:46,358 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=129485.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:27:50,928 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7783, 2.8234, 3.7771, 3.7761, 3.9028, 3.8161, 3.8233, 3.7783], + device='cuda:0'), covar=tensor([0.0033, 0.0166, 0.0034, 0.0037, 0.0028, 0.0035, 0.0055, 0.0047], + device='cuda:0'), in_proj_covar=tensor([0.0052, 0.0074, 0.0061, 0.0060, 0.0057, 0.0063, 0.0050, 0.0082], + device='cuda:0'), out_proj_covar=tensor([8.6044e-05, 1.4667e-04, 1.0801e-04, 1.0160e-04, 9.4276e-05, 1.0680e-04, + 9.2889e-05, 1.4787e-04], device='cuda:0') +2023-03-21 12:27:53,994 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 12:27:57,017 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 12:28:00,533 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.32 vs. limit=5.0 +2023-03-21 12:28:10,911 INFO [train.py:901] (0/2) Epoch 46, batch 2450, loss[loss=0.1129, simple_loss=0.1917, pruned_loss=0.01707, over 7349.00 frames. ], tot_loss[loss=0.1273, simple_loss=0.2091, pruned_loss=0.0228, over 1445513.33 frames. ], batch size: 44, lr: 3.57e-03, grad_scale: 8.0 +2023-03-21 12:28:18,691 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=129546.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 12:28:23,900 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3809, 2.7898, 2.5464, 2.6942, 2.6593, 2.4295, 2.7437, 2.5568], + device='cuda:0'), covar=tensor([0.0848, 0.0532, 0.0845, 0.0804, 0.0731, 0.0605, 0.0850, 0.0846], + device='cuda:0'), in_proj_covar=tensor([0.0060, 0.0061, 0.0070, 0.0062, 0.0058, 0.0065, 0.0059, 0.0056], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 12:28:24,807 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 12:28:27,229 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.108e+02 1.840e+02 2.063e+02 2.420e+02 3.694e+02, threshold=4.127e+02, percent-clipped=0.0 +2023-03-21 12:28:33,988 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6750, 4.3548, 3.8546, 4.1495, 3.7732, 2.6351, 2.4571, 4.6818], + device='cuda:0'), covar=tensor([0.0044, 0.0069, 0.0109, 0.0066, 0.0125, 0.0564, 0.0539, 0.0043], + device='cuda:0'), in_proj_covar=tensor([0.0121, 0.0097, 0.0118, 0.0101, 0.0135, 0.0137, 0.0130, 0.0107], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 12:28:36,384 INFO [train.py:901] (0/2) Epoch 46, batch 2500, loss[loss=0.138, simple_loss=0.2165, pruned_loss=0.0298, over 7284.00 frames. ], tot_loss[loss=0.1268, simple_loss=0.2085, pruned_loss=0.02258, over 1446309.30 frames. ], batch size: 66, lr: 3.57e-03, grad_scale: 8.0 +2023-03-21 12:28:38,648 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.90 vs. limit=2.0 +2023-03-21 12:28:43,650 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=129594.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 12:28:52,158 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 12:28:53,883 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=129612.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:29:03,269 INFO [train.py:901] (0/2) Epoch 46, batch 2550, loss[loss=0.1278, simple_loss=0.2133, pruned_loss=0.02112, over 7346.00 frames. ], tot_loss[loss=0.1277, simple_loss=0.2092, pruned_loss=0.02306, over 1446422.30 frames. ], batch size: 54, lr: 3.57e-03, grad_scale: 16.0 +2023-03-21 12:29:13,687 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5969, 1.8068, 2.2487, 2.4490, 2.4008, 2.4215, 2.2926, 2.4036], + device='cuda:0'), covar=tensor([1.3718, 0.3728, 0.3786, 0.1554, 0.1788, 0.3324, 0.2190, 0.3463], + device='cuda:0'), in_proj_covar=tensor([0.0084, 0.0083, 0.0077, 0.0069, 0.0069, 0.0068, 0.0107, 0.0070], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 12:29:19,655 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.370e+02 1.850e+02 2.116e+02 2.396e+02 5.730e+02, threshold=4.232e+02, percent-clipped=1.0 +2023-03-21 12:29:24,863 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=129673.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:29:29,341 INFO [train.py:901] (0/2) Epoch 46, batch 2600, loss[loss=0.1137, simple_loss=0.1965, pruned_loss=0.01544, over 7269.00 frames. ], tot_loss[loss=0.1275, simple_loss=0.2094, pruned_loss=0.02278, over 1446363.14 frames. ], batch size: 70, lr: 3.57e-03, grad_scale: 16.0 +2023-03-21 12:29:30,665 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.89 vs. limit=2.0 +2023-03-21 12:29:32,997 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0945, 3.4685, 4.3120, 4.2710, 4.3821, 4.2702, 4.3162, 4.2814], + device='cuda:0'), covar=tensor([0.0036, 0.0129, 0.0029, 0.0039, 0.0028, 0.0032, 0.0035, 0.0039], + device='cuda:0'), in_proj_covar=tensor([0.0052, 0.0074, 0.0061, 0.0060, 0.0057, 0.0063, 0.0050, 0.0082], + device='cuda:0'), out_proj_covar=tensor([8.6220e-05, 1.4665e-04, 1.0813e-04, 1.0144e-04, 9.4031e-05, 1.0678e-04, + 9.2222e-05, 1.4785e-04], device='cuda:0') +2023-03-21 12:29:35,980 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.4808, 1.7489, 1.5313, 1.7811, 1.7240, 1.6462, 1.6447, 1.3372], + device='cuda:0'), covar=tensor([0.0216, 0.0215, 0.0434, 0.0228, 0.0279, 0.0190, 0.0194, 0.0283], + device='cuda:0'), in_proj_covar=tensor([0.0041, 0.0041, 0.0038, 0.0040, 0.0039, 0.0037, 0.0039, 0.0049], + device='cuda:0'), out_proj_covar=tensor([4.6100e-05, 4.5297e-05, 4.3361e-05, 4.4496e-05, 4.2896e-05, 4.1229e-05, + 4.3838e-05, 5.3236e-05], device='cuda:0') +2023-03-21 12:29:54,725 INFO [train.py:901] (0/2) Epoch 46, batch 2650, loss[loss=0.1121, simple_loss=0.1727, pruned_loss=0.02574, over 6144.00 frames. ], tot_loss[loss=0.1277, simple_loss=0.2093, pruned_loss=0.02304, over 1445092.86 frames. ], batch size: 27, lr: 3.57e-03, grad_scale: 8.0 +2023-03-21 12:30:09,596 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5055, 4.9776, 5.0567, 4.9941, 4.8546, 4.5153, 5.0833, 4.8637], + device='cuda:0'), covar=tensor([0.0432, 0.0405, 0.0380, 0.0472, 0.0344, 0.0453, 0.0337, 0.0446], + device='cuda:0'), in_proj_covar=tensor([0.0218, 0.0268, 0.0209, 0.0205, 0.0162, 0.0235, 0.0215, 0.0150], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 12:30:10,982 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.153e+02 1.766e+02 2.133e+02 2.471e+02 4.108e+02, threshold=4.267e+02, percent-clipped=0.0 +2023-03-21 12:30:13,558 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=129769.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 12:30:19,401 INFO [train.py:901] (0/2) Epoch 46, batch 2700, loss[loss=0.1412, simple_loss=0.2278, pruned_loss=0.02733, over 7337.00 frames. ], tot_loss[loss=0.1277, simple_loss=0.2093, pruned_loss=0.023, over 1443921.26 frames. ], batch size: 54, lr: 3.57e-03, grad_scale: 8.0 +2023-03-21 12:30:24,488 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=129791.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:30:37,220 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.95 vs. limit=2.0 +2023-03-21 12:30:44,370 INFO [train.py:901] (0/2) Epoch 46, batch 2750, loss[loss=0.1341, simple_loss=0.2142, pruned_loss=0.02705, over 7268.00 frames. ], tot_loss[loss=0.1274, simple_loss=0.2089, pruned_loss=0.02291, over 1441220.06 frames. ], batch size: 70, lr: 3.57e-03, grad_scale: 8.0 +2023-03-21 12:30:54,897 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=129852.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:31:00,579 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.068e+02 1.744e+02 2.096e+02 2.375e+02 4.011e+02, threshold=4.191e+02, percent-clipped=0.0 +2023-03-21 12:31:08,452 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.59 vs. limit=2.0 +2023-03-21 12:31:09,042 INFO [train.py:901] (0/2) Epoch 46, batch 2800, loss[loss=0.1332, simple_loss=0.2079, pruned_loss=0.02924, over 7261.00 frames. ], tot_loss[loss=0.1274, simple_loss=0.2087, pruned_loss=0.02305, over 1439344.17 frames. ], batch size: 55, lr: 3.57e-03, grad_scale: 8.0 +2023-03-21 12:31:21,892 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-46.pt +2023-03-21 12:31:33,960 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 12:31:34,015 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 12:31:34,038 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 12:31:34,056 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 12:31:34,563 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 12:31:34,578 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 12:31:34,590 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 12:31:34,688 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 12:31:34,764 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 12:31:34,779 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 12:31:34,796 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 12:31:34,959 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 12:31:34,965 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 12:31:34,973 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 12:31:35,115 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 12:31:35,118 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 12:31:35,177 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 12:31:37,235 INFO [train.py:901] (0/2) Epoch 47, batch 0, loss[loss=0.1022, simple_loss=0.1715, pruned_loss=0.0164, over 6958.00 frames. ], tot_loss[loss=0.1022, simple_loss=0.1715, pruned_loss=0.0164, over 6958.00 frames. ], batch size: 35, lr: 3.53e-03, grad_scale: 8.0 +2023-03-21 12:31:37,236 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 12:31:48,305 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0724, 3.7947, 3.6750, 3.7259, 3.8548, 3.7020, 3.8912, 3.5874], + device='cuda:0'), covar=tensor([0.0131, 0.0156, 0.0139, 0.0173, 0.0346, 0.0112, 0.0163, 0.0181], + device='cuda:0'), in_proj_covar=tensor([0.0100, 0.0105, 0.0108, 0.0092, 0.0181, 0.0112, 0.0109, 0.0118], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 12:32:02,790 INFO [train.py:935] (0/2) Epoch 47, validation: loss=0.166, simple_loss=0.2594, pruned_loss=0.0363, over 1622729.00 frames. +2023-03-21 12:32:02,790 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 12:32:09,857 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 12:32:20,551 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 12:32:28,808 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 12:32:29,331 INFO [train.py:901] (0/2) Epoch 47, batch 50, loss[loss=0.1324, simple_loss=0.2158, pruned_loss=0.02455, over 7252.00 frames. ], tot_loss[loss=0.1281, simple_loss=0.2102, pruned_loss=0.02305, over 325757.48 frames. ], batch size: 55, lr: 3.53e-03, grad_scale: 8.0 +2023-03-21 12:32:30,863 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 12:32:33,926 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.321e+02 1.737e+02 2.014e+02 2.442e+02 5.110e+02, threshold=4.029e+02, percent-clipped=1.0 +2023-03-21 12:32:33,949 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 12:32:36,004 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=129968.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:32:49,972 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 12:32:50,461 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 12:32:52,447 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=130000.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:32:53,462 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=130002.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:32:54,824 INFO [train.py:901] (0/2) Epoch 47, batch 100, loss[loss=0.1109, simple_loss=0.194, pruned_loss=0.01384, over 7344.00 frames. ], tot_loss[loss=0.1273, simple_loss=0.2096, pruned_loss=0.02251, over 572135.06 frames. ], batch size: 44, lr: 3.53e-03, grad_scale: 8.0 +2023-03-21 12:33:08,482 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6933, 3.0256, 2.5986, 2.8808, 2.8354, 2.5471, 2.8988, 2.6724], + device='cuda:0'), covar=tensor([0.0566, 0.0516, 0.0979, 0.0687, 0.0955, 0.0581, 0.0541, 0.0944], + device='cuda:0'), in_proj_covar=tensor([0.0060, 0.0061, 0.0070, 0.0062, 0.0058, 0.0065, 0.0059, 0.0057], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 12:33:13,691 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.26 vs. limit=2.0 +2023-03-21 12:33:21,476 INFO [train.py:901] (0/2) Epoch 47, batch 150, loss[loss=0.1487, simple_loss=0.2299, pruned_loss=0.03369, over 7272.00 frames. ], tot_loss[loss=0.1281, simple_loss=0.2098, pruned_loss=0.02321, over 765096.51 frames. ], batch size: 64, lr: 3.52e-03, grad_scale: 8.0 +2023-03-21 12:33:24,620 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=130061.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:33:25,647 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=130063.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 12:33:25,990 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.052e+02 1.724e+02 1.922e+02 2.327e+02 4.445e+02, threshold=3.845e+02, percent-clipped=1.0 +2023-03-21 12:33:28,633 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=130069.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 12:33:47,460 INFO [train.py:901] (0/2) Epoch 47, batch 200, loss[loss=0.1226, simple_loss=0.2017, pruned_loss=0.02178, over 7284.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2085, pruned_loss=0.02288, over 913804.21 frames. ], batch size: 68, lr: 3.52e-03, grad_scale: 8.0 +2023-03-21 12:33:47,843 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.63 vs. limit=2.0 +2023-03-21 12:33:51,893 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 12:33:53,993 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=130117.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 12:33:57,509 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 12:34:03,514 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 12:34:09,141 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=130147.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:34:13,052 INFO [train.py:901] (0/2) Epoch 47, batch 250, loss[loss=0.1305, simple_loss=0.206, pruned_loss=0.02753, over 7274.00 frames. ], tot_loss[loss=0.1269, simple_loss=0.2082, pruned_loss=0.02282, over 1032057.55 frames. ], batch size: 57, lr: 3.52e-03, grad_scale: 8.0 +2023-03-21 12:34:16,683 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 12:34:17,638 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.342e+02 1.845e+02 2.146e+02 2.522e+02 4.648e+02, threshold=4.292e+02, percent-clipped=1.0 +2023-03-21 12:34:39,004 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 12:34:39,492 INFO [train.py:901] (0/2) Epoch 47, batch 300, loss[loss=0.1322, simple_loss=0.2107, pruned_loss=0.02681, over 7290.00 frames. ], tot_loss[loss=0.1264, simple_loss=0.2077, pruned_loss=0.02259, over 1122623.11 frames. ], batch size: 80, lr: 3.52e-03, grad_scale: 8.0 +2023-03-21 12:34:49,087 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 12:34:49,349 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.24 vs. limit=2.0 +2023-03-21 12:35:05,231 INFO [train.py:901] (0/2) Epoch 47, batch 350, loss[loss=0.1358, simple_loss=0.2154, pruned_loss=0.02807, over 7275.00 frames. ], tot_loss[loss=0.1254, simple_loss=0.2065, pruned_loss=0.02212, over 1193081.54 frames. ], batch size: 57, lr: 3.52e-03, grad_scale: 8.0 +2023-03-21 12:35:09,770 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.165e+02 1.604e+02 1.967e+02 2.277e+02 3.840e+02, threshold=3.934e+02, percent-clipped=0.0 +2023-03-21 12:35:11,879 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=130268.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:35:20,691 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.69 vs. limit=2.0 +2023-03-21 12:35:23,408 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 12:35:25,809 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-03-21 12:35:31,578 INFO [train.py:901] (0/2) Epoch 47, batch 400, loss[loss=0.1326, simple_loss=0.2199, pruned_loss=0.02265, over 7322.00 frames. ], tot_loss[loss=0.126, simple_loss=0.2071, pruned_loss=0.02243, over 1249283.41 frames. ], batch size: 83, lr: 3.52e-03, grad_scale: 8.0 +2023-03-21 12:35:37,183 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=130316.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:35:37,253 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4556, 4.2037, 3.4627, 3.9486, 3.6354, 2.4008, 1.8803, 4.4384], + device='cuda:0'), covar=tensor([0.0041, 0.0066, 0.0144, 0.0064, 0.0125, 0.0642, 0.0686, 0.0044], + device='cuda:0'), in_proj_covar=tensor([0.0120, 0.0096, 0.0119, 0.0101, 0.0136, 0.0138, 0.0130, 0.0108], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 12:35:56,379 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9054, 4.3595, 4.3640, 4.3478, 4.3984, 3.9916, 4.4348, 4.3316], + device='cuda:0'), covar=tensor([0.0557, 0.0451, 0.0439, 0.0517, 0.0341, 0.0492, 0.0370, 0.0423], + device='cuda:0'), in_proj_covar=tensor([0.0220, 0.0270, 0.0210, 0.0206, 0.0163, 0.0237, 0.0216, 0.0152], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 12:35:56,785 INFO [train.py:901] (0/2) Epoch 47, batch 450, loss[loss=0.1343, simple_loss=0.2251, pruned_loss=0.02174, over 6707.00 frames. ], tot_loss[loss=0.1257, simple_loss=0.207, pruned_loss=0.02217, over 1291832.59 frames. ], batch size: 106, lr: 3.52e-03, grad_scale: 8.0 +2023-03-21 12:35:57,369 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4396, 4.9060, 4.9644, 4.9400, 4.8196, 4.4666, 4.9964, 4.8209], + device='cuda:0'), covar=tensor([0.0468, 0.0400, 0.0382, 0.0419, 0.0355, 0.0440, 0.0337, 0.0435], + device='cuda:0'), in_proj_covar=tensor([0.0220, 0.0270, 0.0210, 0.0206, 0.0163, 0.0237, 0.0216, 0.0152], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 12:35:57,373 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=130356.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:35:58,366 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=130358.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 12:36:00,955 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6605, 5.1386, 5.2116, 5.1691, 4.9948, 4.6731, 5.2433, 5.0439], + device='cuda:0'), covar=tensor([0.0483, 0.0381, 0.0361, 0.0423, 0.0360, 0.0421, 0.0320, 0.0423], + device='cuda:0'), in_proj_covar=tensor([0.0221, 0.0271, 0.0210, 0.0206, 0.0163, 0.0238, 0.0216, 0.0152], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 12:36:01,352 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.257e+02 1.656e+02 1.997e+02 2.369e+02 3.103e+02, threshold=3.993e+02, percent-clipped=0.0 +2023-03-21 12:36:04,349 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 12:36:05,469 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 12:36:12,126 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=130384.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:36:23,528 INFO [train.py:901] (0/2) Epoch 47, batch 500, loss[loss=0.1267, simple_loss=0.2078, pruned_loss=0.02279, over 7280.00 frames. ], tot_loss[loss=0.1259, simple_loss=0.2075, pruned_loss=0.02222, over 1327277.15 frames. ], batch size: 70, lr: 3.52e-03, grad_scale: 8.0 +2023-03-21 12:36:37,446 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=130432.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:36:37,857 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 12:36:38,911 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 12:36:39,890 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 12:36:40,450 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=130438.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:36:41,891 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 12:36:44,061 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=130445.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:36:45,063 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=130447.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:36:46,467 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 12:36:48,933 INFO [train.py:901] (0/2) Epoch 47, batch 550, loss[loss=0.1327, simple_loss=0.2146, pruned_loss=0.02539, over 7329.00 frames. ], tot_loss[loss=0.1262, simple_loss=0.2076, pruned_loss=0.02244, over 1351712.18 frames. ], batch size: 59, lr: 3.52e-03, grad_scale: 8.0 +2023-03-21 12:36:54,107 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.358e+02 1.782e+02 2.105e+02 2.529e+02 4.728e+02, threshold=4.210e+02, percent-clipped=3.0 +2023-03-21 12:36:54,265 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7174, 2.5059, 2.7451, 2.5734, 2.5777, 2.5215, 2.4979, 2.1329], + device='cuda:0'), covar=tensor([0.0318, 0.0580, 0.0390, 0.0322, 0.0632, 0.0507, 0.0353, 0.0380], + device='cuda:0'), in_proj_covar=tensor([0.0045, 0.0043, 0.0044, 0.0043, 0.0040, 0.0041, 0.0047, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 12:36:59,372 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 12:37:07,887 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 12:37:09,574 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=130493.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:37:10,515 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=130495.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:37:10,990 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 12:37:12,623 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=130499.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:37:15,524 INFO [train.py:901] (0/2) Epoch 47, batch 600, loss[loss=0.1287, simple_loss=0.2104, pruned_loss=0.02348, over 7312.00 frames. ], tot_loss[loss=0.1258, simple_loss=0.207, pruned_loss=0.02228, over 1372669.27 frames. ], batch size: 80, lr: 3.52e-03, grad_scale: 8.0 +2023-03-21 12:37:16,218 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3896, 3.4858, 2.3497, 3.7562, 2.9516, 3.4219, 1.6336, 2.3961], + device='cuda:0'), covar=tensor([0.0611, 0.1252, 0.3403, 0.0761, 0.0527, 0.0958, 0.4627, 0.2436], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0257, 0.0279, 0.0269, 0.0266, 0.0262, 0.0229, 0.0257], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0004, 0.0004, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 12:37:18,084 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 12:37:34,334 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 12:37:38,133 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=130548.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:37:42,154 INFO [train.py:901] (0/2) Epoch 47, batch 650, loss[loss=0.1433, simple_loss=0.2256, pruned_loss=0.03044, over 7116.00 frames. ], tot_loss[loss=0.1264, simple_loss=0.2074, pruned_loss=0.02268, over 1385984.02 frames. ], batch size: 98, lr: 3.52e-03, grad_scale: 8.0 +2023-03-21 12:37:43,723 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 12:37:46,754 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.164e+02 1.631e+02 1.961e+02 2.425e+02 4.775e+02, threshold=3.922e+02, percent-clipped=1.0 +2023-03-21 12:37:47,362 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=130565.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:38:00,773 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 12:38:07,801 INFO [train.py:901] (0/2) Epoch 47, batch 700, loss[loss=0.1347, simple_loss=0.2206, pruned_loss=0.02441, over 7283.00 frames. ], tot_loss[loss=0.1265, simple_loss=0.2077, pruned_loss=0.02266, over 1400229.86 frames. ], batch size: 86, lr: 3.52e-03, grad_scale: 8.0 +2023-03-21 12:38:09,971 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=130609.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 12:38:10,305 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 12:38:18,371 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=130626.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:38:20,378 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7135, 2.3774, 3.1730, 2.7534, 3.1179, 2.7835, 2.4210, 2.9661], + device='cuda:0'), covar=tensor([0.1738, 0.1235, 0.0675, 0.1278, 0.0739, 0.1099, 0.2279, 0.1230], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0071, 0.0054, 0.0053, 0.0053, 0.0052, 0.0070, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 12:38:33,009 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0636, 2.7933, 2.7143, 4.0159, 2.2104, 3.8574, 1.5987, 3.5553], + device='cuda:0'), covar=tensor([0.0242, 0.1333, 0.1961, 0.0243, 0.4168, 0.0320, 0.1373, 0.0494], + device='cuda:0'), in_proj_covar=tensor([0.0208, 0.0242, 0.0254, 0.0212, 0.0248, 0.0219, 0.0220, 0.0228], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 12:38:34,374 INFO [train.py:901] (0/2) Epoch 47, batch 750, loss[loss=0.1048, simple_loss=0.1811, pruned_loss=0.01428, over 7132.00 frames. ], tot_loss[loss=0.1269, simple_loss=0.2081, pruned_loss=0.02284, over 1409894.78 frames. ], batch size: 41, lr: 3.52e-03, grad_scale: 8.0 +2023-03-21 12:38:34,981 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=130656.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:38:35,378 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 12:38:35,870 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 12:38:35,950 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=130658.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 12:38:38,840 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.113e+02 1.747e+02 2.015e+02 2.453e+02 3.447e+02, threshold=4.031e+02, percent-clipped=0.0 +2023-03-21 12:38:39,478 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7351, 3.9013, 3.6740, 3.8602, 3.4699, 3.7759, 4.1234, 4.1059], + device='cuda:0'), covar=tensor([0.0253, 0.0170, 0.0256, 0.0176, 0.0373, 0.0329, 0.0256, 0.0233], + device='cuda:0'), in_proj_covar=tensor([0.0126, 0.0129, 0.0123, 0.0128, 0.0116, 0.0103, 0.0101, 0.0105], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 12:38:49,447 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 12:38:54,001 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 12:38:58,980 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=130704.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:38:59,434 INFO [train.py:901] (0/2) Epoch 47, batch 800, loss[loss=0.1318, simple_loss=0.2134, pruned_loss=0.02513, over 7259.00 frames. ], tot_loss[loss=0.127, simple_loss=0.2081, pruned_loss=0.02291, over 1417761.27 frames. ], batch size: 52, lr: 3.52e-03, grad_scale: 8.0 +2023-03-21 12:38:59,991 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 12:39:00,028 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=130706.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:39:01,999 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 12:39:14,480 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 12:39:18,575 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=130740.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:39:25,940 INFO [train.py:901] (0/2) Epoch 47, batch 850, loss[loss=0.1174, simple_loss=0.1991, pruned_loss=0.01784, over 7356.00 frames. ], tot_loss[loss=0.1263, simple_loss=0.2074, pruned_loss=0.0226, over 1424262.81 frames. ], batch size: 73, lr: 3.52e-03, grad_scale: 8.0 +2023-03-21 12:39:30,563 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.147e+02 1.714e+02 1.949e+02 2.244e+02 4.370e+02, threshold=3.898e+02, percent-clipped=1.0 +2023-03-21 12:39:32,165 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 12:39:32,177 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 12:39:37,692 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 12:39:40,740 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 12:39:42,766 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=130788.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:39:45,818 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=130794.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:39:48,198 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.85 vs. limit=5.0 +2023-03-21 12:39:52,269 INFO [train.py:901] (0/2) Epoch 47, batch 900, loss[loss=0.1166, simple_loss=0.2017, pruned_loss=0.01571, over 7133.00 frames. ], tot_loss[loss=0.1267, simple_loss=0.2081, pruned_loss=0.02267, over 1426865.69 frames. ], batch size: 41, lr: 3.51e-03, grad_scale: 8.0 +2023-03-21 12:40:18,166 INFO [train.py:901] (0/2) Epoch 47, batch 950, loss[loss=0.1373, simple_loss=0.2236, pruned_loss=0.02549, over 6782.00 frames. ], tot_loss[loss=0.1276, simple_loss=0.2091, pruned_loss=0.02309, over 1430624.23 frames. ], batch size: 107, lr: 3.51e-03, grad_scale: 8.0 +2023-03-21 12:40:19,139 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 12:40:22,689 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.245e+02 1.741e+02 1.966e+02 2.320e+02 4.605e+02, threshold=3.932e+02, percent-clipped=1.0 +2023-03-21 12:40:30,146 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.78 vs. limit=5.0 +2023-03-21 12:40:33,472 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9326, 3.2342, 2.8499, 3.0943, 3.1167, 3.0095, 3.1663, 2.9384], + device='cuda:0'), covar=tensor([0.1109, 0.0993, 0.0955, 0.0900, 0.1085, 0.0550, 0.0825, 0.0919], + device='cuda:0'), in_proj_covar=tensor([0.0061, 0.0063, 0.0071, 0.0063, 0.0059, 0.0066, 0.0060, 0.0057], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 12:40:42,317 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 12:40:44,421 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=130904.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 12:40:44,849 INFO [train.py:901] (0/2) Epoch 47, batch 1000, loss[loss=0.1041, simple_loss=0.1788, pruned_loss=0.01473, over 7005.00 frames. ], tot_loss[loss=0.1272, simple_loss=0.2089, pruned_loss=0.02279, over 1431059.56 frames. ], batch size: 35, lr: 3.51e-03, grad_scale: 8.0 +2023-03-21 12:40:53,071 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=130921.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:41:02,612 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 12:41:10,221 INFO [train.py:901] (0/2) Epoch 47, batch 1050, loss[loss=0.1141, simple_loss=0.1977, pruned_loss=0.01521, over 7145.00 frames. ], tot_loss[loss=0.1268, simple_loss=0.2082, pruned_loss=0.02268, over 1433799.41 frames. ], batch size: 41, lr: 3.51e-03, grad_scale: 8.0 +2023-03-21 12:41:14,744 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.118e+02 1.723e+02 1.973e+02 2.315e+02 3.318e+02, threshold=3.946e+02, percent-clipped=0.0 +2023-03-21 12:41:20,982 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9732, 3.1430, 2.8058, 3.1255, 3.1444, 2.9374, 3.2098, 2.9011], + device='cuda:0'), covar=tensor([0.0622, 0.1154, 0.0921, 0.0735, 0.0766, 0.0695, 0.0666, 0.1169], + device='cuda:0'), in_proj_covar=tensor([0.0061, 0.0062, 0.0071, 0.0062, 0.0059, 0.0066, 0.0060, 0.0057], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 12:41:24,981 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 12:41:29,043 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 12:41:37,123 INFO [train.py:901] (0/2) Epoch 47, batch 1100, loss[loss=0.1205, simple_loss=0.2102, pruned_loss=0.0154, over 7282.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2084, pruned_loss=0.02289, over 1436745.26 frames. ], batch size: 68, lr: 3.51e-03, grad_scale: 8.0 +2023-03-21 12:41:38,290 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5577, 2.3738, 2.7459, 2.3753, 2.6493, 2.6458, 2.3384, 2.0173], + device='cuda:0'), covar=tensor([0.0466, 0.0530, 0.0302, 0.0389, 0.0629, 0.0425, 0.0384, 0.0357], + device='cuda:0'), in_proj_covar=tensor([0.0045, 0.0043, 0.0043, 0.0043, 0.0040, 0.0041, 0.0047, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 12:41:38,426 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.22 vs. limit=2.0 +2023-03-21 12:41:45,960 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0414, 2.9712, 4.0358, 3.9809, 4.1076, 4.0180, 4.0432, 3.9118], + device='cuda:0'), covar=tensor([0.0029, 0.0163, 0.0033, 0.0033, 0.0027, 0.0033, 0.0045, 0.0051], + device='cuda:0'), in_proj_covar=tensor([0.0052, 0.0075, 0.0062, 0.0060, 0.0056, 0.0063, 0.0050, 0.0082], + device='cuda:0'), out_proj_covar=tensor([8.6072e-05, 1.4741e-04, 1.0829e-04, 1.0073e-04, 9.2208e-05, 1.0703e-04, + 9.2908e-05, 1.4852e-04], device='cuda:0') +2023-03-21 12:41:54,989 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=131040.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:41:57,969 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 12:41:58,465 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 12:42:02,457 INFO [train.py:901] (0/2) Epoch 47, batch 1150, loss[loss=0.149, simple_loss=0.234, pruned_loss=0.03195, over 7099.00 frames. ], tot_loss[loss=0.1273, simple_loss=0.2088, pruned_loss=0.02286, over 1437197.53 frames. ], batch size: 98, lr: 3.51e-03, grad_scale: 8.0 +2023-03-21 12:42:07,684 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.315e+02 1.713e+02 2.076e+02 2.452e+02 6.009e+02, threshold=4.153e+02, percent-clipped=1.0 +2023-03-21 12:42:11,912 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 12:42:12,426 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 12:42:14,979 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.1633, 4.6535, 4.5143, 5.0894, 4.9767, 5.0364, 4.5335, 4.7157], + device='cuda:0'), covar=tensor([0.0794, 0.2325, 0.2253, 0.0874, 0.0783, 0.1124, 0.0720, 0.0990], + device='cuda:0'), in_proj_covar=tensor([0.0265, 0.0400, 0.0302, 0.0314, 0.0233, 0.0374, 0.0233, 0.0282], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 12:42:17,011 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7086, 1.8700, 2.2181, 2.3162, 2.2895, 2.4251, 2.1830, 2.3142], + device='cuda:0'), covar=tensor([0.2721, 0.4622, 0.1858, 0.0855, 0.1656, 0.2522, 0.2232, 0.2170], + device='cuda:0'), in_proj_covar=tensor([0.0086, 0.0086, 0.0079, 0.0070, 0.0071, 0.0069, 0.0110, 0.0072], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 12:42:20,536 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=131088.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:42:20,583 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=131088.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:42:23,676 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=131094.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:42:29,208 INFO [train.py:901] (0/2) Epoch 47, batch 1200, loss[loss=0.1221, simple_loss=0.2007, pruned_loss=0.0218, over 7297.00 frames. ], tot_loss[loss=0.1275, simple_loss=0.2094, pruned_loss=0.02281, over 1437909.55 frames. ], batch size: 49, lr: 3.51e-03, grad_scale: 8.0 +2023-03-21 12:42:43,908 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 12:42:44,997 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=131136.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:42:48,005 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=131142.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:42:55,861 INFO [train.py:901] (0/2) Epoch 47, batch 1250, loss[loss=0.1287, simple_loss=0.212, pruned_loss=0.02273, over 7326.00 frames. ], tot_loss[loss=0.1272, simple_loss=0.2093, pruned_loss=0.0226, over 1439252.45 frames. ], batch size: 54, lr: 3.51e-03, grad_scale: 8.0 +2023-03-21 12:43:00,374 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.326e+02 1.727e+02 1.988e+02 2.304e+02 5.808e+02, threshold=3.975e+02, percent-clipped=1.0 +2023-03-21 12:43:05,742 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=131174.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:43:09,167 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 12:43:13,276 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 12:43:14,816 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 12:43:21,461 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=131204.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 12:43:21,864 INFO [train.py:901] (0/2) Epoch 47, batch 1300, loss[loss=0.1349, simple_loss=0.2116, pruned_loss=0.02913, over 7329.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2091, pruned_loss=0.0225, over 1440115.95 frames. ], batch size: 49, lr: 3.51e-03, grad_scale: 8.0 +2023-03-21 12:43:30,128 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=131221.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:43:37,802 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=131235.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 12:43:38,135 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 12:43:41,326 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 12:43:44,455 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 12:43:47,082 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=131252.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:43:48,582 INFO [train.py:901] (0/2) Epoch 47, batch 1350, loss[loss=0.1257, simple_loss=0.2064, pruned_loss=0.02253, over 7259.00 frames. ], tot_loss[loss=0.1275, simple_loss=0.2097, pruned_loss=0.02264, over 1441756.88 frames. ], batch size: 52, lr: 3.51e-03, grad_scale: 8.0 +2023-03-21 12:43:53,126 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.291e+02 1.774e+02 1.993e+02 2.320e+02 3.599e+02, threshold=3.986e+02, percent-clipped=0.0 +2023-03-21 12:43:53,649 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 12:43:55,646 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=131269.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:44:07,274 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9065, 4.1266, 3.8965, 4.1102, 3.6504, 4.0314, 4.4203, 4.4447], + device='cuda:0'), covar=tensor([0.0262, 0.0158, 0.0250, 0.0184, 0.0379, 0.0326, 0.0189, 0.0165], + device='cuda:0'), in_proj_covar=tensor([0.0130, 0.0132, 0.0126, 0.0130, 0.0119, 0.0106, 0.0103, 0.0107], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 12:44:13,895 INFO [train.py:901] (0/2) Epoch 47, batch 1400, loss[loss=0.09762, simple_loss=0.1697, pruned_loss=0.01277, over 6966.00 frames. ], tot_loss[loss=0.1274, simple_loss=0.2093, pruned_loss=0.02274, over 1442284.43 frames. ], batch size: 35, lr: 3.51e-03, grad_scale: 8.0 +2023-03-21 12:44:27,234 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 12:44:35,935 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5800, 1.7757, 1.4609, 1.8316, 1.8360, 1.7620, 1.7420, 1.3964], + device='cuda:0'), covar=tensor([0.0174, 0.0201, 0.0365, 0.0190, 0.0135, 0.0192, 0.0190, 0.0270], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0041, 0.0038, 0.0041, 0.0039, 0.0038, 0.0039, 0.0049], + device='cuda:0'), out_proj_covar=tensor([4.6707e-05, 4.5756e-05, 4.3440e-05, 4.5022e-05, 4.2756e-05, 4.1627e-05, + 4.4375e-05, 5.3522e-05], device='cuda:0') +2023-03-21 12:44:37,671 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.76 vs. limit=2.0 +2023-03-21 12:44:38,229 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.55 vs. limit=2.0 +2023-03-21 12:44:40,403 INFO [train.py:901] (0/2) Epoch 47, batch 1450, loss[loss=0.1275, simple_loss=0.2093, pruned_loss=0.02281, over 7391.00 frames. ], tot_loss[loss=0.1276, simple_loss=0.2094, pruned_loss=0.0229, over 1442970.64 frames. ], batch size: 67, lr: 3.51e-03, grad_scale: 8.0 +2023-03-21 12:44:45,045 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.405e+02 1.823e+02 2.131e+02 2.524e+02 5.111e+02, threshold=4.261e+02, percent-clipped=2.0 +2023-03-21 12:44:51,648 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 12:45:06,081 INFO [train.py:901] (0/2) Epoch 47, batch 1500, loss[loss=0.1229, simple_loss=0.205, pruned_loss=0.02044, over 7314.00 frames. ], tot_loss[loss=0.1267, simple_loss=0.2083, pruned_loss=0.02259, over 1442667.43 frames. ], batch size: 59, lr: 3.51e-03, grad_scale: 8.0 +2023-03-21 12:45:08,826 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 12:45:08,954 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=131409.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:45:13,147 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=131416.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:45:32,714 INFO [train.py:901] (0/2) Epoch 47, batch 1550, loss[loss=0.1346, simple_loss=0.2174, pruned_loss=0.02593, over 7254.00 frames. ], tot_loss[loss=0.1273, simple_loss=0.2087, pruned_loss=0.02296, over 1442530.31 frames. ], batch size: 55, lr: 3.51e-03, grad_scale: 8.0 +2023-03-21 12:45:33,718 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 12:45:37,222 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.357e+02 1.828e+02 2.189e+02 2.514e+02 5.463e+02, threshold=4.378e+02, percent-clipped=3.0 +2023-03-21 12:45:40,370 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=131470.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:45:43,981 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=131477.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:45:51,463 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=131492.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:45:59,276 INFO [train.py:901] (0/2) Epoch 47, batch 1600, loss[loss=0.1312, simple_loss=0.2141, pruned_loss=0.02413, over 7309.00 frames. ], tot_loss[loss=0.1269, simple_loss=0.2081, pruned_loss=0.02287, over 1439683.82 frames. ], batch size: 83, lr: 3.51e-03, grad_scale: 8.0 +2023-03-21 12:46:04,834 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 12:46:05,314 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 12:46:08,411 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 12:46:11,986 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=131530.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 12:46:18,021 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 12:46:22,519 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 12:46:23,677 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=131553.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:46:24,576 INFO [train.py:901] (0/2) Epoch 47, batch 1650, loss[loss=0.1413, simple_loss=0.2285, pruned_loss=0.02704, over 7357.00 frames. ], tot_loss[loss=0.1275, simple_loss=0.2089, pruned_loss=0.02303, over 1441221.56 frames. ], batch size: 51, lr: 3.50e-03, grad_scale: 8.0 +2023-03-21 12:46:28,968 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.299e+02 1.715e+02 1.954e+02 2.297e+02 4.691e+02, threshold=3.908e+02, percent-clipped=1.0 +2023-03-21 12:46:29,404 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.60 vs. limit=2.0 +2023-03-21 12:46:29,947 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 12:46:47,898 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 12:46:51,290 INFO [train.py:901] (0/2) Epoch 47, batch 1700, loss[loss=0.1182, simple_loss=0.1999, pruned_loss=0.01826, over 7275.00 frames. ], tot_loss[loss=0.1278, simple_loss=0.2092, pruned_loss=0.02317, over 1439519.27 frames. ], batch size: 70, lr: 3.50e-03, grad_scale: 8.0 +2023-03-21 12:46:52,341 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 12:46:56,927 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6634, 1.7207, 2.0226, 2.1628, 2.0451, 2.1248, 1.6789, 2.2206], + device='cuda:0'), covar=tensor([0.5116, 0.2987, 0.1150, 0.0848, 0.2580, 0.1765, 0.2352, 0.2017], + device='cuda:0'), in_proj_covar=tensor([0.0087, 0.0087, 0.0080, 0.0070, 0.0071, 0.0071, 0.0112, 0.0073], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 12:47:02,910 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 12:47:16,406 INFO [train.py:901] (0/2) Epoch 47, batch 1750, loss[loss=0.1296, simple_loss=0.2139, pruned_loss=0.0226, over 7285.00 frames. ], tot_loss[loss=0.1275, simple_loss=0.2088, pruned_loss=0.02313, over 1438067.30 frames. ], batch size: 70, lr: 3.50e-03, grad_scale: 8.0 +2023-03-21 12:47:20,892 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.170e+02 1.778e+02 2.043e+02 2.310e+02 3.986e+02, threshold=4.086e+02, percent-clipped=1.0 +2023-03-21 12:47:28,175 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 12:47:29,158 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 12:47:42,810 INFO [train.py:901] (0/2) Epoch 47, batch 1800, loss[loss=0.1442, simple_loss=0.2325, pruned_loss=0.028, over 6533.00 frames. ], tot_loss[loss=0.1274, simple_loss=0.2089, pruned_loss=0.02291, over 1438710.77 frames. ], batch size: 106, lr: 3.50e-03, grad_scale: 8.0 +2023-03-21 12:47:49,955 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 12:48:03,383 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 12:48:07,916 INFO [train.py:901] (0/2) Epoch 47, batch 1850, loss[loss=0.1452, simple_loss=0.214, pruned_loss=0.03822, over 7232.00 frames. ], tot_loss[loss=0.1264, simple_loss=0.2077, pruned_loss=0.02256, over 1438016.64 frames. ], batch size: 45, lr: 3.50e-03, grad_scale: 16.0 +2023-03-21 12:48:13,549 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.241e+02 1.650e+02 1.902e+02 2.256e+02 3.427e+02, threshold=3.803e+02, percent-clipped=0.0 +2023-03-21 12:48:14,167 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=131765.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:48:14,602 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 12:48:17,773 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=131772.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:48:19,084 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-21 12:48:30,423 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 12:48:34,816 INFO [train.py:901] (0/2) Epoch 47, batch 1900, loss[loss=0.1353, simple_loss=0.2222, pruned_loss=0.02414, over 7277.00 frames. ], tot_loss[loss=0.1259, simple_loss=0.2072, pruned_loss=0.02231, over 1437056.72 frames. ], batch size: 52, lr: 3.50e-03, grad_scale: 16.0 +2023-03-21 12:48:37,624 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.01 vs. limit=2.0 +2023-03-21 12:48:47,634 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=131830.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:48:56,794 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 12:48:57,355 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=131848.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:48:58,608 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6645, 1.9909, 1.4650, 2.0266, 1.9762, 1.8849, 1.8369, 1.4181], + device='cuda:0'), covar=tensor([0.0214, 0.0184, 0.0447, 0.0158, 0.0197, 0.0190, 0.0241, 0.0262], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0042, 0.0039, 0.0042, 0.0039, 0.0038, 0.0040, 0.0050], + device='cuda:0'), out_proj_covar=tensor([4.7700e-05, 4.6656e-05, 4.4364e-05, 4.5927e-05, 4.3384e-05, 4.2246e-05, + 4.5369e-05, 5.4784e-05], device='cuda:0') +2023-03-21 12:49:01,441 INFO [train.py:901] (0/2) Epoch 47, batch 1950, loss[loss=0.1286, simple_loss=0.2102, pruned_loss=0.02347, over 7335.00 frames. ], tot_loss[loss=0.1255, simple_loss=0.2068, pruned_loss=0.02214, over 1437902.16 frames. ], batch size: 61, lr: 3.50e-03, grad_scale: 8.0 +2023-03-21 12:49:02,624 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=131857.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:49:06,462 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.255e+02 1.657e+02 1.977e+02 2.340e+02 3.243e+02, threshold=3.954e+02, percent-clipped=0.0 +2023-03-21 12:49:06,630 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3402, 2.1750, 2.4316, 2.1540, 2.4670, 2.4147, 2.1890, 1.7839], + device='cuda:0'), covar=tensor([0.0427, 0.0408, 0.0500, 0.0306, 0.0376, 0.0326, 0.0365, 0.0432], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0042, 0.0044, 0.0043, 0.0040, 0.0041, 0.0047, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 12:49:07,999 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 12:49:13,112 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 12:49:13,159 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=131878.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:49:13,650 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 12:49:26,685 INFO [train.py:901] (0/2) Epoch 47, batch 2000, loss[loss=0.1321, simple_loss=0.2142, pruned_loss=0.02497, over 7304.00 frames. ], tot_loss[loss=0.1255, simple_loss=0.207, pruned_loss=0.02201, over 1438939.45 frames. ], batch size: 49, lr: 3.50e-03, grad_scale: 8.0 +2023-03-21 12:49:29,235 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 12:49:33,315 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=131918.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:49:41,400 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 12:49:50,849 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 12:49:53,420 INFO [train.py:901] (0/2) Epoch 47, batch 2050, loss[loss=0.1072, simple_loss=0.1876, pruned_loss=0.01339, over 7135.00 frames. ], tot_loss[loss=0.1256, simple_loss=0.207, pruned_loss=0.02212, over 1439652.76 frames. ], batch size: 41, lr: 3.50e-03, grad_scale: 8.0 +2023-03-21 12:49:58,475 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.295e+02 1.826e+02 2.096e+02 2.396e+02 3.231e+02, threshold=4.192e+02, percent-clipped=0.0 +2023-03-21 12:50:07,273 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0639, 2.1790, 2.3017, 3.4591, 1.9216, 3.1377, 1.4087, 3.1456], + device='cuda:0'), covar=tensor([0.0228, 0.1767, 0.2235, 0.0291, 0.4513, 0.0329, 0.1642, 0.0515], + device='cuda:0'), in_proj_covar=tensor([0.0208, 0.0243, 0.0254, 0.0211, 0.0248, 0.0219, 0.0220, 0.0230], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 12:50:16,568 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-132000.pt +2023-03-21 12:50:22,452 INFO [train.py:901] (0/2) Epoch 47, batch 2100, loss[loss=0.1246, simple_loss=0.2086, pruned_loss=0.02025, over 7361.00 frames. ], tot_loss[loss=0.1255, simple_loss=0.2069, pruned_loss=0.022, over 1441001.33 frames. ], batch size: 73, lr: 3.50e-03, grad_scale: 8.0 +2023-03-21 12:50:27,464 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 12:50:31,747 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 12:50:36,843 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=132031.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:50:42,888 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7145, 2.2332, 2.9429, 2.6615, 2.7059, 2.4837, 2.2349, 2.7851], + device='cuda:0'), covar=tensor([0.1160, 0.1101, 0.0634, 0.1312, 0.1022, 0.1324, 0.2347, 0.1046], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0073, 0.0055, 0.0054, 0.0054, 0.0052, 0.0072, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 12:50:43,935 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5772, 1.7425, 1.7150, 1.9779, 1.8471, 1.9969, 1.4645, 2.0428], + device='cuda:0'), covar=tensor([0.2768, 0.2387, 0.1388, 0.1167, 0.2519, 0.1117, 0.1729, 0.1167], + device='cuda:0'), in_proj_covar=tensor([0.0086, 0.0086, 0.0079, 0.0070, 0.0071, 0.0071, 0.0110, 0.0072], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 12:50:48,418 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=132054.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:50:48,797 INFO [train.py:901] (0/2) Epoch 47, batch 2150, loss[loss=0.119, simple_loss=0.2013, pruned_loss=0.01838, over 7231.00 frames. ], tot_loss[loss=0.1257, simple_loss=0.2072, pruned_loss=0.02211, over 1441698.81 frames. ], batch size: 93, lr: 3.50e-03, grad_scale: 8.0 +2023-03-21 12:50:53,764 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.210e+02 1.654e+02 1.954e+02 2.342e+02 4.874e+02, threshold=3.908e+02, percent-clipped=1.0 +2023-03-21 12:50:53,879 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=132065.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:50:57,479 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=132072.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:51:07,545 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=132092.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:51:13,870 INFO [train.py:901] (0/2) Epoch 47, batch 2200, loss[loss=0.1258, simple_loss=0.2092, pruned_loss=0.02118, over 7282.00 frames. ], tot_loss[loss=0.1258, simple_loss=0.2074, pruned_loss=0.0221, over 1441168.18 frames. ], batch size: 57, lr: 3.50e-03, grad_scale: 8.0 +2023-03-21 12:51:18,672 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 12:51:19,219 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=132113.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:51:20,314 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=132115.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 12:51:22,726 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=132120.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:51:36,925 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=132148.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:51:40,435 INFO [train.py:901] (0/2) Epoch 47, batch 2250, loss[loss=0.1121, simple_loss=0.1861, pruned_loss=0.01899, over 7169.00 frames. ], tot_loss[loss=0.1258, simple_loss=0.2071, pruned_loss=0.02221, over 1436568.35 frames. ], batch size: 39, lr: 3.50e-03, grad_scale: 8.0 +2023-03-21 12:51:43,843 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.57 vs. limit=2.0 +2023-03-21 12:51:45,424 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.229e+02 1.779e+02 2.110e+02 2.462e+02 5.777e+02, threshold=4.221e+02, percent-clipped=3.0 +2023-03-21 12:51:51,618 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 12:51:52,072 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 12:52:02,423 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=132196.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:52:05,616 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 12:52:07,163 INFO [train.py:901] (0/2) Epoch 47, batch 2300, loss[loss=0.1238, simple_loss=0.207, pruned_loss=0.02025, over 7282.00 frames. ], tot_loss[loss=0.1264, simple_loss=0.2079, pruned_loss=0.02249, over 1437218.19 frames. ], batch size: 57, lr: 3.50e-03, grad_scale: 8.0 +2023-03-21 12:52:11,246 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=132213.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:52:32,369 INFO [train.py:901] (0/2) Epoch 47, batch 2350, loss[loss=0.122, simple_loss=0.2054, pruned_loss=0.01924, over 7307.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2087, pruned_loss=0.02275, over 1440913.78 frames. ], batch size: 49, lr: 3.50e-03, grad_scale: 8.0 +2023-03-21 12:52:36,113 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8551, 2.7174, 3.0925, 3.0546, 2.9336, 2.7567, 3.0802, 2.2210], + device='cuda:0'), covar=tensor([0.0568, 0.0724, 0.1019, 0.0849, 0.0851, 0.1254, 0.0763, 0.3370], + device='cuda:0'), in_proj_covar=tensor([0.0329, 0.0338, 0.0269, 0.0348, 0.0278, 0.0286, 0.0342, 0.0236], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 12:52:37,446 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.200e+02 1.770e+02 2.102e+02 2.451e+02 5.297e+02, threshold=4.204e+02, percent-clipped=1.0 +2023-03-21 12:52:52,500 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 12:52:58,545 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 12:52:59,022 INFO [train.py:901] (0/2) Epoch 47, batch 2400, loss[loss=0.1222, simple_loss=0.2108, pruned_loss=0.01684, over 7132.00 frames. ], tot_loss[loss=0.1268, simple_loss=0.2087, pruned_loss=0.02251, over 1440776.16 frames. ], batch size: 98, lr: 3.49e-03, grad_scale: 8.0 +2023-03-21 12:53:09,180 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 12:53:12,222 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 12:53:21,523 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6454, 2.0075, 1.5643, 2.0039, 2.0352, 1.8316, 1.7547, 1.3966], + device='cuda:0'), covar=tensor([0.0311, 0.0212, 0.0367, 0.0242, 0.0204, 0.0203, 0.0278, 0.0301], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0041, 0.0039, 0.0041, 0.0038, 0.0037, 0.0040, 0.0049], + device='cuda:0'), out_proj_covar=tensor([4.6384e-05, 4.5262e-05, 4.3734e-05, 4.4763e-05, 4.2159e-05, 4.1481e-05, + 4.4390e-05, 5.3776e-05], device='cuda:0') +2023-03-21 12:53:24,314 INFO [train.py:901] (0/2) Epoch 47, batch 2450, loss[loss=0.153, simple_loss=0.2303, pruned_loss=0.03786, over 7326.00 frames. ], tot_loss[loss=0.1269, simple_loss=0.2085, pruned_loss=0.02263, over 1439368.57 frames. ], batch size: 59, lr: 3.49e-03, grad_scale: 8.0 +2023-03-21 12:53:29,371 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.194e+02 1.852e+02 2.168e+02 2.561e+02 3.828e+02, threshold=4.336e+02, percent-clipped=0.0 +2023-03-21 12:53:39,246 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 12:53:41,734 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=132387.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:53:51,203 INFO [train.py:901] (0/2) Epoch 47, batch 2500, loss[loss=0.1268, simple_loss=0.2141, pruned_loss=0.01972, over 7267.00 frames. ], tot_loss[loss=0.1269, simple_loss=0.2085, pruned_loss=0.02263, over 1439588.46 frames. ], batch size: 52, lr: 3.49e-03, grad_scale: 8.0 +2023-03-21 12:53:53,395 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0255, 3.8830, 2.9830, 3.5994, 2.8846, 2.2161, 1.6836, 3.9947], + device='cuda:0'), covar=tensor([0.0052, 0.0061, 0.0198, 0.0089, 0.0217, 0.0637, 0.0789, 0.0060], + device='cuda:0'), in_proj_covar=tensor([0.0122, 0.0097, 0.0120, 0.0103, 0.0138, 0.0139, 0.0133, 0.0110], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 12:53:53,833 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=132410.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 12:54:03,696 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.22 vs. limit=2.0 +2023-03-21 12:54:04,883 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 12:54:17,080 INFO [train.py:901] (0/2) Epoch 47, batch 2550, loss[loss=0.1199, simple_loss=0.2028, pruned_loss=0.01849, over 7323.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2088, pruned_loss=0.02267, over 1442360.61 frames. ], batch size: 83, lr: 3.49e-03, grad_scale: 8.0 +2023-03-21 12:54:22,810 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.239e+02 1.683e+02 1.964e+02 2.400e+02 4.092e+02, threshold=3.929e+02, percent-clipped=0.0 +2023-03-21 12:54:43,030 INFO [train.py:901] (0/2) Epoch 47, batch 2600, loss[loss=0.162, simple_loss=0.2421, pruned_loss=0.04095, over 6813.00 frames. ], tot_loss[loss=0.1269, simple_loss=0.2083, pruned_loss=0.02274, over 1441796.11 frames. ], batch size: 107, lr: 3.49e-03, grad_scale: 8.0 +2023-03-21 12:54:47,106 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=132513.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:55:08,063 INFO [train.py:901] (0/2) Epoch 47, batch 2650, loss[loss=0.1343, simple_loss=0.2177, pruned_loss=0.02549, over 7255.00 frames. ], tot_loss[loss=0.127, simple_loss=0.2083, pruned_loss=0.02283, over 1443018.03 frames. ], batch size: 55, lr: 3.49e-03, grad_scale: 8.0 +2023-03-21 12:55:11,116 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=132561.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:55:13,048 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.259e+02 1.743e+02 2.078e+02 2.512e+02 4.321e+02, threshold=4.155e+02, percent-clipped=2.0 +2023-03-21 12:55:25,632 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.78 vs. limit=2.0 +2023-03-21 12:55:33,196 INFO [train.py:901] (0/2) Epoch 47, batch 2700, loss[loss=0.1481, simple_loss=0.2255, pruned_loss=0.03529, over 7269.00 frames. ], tot_loss[loss=0.1273, simple_loss=0.2085, pruned_loss=0.02303, over 1442413.34 frames. ], batch size: 70, lr: 3.49e-03, grad_scale: 8.0 +2023-03-21 12:55:44,298 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2427, 3.1763, 3.6950, 3.2835, 3.6294, 3.2061, 2.9240, 3.4827], + device='cuda:0'), covar=tensor([0.1575, 0.0588, 0.0675, 0.0943, 0.0824, 0.1013, 0.1668, 0.0903], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0073, 0.0055, 0.0054, 0.0053, 0.0052, 0.0072, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 12:55:59,153 INFO [train.py:901] (0/2) Epoch 47, batch 2750, loss[loss=0.1457, simple_loss=0.2194, pruned_loss=0.03601, over 7196.00 frames. ], tot_loss[loss=0.127, simple_loss=0.2086, pruned_loss=0.02274, over 1443011.90 frames. ], batch size: 50, lr: 3.49e-03, grad_scale: 8.0 +2023-03-21 12:56:01,818 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=132660.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:56:04,147 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.232e+02 1.754e+02 2.060e+02 2.369e+02 4.641e+02, threshold=4.120e+02, percent-clipped=1.0 +2023-03-21 12:56:14,006 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.91 vs. limit=2.0 +2023-03-21 12:56:15,180 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=132687.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:56:23,843 INFO [train.py:901] (0/2) Epoch 47, batch 2800, loss[loss=0.1358, simple_loss=0.2267, pruned_loss=0.02242, over 6675.00 frames. ], tot_loss[loss=0.1273, simple_loss=0.2087, pruned_loss=0.02288, over 1441264.97 frames. ], batch size: 107, lr: 3.49e-03, grad_scale: 8.0 +2023-03-21 12:56:26,418 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=132710.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 12:56:31,927 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=132721.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:56:36,594 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-47.pt +2023-03-21 12:56:49,379 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 12:56:49,437 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 12:56:49,465 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 12:56:49,485 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 12:56:49,673 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 12:56:49,690 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 12:56:49,702 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 12:56:49,810 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 12:56:49,890 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 12:56:49,907 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 12:56:49,925 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 12:56:50,098 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 12:56:50,103 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 12:56:50,111 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 12:56:50,640 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 12:56:50,643 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 12:56:50,707 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 12:56:53,004 INFO [train.py:901] (0/2) Epoch 48, batch 0, loss[loss=0.1283, simple_loss=0.2198, pruned_loss=0.01837, over 7301.00 frames. ], tot_loss[loss=0.1283, simple_loss=0.2198, pruned_loss=0.01837, over 7301.00 frames. ], batch size: 68, lr: 3.45e-03, grad_scale: 8.0 +2023-03-21 12:56:53,006 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 12:57:19,428 INFO [train.py:935] (0/2) Epoch 48, validation: loss=0.1658, simple_loss=0.2588, pruned_loss=0.03638, over 1622729.00 frames. +2023-03-21 12:57:19,429 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 12:57:22,410 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=132735.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:57:25,881 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 12:57:28,054 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=132746.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:57:34,016 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=132758.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:57:36,968 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 12:57:37,424 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.399e+02 1.668e+02 1.991e+02 2.294e+02 3.667e+02, threshold=3.982e+02, percent-clipped=0.0 +2023-03-21 12:57:44,520 INFO [train.py:901] (0/2) Epoch 48, batch 50, loss[loss=0.1213, simple_loss=0.2004, pruned_loss=0.02117, over 7354.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2092, pruned_loss=0.02252, over 324538.57 frames. ], batch size: 73, lr: 3.45e-03, grad_scale: 8.0 +2023-03-21 12:57:44,540 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 12:57:47,108 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 12:57:49,652 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 12:57:57,249 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2925, 3.4822, 2.5082, 3.8350, 2.9366, 3.2620, 1.7185, 2.6983], + device='cuda:0'), covar=tensor([0.0478, 0.0901, 0.2635, 0.0564, 0.0543, 0.0554, 0.4401, 0.1921], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0255, 0.0275, 0.0268, 0.0264, 0.0260, 0.0228, 0.0254], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 12:57:59,264 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=132807.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:58:06,752 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 12:58:07,739 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 12:58:10,805 INFO [train.py:901] (0/2) Epoch 48, batch 100, loss[loss=0.1326, simple_loss=0.2185, pruned_loss=0.02341, over 7351.00 frames. ], tot_loss[loss=0.1272, simple_loss=0.2089, pruned_loss=0.0227, over 573554.38 frames. ], batch size: 63, lr: 3.45e-03, grad_scale: 8.0 +2023-03-21 12:58:28,999 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.188e+02 1.757e+02 1.926e+02 2.244e+02 3.383e+02, threshold=3.853e+02, percent-clipped=0.0 +2023-03-21 12:58:36,324 INFO [train.py:901] (0/2) Epoch 48, batch 150, loss[loss=0.1373, simple_loss=0.2241, pruned_loss=0.02526, over 6559.00 frames. ], tot_loss[loss=0.1266, simple_loss=0.2088, pruned_loss=0.02224, over 764840.37 frames. ], batch size: 106, lr: 3.45e-03, grad_scale: 8.0 +2023-03-21 12:58:39,460 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9590, 5.3960, 5.4983, 5.4266, 5.2485, 4.9649, 5.5238, 5.3151], + device='cuda:0'), covar=tensor([0.0343, 0.0333, 0.0358, 0.0400, 0.0282, 0.0341, 0.0297, 0.0373], + device='cuda:0'), in_proj_covar=tensor([0.0221, 0.0271, 0.0211, 0.0210, 0.0162, 0.0239, 0.0218, 0.0151], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 12:58:51,213 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9150, 3.8933, 2.9919, 3.5138, 2.8961, 2.1320, 1.7852, 3.8652], + device='cuda:0'), covar=tensor([0.0070, 0.0067, 0.0263, 0.0111, 0.0291, 0.0797, 0.0863, 0.0086], + device='cuda:0'), in_proj_covar=tensor([0.0122, 0.0097, 0.0120, 0.0102, 0.0138, 0.0138, 0.0134, 0.0109], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 12:59:02,772 INFO [train.py:901] (0/2) Epoch 48, batch 200, loss[loss=0.1199, simple_loss=0.2038, pruned_loss=0.01799, over 7314.00 frames. ], tot_loss[loss=0.1268, simple_loss=0.2085, pruned_loss=0.02253, over 915687.29 frames. ], batch size: 59, lr: 3.45e-03, grad_scale: 8.0 +2023-03-21 12:59:08,987 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2427, 4.3190, 4.1408, 4.2981, 3.9630, 4.3245, 4.6168, 4.6268], + device='cuda:0'), covar=tensor([0.0172, 0.0124, 0.0193, 0.0161, 0.0301, 0.0210, 0.0174, 0.0156], + device='cuda:0'), in_proj_covar=tensor([0.0126, 0.0129, 0.0123, 0.0128, 0.0116, 0.0103, 0.0101, 0.0104], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 12:59:09,926 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 12:59:13,912 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 12:59:20,380 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 12:59:20,851 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.260e+02 1.731e+02 2.035e+02 2.345e+02 3.653e+02, threshold=4.069e+02, percent-clipped=0.0 +2023-03-21 12:59:27,891 INFO [train.py:901] (0/2) Epoch 48, batch 250, loss[loss=0.1342, simple_loss=0.2121, pruned_loss=0.0282, over 7300.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2085, pruned_loss=0.02289, over 1032463.52 frames. ], batch size: 66, lr: 3.45e-03, grad_scale: 8.0 +2023-03-21 12:59:32,917 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 12:59:48,094 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=133016.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 12:59:48,812 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.70 vs. limit=2.0 +2023-03-21 12:59:54,515 INFO [train.py:901] (0/2) Epoch 48, batch 300, loss[loss=0.1194, simple_loss=0.2009, pruned_loss=0.01894, over 7344.00 frames. ], tot_loss[loss=0.1267, simple_loss=0.2083, pruned_loss=0.02258, over 1124397.00 frames. ], batch size: 63, lr: 3.45e-03, grad_scale: 8.0 +2023-03-21 12:59:54,523 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 12:59:56,185 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9769, 2.9263, 3.1548, 3.1185, 2.8891, 2.7617, 3.2040, 2.3407], + device='cuda:0'), covar=tensor([0.0456, 0.0686, 0.0819, 0.0773, 0.0775, 0.1177, 0.0640, 0.2832], + device='cuda:0'), in_proj_covar=tensor([0.0328, 0.0337, 0.0269, 0.0349, 0.0279, 0.0286, 0.0343, 0.0237], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:00:03,630 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 13:00:12,663 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.276e+02 1.775e+02 2.154e+02 2.496e+02 3.947e+02, threshold=4.307e+02, percent-clipped=0.0 +2023-03-21 13:00:14,034 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.22 vs. limit=2.0 +2023-03-21 13:00:19,801 INFO [train.py:901] (0/2) Epoch 48, batch 350, loss[loss=0.1129, simple_loss=0.1932, pruned_loss=0.01631, over 7344.00 frames. ], tot_loss[loss=0.1261, simple_loss=0.2074, pruned_loss=0.02243, over 1194501.68 frames. ], batch size: 44, lr: 3.45e-03, grad_scale: 8.0 +2023-03-21 13:00:19,872 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8848, 4.2818, 4.1162, 4.8520, 4.6361, 4.6751, 4.1538, 4.2779], + device='cuda:0'), covar=tensor([0.0879, 0.2795, 0.2654, 0.1093, 0.0976, 0.1167, 0.0869, 0.1124], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0407, 0.0308, 0.0319, 0.0237, 0.0381, 0.0241, 0.0289], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 13:00:32,667 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=133102.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:00:37,626 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 13:00:46,192 INFO [train.py:901] (0/2) Epoch 48, batch 400, loss[loss=0.09477, simple_loss=0.1612, pruned_loss=0.01419, over 5915.00 frames. ], tot_loss[loss=0.1264, simple_loss=0.2078, pruned_loss=0.02253, over 1246181.83 frames. ], batch size: 26, lr: 3.45e-03, grad_scale: 8.0 +2023-03-21 13:01:04,079 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.252e+02 1.745e+02 2.126e+02 2.539e+02 4.672e+02, threshold=4.251e+02, percent-clipped=1.0 +2023-03-21 13:01:12,432 INFO [train.py:901] (0/2) Epoch 48, batch 450, loss[loss=0.1284, simple_loss=0.2157, pruned_loss=0.02053, over 7258.00 frames. ], tot_loss[loss=0.1263, simple_loss=0.2076, pruned_loss=0.0225, over 1289485.78 frames. ], batch size: 64, lr: 3.45e-03, grad_scale: 8.0 +2023-03-21 13:01:19,489 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 13:01:19,979 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 13:01:35,635 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7433, 3.8412, 3.6665, 3.8679, 3.5561, 3.7764, 4.0920, 4.0608], + device='cuda:0'), covar=tensor([0.0213, 0.0192, 0.0255, 0.0174, 0.0351, 0.0342, 0.0234, 0.0233], + device='cuda:0'), in_proj_covar=tensor([0.0126, 0.0129, 0.0123, 0.0127, 0.0115, 0.0102, 0.0100, 0.0103], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 13:01:38,010 INFO [train.py:901] (0/2) Epoch 48, batch 500, loss[loss=0.1289, simple_loss=0.2076, pruned_loss=0.02509, over 7353.00 frames. ], tot_loss[loss=0.1259, simple_loss=0.2072, pruned_loss=0.02233, over 1324531.15 frames. ], batch size: 44, lr: 3.45e-03, grad_scale: 8.0 +2023-03-21 13:01:53,335 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 13:01:54,870 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 13:01:55,368 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 13:01:57,516 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.202e+02 1.728e+02 2.008e+02 2.239e+02 3.859e+02, threshold=4.016e+02, percent-clipped=0.0 +2023-03-21 13:01:58,069 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 13:01:59,437 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-03-21 13:02:02,712 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 13:02:03,632 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.88 vs. limit=5.0 +2023-03-21 13:02:04,696 INFO [train.py:901] (0/2) Epoch 48, batch 550, loss[loss=0.1357, simple_loss=0.221, pruned_loss=0.0252, over 7283.00 frames. ], tot_loss[loss=0.1265, simple_loss=0.2081, pruned_loss=0.02248, over 1353861.55 frames. ], batch size: 66, lr: 3.45e-03, grad_scale: 8.0 +2023-03-21 13:02:14,818 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 13:02:23,479 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 13:02:23,572 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=133316.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:02:26,590 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 13:02:30,056 INFO [train.py:901] (0/2) Epoch 48, batch 600, loss[loss=0.1498, simple_loss=0.2316, pruned_loss=0.034, over 7286.00 frames. ], tot_loss[loss=0.126, simple_loss=0.2074, pruned_loss=0.02232, over 1374236.82 frames. ], batch size: 68, lr: 3.44e-03, grad_scale: 8.0 +2023-03-21 13:02:33,136 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 13:02:40,465 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.15 vs. limit=5.0 +2023-03-21 13:02:48,137 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.21 vs. limit=2.0 +2023-03-21 13:02:48,948 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=133364.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:02:49,398 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.220e+02 1.711e+02 1.911e+02 2.129e+02 5.937e+02, threshold=3.823e+02, percent-clipped=2.0 +2023-03-21 13:02:50,430 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 13:02:54,105 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=133374.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 13:02:55,124 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=133376.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:02:56,504 INFO [train.py:901] (0/2) Epoch 48, batch 650, loss[loss=0.1338, simple_loss=0.2149, pruned_loss=0.02632, over 7296.00 frames. ], tot_loss[loss=0.1254, simple_loss=0.2067, pruned_loss=0.02202, over 1389392.22 frames. ], batch size: 68, lr: 3.44e-03, grad_scale: 8.0 +2023-03-21 13:02:59,605 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 13:03:08,678 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=133402.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:03:12,810 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7032, 3.5292, 3.5344, 3.3989, 3.5361, 3.3198, 3.6659, 3.3856], + device='cuda:0'), covar=tensor([0.0164, 0.0204, 0.0113, 0.0246, 0.0369, 0.0115, 0.0140, 0.0140], + device='cuda:0'), in_proj_covar=tensor([0.0101, 0.0106, 0.0108, 0.0094, 0.0184, 0.0114, 0.0110, 0.0118], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 13:03:16,805 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 13:03:22,960 INFO [train.py:901] (0/2) Epoch 48, batch 700, loss[loss=0.1362, simple_loss=0.2191, pruned_loss=0.02659, over 7169.00 frames. ], tot_loss[loss=0.1262, simple_loss=0.2077, pruned_loss=0.02239, over 1400901.29 frames. ], batch size: 98, lr: 3.44e-03, grad_scale: 8.0 +2023-03-21 13:03:26,281 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 13:03:26,418 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=133435.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 13:03:27,395 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=133437.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:03:34,488 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=133450.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:03:41,876 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.129e+02 1.729e+02 1.961e+02 2.406e+02 3.569e+02, threshold=3.922e+02, percent-clipped=0.0 +2023-03-21 13:03:46,626 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=133474.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:03:48,115 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9615, 4.1054, 3.9722, 4.1449, 3.7312, 4.0649, 4.3747, 4.3621], + device='cuda:0'), covar=tensor([0.0222, 0.0188, 0.0216, 0.0175, 0.0373, 0.0357, 0.0235, 0.0196], + device='cuda:0'), in_proj_covar=tensor([0.0127, 0.0130, 0.0124, 0.0127, 0.0116, 0.0103, 0.0101, 0.0104], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 13:03:49,050 INFO [train.py:901] (0/2) Epoch 48, batch 750, loss[loss=0.118, simple_loss=0.2028, pruned_loss=0.01658, over 7267.00 frames. ], tot_loss[loss=0.1262, simple_loss=0.2076, pruned_loss=0.02237, over 1412046.29 frames. ], batch size: 70, lr: 3.44e-03, grad_scale: 8.0 +2023-03-21 13:03:50,048 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 13:03:50,550 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 13:04:03,492 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 13:04:07,851 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6119, 1.9935, 1.6817, 1.9304, 1.9539, 1.8310, 1.5966, 1.5060], + device='cuda:0'), covar=tensor([0.0255, 0.0204, 0.0344, 0.0251, 0.0183, 0.0203, 0.0278, 0.0330], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0041, 0.0039, 0.0041, 0.0039, 0.0038, 0.0040, 0.0049], + device='cuda:0'), out_proj_covar=tensor([4.6892e-05, 4.5066e-05, 4.3911e-05, 4.5130e-05, 4.2669e-05, 4.1614e-05, + 4.4936e-05, 5.4182e-05], device='cuda:0') +2023-03-21 13:04:08,490 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.81 vs. limit=2.0 +2023-03-21 13:04:09,247 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 13:04:15,414 INFO [train.py:901] (0/2) Epoch 48, batch 800, loss[loss=0.1236, simple_loss=0.2031, pruned_loss=0.02199, over 7337.00 frames. ], tot_loss[loss=0.126, simple_loss=0.2072, pruned_loss=0.02236, over 1416869.08 frames. ], batch size: 54, lr: 3.44e-03, grad_scale: 8.0 +2023-03-21 13:04:15,946 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 13:04:16,929 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 13:04:18,575 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=133535.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:04:27,402 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 13:04:33,574 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.190e+02 1.718e+02 2.033e+02 2.465e+02 3.831e+02, threshold=4.067e+02, percent-clipped=0.0 +2023-03-21 13:04:40,720 INFO [train.py:901] (0/2) Epoch 48, batch 850, loss[loss=0.09731, simple_loss=0.161, pruned_loss=0.01681, over 5872.00 frames. ], tot_loss[loss=0.1258, simple_loss=0.2071, pruned_loss=0.0222, over 1423722.75 frames. ], batch size: 25, lr: 3.44e-03, grad_scale: 8.0 +2023-03-21 13:04:46,370 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 13:04:46,850 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 13:04:52,855 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 13:04:56,321 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 13:05:06,318 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.94 vs. limit=2.0 +2023-03-21 13:05:07,529 INFO [train.py:901] (0/2) Epoch 48, batch 900, loss[loss=0.1313, simple_loss=0.2212, pruned_loss=0.0207, over 7316.00 frames. ], tot_loss[loss=0.1266, simple_loss=0.208, pruned_loss=0.02257, over 1429733.69 frames. ], batch size: 59, lr: 3.44e-03, grad_scale: 8.0 +2023-03-21 13:05:19,376 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=133652.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:05:25,802 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.155e+02 1.711e+02 1.987e+02 2.449e+02 8.050e+02, threshold=3.974e+02, percent-clipped=6.0 +2023-03-21 13:05:32,849 INFO [train.py:901] (0/2) Epoch 48, batch 950, loss[loss=0.1296, simple_loss=0.2168, pruned_loss=0.0212, over 7274.00 frames. ], tot_loss[loss=0.1267, simple_loss=0.2081, pruned_loss=0.02266, over 1433506.18 frames. ], batch size: 52, lr: 3.44e-03, grad_scale: 8.0 +2023-03-21 13:05:35,451 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 13:05:50,319 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.52 vs. limit=2.0 +2023-03-21 13:05:51,672 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=133713.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:05:59,180 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 13:05:59,683 INFO [train.py:901] (0/2) Epoch 48, batch 1000, loss[loss=0.1206, simple_loss=0.202, pruned_loss=0.01956, over 7292.00 frames. ], tot_loss[loss=0.1264, simple_loss=0.2078, pruned_loss=0.02251, over 1433599.22 frames. ], batch size: 70, lr: 3.44e-03, grad_scale: 8.0 +2023-03-21 13:06:00,287 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=133730.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 13:06:01,344 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=133732.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:06:17,969 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.135e+02 1.525e+02 1.854e+02 2.105e+02 4.157e+02, threshold=3.709e+02, percent-clipped=1.0 +2023-03-21 13:06:18,524 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 13:06:25,680 INFO [train.py:901] (0/2) Epoch 48, batch 1050, loss[loss=0.1353, simple_loss=0.217, pruned_loss=0.02677, over 7216.00 frames. ], tot_loss[loss=0.1264, simple_loss=0.2081, pruned_loss=0.02235, over 1437529.35 frames. ], batch size: 93, lr: 3.44e-03, grad_scale: 8.0 +2023-03-21 13:06:28,228 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.99 vs. limit=5.0 +2023-03-21 13:06:29,954 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2157, 2.7531, 3.3559, 2.9918, 3.2723, 3.0677, 2.7118, 3.2038], + device='cuda:0'), covar=tensor([0.0867, 0.0735, 0.0874, 0.1433, 0.0597, 0.0905, 0.1720, 0.0907], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0073, 0.0055, 0.0054, 0.0053, 0.0052, 0.0072, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 13:06:41,708 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 13:06:46,189 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 13:06:51,717 INFO [train.py:901] (0/2) Epoch 48, batch 1100, loss[loss=0.1302, simple_loss=0.2138, pruned_loss=0.02332, over 7263.00 frames. ], tot_loss[loss=0.1264, simple_loss=0.2082, pruned_loss=0.02234, over 1437166.61 frames. ], batch size: 77, lr: 3.44e-03, grad_scale: 8.0 +2023-03-21 13:06:52,296 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=133830.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:06:53,375 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7604, 2.3123, 3.0274, 2.7318, 2.8776, 2.7067, 2.2772, 2.9418], + device='cuda:0'), covar=tensor([0.1191, 0.1087, 0.0964, 0.1354, 0.0689, 0.1171, 0.2136, 0.1072], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0073, 0.0055, 0.0054, 0.0054, 0.0053, 0.0073, 0.0054], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 13:07:11,123 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.249e+02 1.881e+02 2.184e+02 2.691e+02 4.315e+02, threshold=4.367e+02, percent-clipped=4.0 +2023-03-21 13:07:13,781 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0814, 3.9135, 3.0986, 3.6359, 3.1037, 2.2478, 1.7994, 4.0136], + device='cuda:0'), covar=tensor([0.0052, 0.0069, 0.0189, 0.0081, 0.0189, 0.0729, 0.0771, 0.0066], + device='cuda:0'), in_proj_covar=tensor([0.0122, 0.0097, 0.0119, 0.0101, 0.0138, 0.0138, 0.0134, 0.0110], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 13:07:14,627 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 13:07:15,176 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 13:07:18,213 INFO [train.py:901] (0/2) Epoch 48, batch 1150, loss[loss=0.1423, simple_loss=0.2261, pruned_loss=0.02924, over 7292.00 frames. ], tot_loss[loss=0.1269, simple_loss=0.2087, pruned_loss=0.02254, over 1439093.65 frames. ], batch size: 77, lr: 3.44e-03, grad_scale: 16.0 +2023-03-21 13:07:22,377 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3744, 3.2259, 3.1177, 3.3176, 3.0968, 2.7844, 3.5750, 2.2668], + device='cuda:0'), covar=tensor([0.0617, 0.0746, 0.0985, 0.0861, 0.0927, 0.1285, 0.0736, 0.3373], + device='cuda:0'), in_proj_covar=tensor([0.0327, 0.0335, 0.0267, 0.0346, 0.0277, 0.0284, 0.0342, 0.0235], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:07:26,698 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 13:07:27,687 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 13:07:39,601 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1707, 3.3650, 2.4407, 3.7602, 2.9027, 3.1913, 1.8548, 2.7255], + device='cuda:0'), covar=tensor([0.0456, 0.0756, 0.2756, 0.0548, 0.0453, 0.0528, 0.4056, 0.1889], + device='cuda:0'), in_proj_covar=tensor([0.0271, 0.0252, 0.0273, 0.0266, 0.0262, 0.0259, 0.0225, 0.0252], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:07:43,073 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6243, 1.6073, 1.5576, 1.9080, 1.7192, 1.9970, 1.5094, 1.9202], + device='cuda:0'), covar=tensor([0.2141, 0.3372, 0.2339, 0.1069, 0.1779, 0.1162, 0.2336, 0.1930], + device='cuda:0'), in_proj_covar=tensor([0.0085, 0.0086, 0.0080, 0.0069, 0.0070, 0.0070, 0.0110, 0.0071], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:07:43,419 INFO [train.py:901] (0/2) Epoch 48, batch 1200, loss[loss=0.1403, simple_loss=0.2185, pruned_loss=0.03103, over 6651.00 frames. ], tot_loss[loss=0.1266, simple_loss=0.2086, pruned_loss=0.02229, over 1441886.78 frames. ], batch size: 106, lr: 3.44e-03, grad_scale: 16.0 +2023-03-21 13:07:53,252 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1889, 3.0756, 3.0192, 3.2234, 3.0599, 2.7844, 3.3862, 2.2349], + device='cuda:0'), covar=tensor([0.0512, 0.0632, 0.0975, 0.0813, 0.0850, 0.1190, 0.0786, 0.3166], + device='cuda:0'), in_proj_covar=tensor([0.0325, 0.0333, 0.0266, 0.0344, 0.0276, 0.0282, 0.0340, 0.0233], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:08:02,804 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 13:08:03,293 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.228e+02 1.806e+02 2.192e+02 2.530e+02 3.927e+02, threshold=4.384e+02, percent-clipped=0.0 +2023-03-21 13:08:09,887 INFO [train.py:901] (0/2) Epoch 48, batch 1250, loss[loss=0.1127, simple_loss=0.1924, pruned_loss=0.01649, over 7137.00 frames. ], tot_loss[loss=0.1261, simple_loss=0.2079, pruned_loss=0.0222, over 1441298.96 frames. ], batch size: 41, lr: 3.44e-03, grad_scale: 8.0 +2023-03-21 13:08:19,034 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=133997.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:08:24,931 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=134008.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:08:26,376 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 13:08:30,990 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 13:08:32,029 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 13:08:36,306 INFO [train.py:901] (0/2) Epoch 48, batch 1300, loss[loss=0.1486, simple_loss=0.235, pruned_loss=0.03108, over 6755.00 frames. ], tot_loss[loss=0.1256, simple_loss=0.2071, pruned_loss=0.02207, over 1440478.83 frames. ], batch size: 107, lr: 3.44e-03, grad_scale: 8.0 +2023-03-21 13:08:36,947 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=134030.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 13:08:37,996 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=134032.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:08:51,814 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=134058.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:08:55,690 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.358e+02 1.728e+02 2.012e+02 2.304e+02 3.451e+02, threshold=4.025e+02, percent-clipped=0.0 +2023-03-21 13:08:56,226 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 13:08:58,299 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 13:09:01,388 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0212, 3.5377, 4.1994, 4.2220, 4.2235, 4.1447, 4.4090, 4.1785], + device='cuda:0'), covar=tensor([0.0037, 0.0112, 0.0029, 0.0026, 0.0027, 0.0030, 0.0028, 0.0040], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0075, 0.0062, 0.0060, 0.0057, 0.0063, 0.0049, 0.0082], + device='cuda:0'), out_proj_covar=tensor([8.6169e-05, 1.4687e-04, 1.0827e-04, 9.9342e-05, 9.2447e-05, 1.0607e-04, + 9.1096e-05, 1.4703e-04], device='cuda:0') +2023-03-21 13:09:01,802 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 13:09:01,854 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=134078.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 13:09:02,248 INFO [train.py:901] (0/2) Epoch 48, batch 1350, loss[loss=0.153, simple_loss=0.2251, pruned_loss=0.04041, over 7271.00 frames. ], tot_loss[loss=0.1257, simple_loss=0.2072, pruned_loss=0.02214, over 1440479.53 frames. ], batch size: 52, lr: 3.43e-03, grad_scale: 8.0 +2023-03-21 13:09:02,818 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=134080.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:09:11,545 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 13:09:24,165 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.65 vs. limit=2.0 +2023-03-21 13:09:28,302 INFO [train.py:901] (0/2) Epoch 48, batch 1400, loss[loss=0.128, simple_loss=0.2092, pruned_loss=0.02341, over 7344.00 frames. ], tot_loss[loss=0.1263, simple_loss=0.2079, pruned_loss=0.02234, over 1441234.43 frames. ], batch size: 75, lr: 3.43e-03, grad_scale: 8.0 +2023-03-21 13:09:29,570 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=134130.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:09:41,865 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9918, 4.4056, 4.4136, 4.3616, 4.3650, 4.0284, 4.4519, 4.3306], + device='cuda:0'), covar=tensor([0.0479, 0.0416, 0.0408, 0.0547, 0.0381, 0.0495, 0.0374, 0.0439], + device='cuda:0'), in_proj_covar=tensor([0.0225, 0.0275, 0.0214, 0.0212, 0.0164, 0.0244, 0.0223, 0.0154], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:09:44,778 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 13:09:46,924 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7020, 3.7981, 3.5830, 3.7639, 3.4814, 3.6112, 4.0798, 4.0842], + device='cuda:0'), covar=tensor([0.0239, 0.0186, 0.0254, 0.0190, 0.0331, 0.0442, 0.0231, 0.0193], + device='cuda:0'), in_proj_covar=tensor([0.0129, 0.0132, 0.0127, 0.0129, 0.0117, 0.0105, 0.0102, 0.0106], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 13:09:47,823 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.303e+02 1.834e+02 2.048e+02 2.397e+02 3.522e+02, threshold=4.097e+02, percent-clipped=0.0 +2023-03-21 13:09:54,038 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=134178.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:09:54,483 INFO [train.py:901] (0/2) Epoch 48, batch 1450, loss[loss=0.1134, simple_loss=0.2006, pruned_loss=0.01313, over 7275.00 frames. ], tot_loss[loss=0.1258, simple_loss=0.2071, pruned_loss=0.0222, over 1439511.55 frames. ], batch size: 70, lr: 3.43e-03, grad_scale: 8.0 +2023-03-21 13:09:56,908 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.90 vs. limit=5.0 +2023-03-21 13:10:01,228 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2184, 2.1258, 2.1865, 3.2398, 1.7652, 3.1653, 1.3334, 3.0311], + device='cuda:0'), covar=tensor([0.0276, 0.1571, 0.2056, 0.0317, 0.4319, 0.0349, 0.1584, 0.0520], + device='cuda:0'), in_proj_covar=tensor([0.0207, 0.0239, 0.0252, 0.0210, 0.0245, 0.0219, 0.0216, 0.0228], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:10:09,708 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 13:10:15,241 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.16 vs. limit=5.0 +2023-03-21 13:10:21,518 INFO [train.py:901] (0/2) Epoch 48, batch 1500, loss[loss=0.1223, simple_loss=0.2037, pruned_loss=0.02039, over 7274.00 frames. ], tot_loss[loss=0.1255, simple_loss=0.2071, pruned_loss=0.02194, over 1441361.00 frames. ], batch size: 47, lr: 3.43e-03, grad_scale: 8.0 +2023-03-21 13:10:27,074 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 13:10:27,161 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2144, 4.3212, 4.1150, 4.2636, 3.9418, 4.1923, 4.6530, 4.6816], + device='cuda:0'), covar=tensor([0.0201, 0.0160, 0.0204, 0.0171, 0.0324, 0.0400, 0.0202, 0.0162], + device='cuda:0'), in_proj_covar=tensor([0.0129, 0.0131, 0.0126, 0.0129, 0.0117, 0.0105, 0.0102, 0.0105], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 13:10:40,287 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.207e+02 1.705e+02 1.935e+02 2.252e+02 3.749e+02, threshold=3.870e+02, percent-clipped=0.0 +2023-03-21 13:10:44,198 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.60 vs. limit=2.0 +2023-03-21 13:10:46,854 INFO [train.py:901] (0/2) Epoch 48, batch 1550, loss[loss=0.1152, simple_loss=0.1973, pruned_loss=0.01658, over 7140.00 frames. ], tot_loss[loss=0.1257, simple_loss=0.2072, pruned_loss=0.02214, over 1441805.64 frames. ], batch size: 41, lr: 3.43e-03, grad_scale: 8.0 +2023-03-21 13:10:52,060 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 13:11:00,304 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=134304.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:11:03,029 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=134308.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:11:03,571 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=134309.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:11:13,629 INFO [train.py:901] (0/2) Epoch 48, batch 1600, loss[loss=0.1301, simple_loss=0.2242, pruned_loss=0.018, over 7233.00 frames. ], tot_loss[loss=0.126, simple_loss=0.2073, pruned_loss=0.02235, over 1440279.38 frames. ], batch size: 55, lr: 3.43e-03, grad_scale: 8.0 +2023-03-21 13:11:23,836 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 13:11:24,851 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 13:11:25,947 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=134353.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:11:27,413 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 13:11:27,451 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=134356.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:11:32,122 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=134365.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:11:32,448 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.266e+02 1.662e+02 1.962e+02 2.325e+02 5.415e+02, threshold=3.925e+02, percent-clipped=1.0 +2023-03-21 13:11:35,290 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=134370.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:11:37,245 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 13:11:39,767 INFO [train.py:901] (0/2) Epoch 48, batch 1650, loss[loss=0.1264, simple_loss=0.2031, pruned_loss=0.02486, over 7201.00 frames. ], tot_loss[loss=0.1255, simple_loss=0.2069, pruned_loss=0.02209, over 1442312.33 frames. ], batch size: 50, lr: 3.43e-03, grad_scale: 8.0 +2023-03-21 13:11:41,871 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 13:11:46,804 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.98 vs. limit=5.0 +2023-03-21 13:11:50,669 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 13:12:06,338 INFO [train.py:901] (0/2) Epoch 48, batch 1700, loss[loss=0.1422, simple_loss=0.2253, pruned_loss=0.02957, over 7363.00 frames. ], tot_loss[loss=0.1254, simple_loss=0.2067, pruned_loss=0.02204, over 1441819.27 frames. ], batch size: 63, lr: 3.43e-03, grad_scale: 8.0 +2023-03-21 13:12:07,900 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 13:12:11,950 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 13:12:23,055 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 13:12:25,575 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.287e+02 1.707e+02 1.940e+02 2.259e+02 3.650e+02, threshold=3.879e+02, percent-clipped=0.0 +2023-03-21 13:12:32,346 INFO [train.py:901] (0/2) Epoch 48, batch 1750, loss[loss=0.1206, simple_loss=0.2088, pruned_loss=0.01613, over 7275.00 frames. ], tot_loss[loss=0.1248, simple_loss=0.2062, pruned_loss=0.02175, over 1440595.99 frames. ], batch size: 70, lr: 3.43e-03, grad_scale: 8.0 +2023-03-21 13:12:48,236 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 13:12:48,772 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 13:12:53,875 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6320, 1.9398, 1.5754, 1.7883, 1.8835, 1.8293, 1.7561, 1.4363], + device='cuda:0'), covar=tensor([0.0207, 0.0175, 0.0496, 0.0220, 0.0210, 0.0159, 0.0260, 0.0228], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0041, 0.0039, 0.0041, 0.0039, 0.0038, 0.0040, 0.0050], + device='cuda:0'), out_proj_covar=tensor([4.7067e-05, 4.5604e-05, 4.4103e-05, 4.5319e-05, 4.3117e-05, 4.2232e-05, + 4.5112e-05, 5.4602e-05], device='cuda:0') +2023-03-21 13:12:58,289 INFO [train.py:901] (0/2) Epoch 48, batch 1800, loss[loss=0.1318, simple_loss=0.22, pruned_loss=0.02185, over 7321.00 frames. ], tot_loss[loss=0.125, simple_loss=0.2065, pruned_loss=0.02177, over 1441024.55 frames. ], batch size: 59, lr: 3.43e-03, grad_scale: 8.0 +2023-03-21 13:13:05,716 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.9026, 1.7981, 2.1399, 2.3380, 2.1548, 2.2757, 2.0616, 2.3361], + device='cuda:0'), covar=tensor([0.3706, 0.4392, 0.2981, 0.1584, 0.1170, 0.2731, 0.2517, 0.2918], + device='cuda:0'), in_proj_covar=tensor([0.0086, 0.0088, 0.0081, 0.0070, 0.0071, 0.0071, 0.0111, 0.0072], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:13:09,656 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 13:13:17,412 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3519, 4.0863, 3.3252, 3.8906, 3.1440, 2.2917, 1.8102, 4.2590], + device='cuda:0'), covar=tensor([0.0054, 0.0089, 0.0180, 0.0077, 0.0205, 0.0673, 0.0817, 0.0062], + device='cuda:0'), in_proj_covar=tensor([0.0121, 0.0096, 0.0119, 0.0101, 0.0137, 0.0136, 0.0134, 0.0109], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 13:13:17,787 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.203e+02 1.829e+02 2.119e+02 2.568e+02 4.020e+02, threshold=4.237e+02, percent-clipped=1.0 +2023-03-21 13:13:24,628 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 13:13:25,127 INFO [train.py:901] (0/2) Epoch 48, batch 1850, loss[loss=0.1287, simple_loss=0.2052, pruned_loss=0.02615, over 7360.00 frames. ], tot_loss[loss=0.1247, simple_loss=0.2059, pruned_loss=0.02177, over 1438258.69 frames. ], batch size: 51, lr: 3.43e-03, grad_scale: 8.0 +2023-03-21 13:13:34,372 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 13:13:51,704 INFO [train.py:901] (0/2) Epoch 48, batch 1900, loss[loss=0.114, simple_loss=0.2008, pruned_loss=0.01356, over 7310.00 frames. ], tot_loss[loss=0.1243, simple_loss=0.2056, pruned_loss=0.02152, over 1438597.30 frames. ], batch size: 86, lr: 3.43e-03, grad_scale: 8.0 +2023-03-21 13:13:51,710 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 13:13:52,567 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.91 vs. limit=2.0 +2023-03-21 13:14:04,109 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=134653.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:14:08,307 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=134660.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:14:10,853 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=134665.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:14:11,274 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.369e+02 1.733e+02 1.999e+02 2.386e+02 3.499e+02, threshold=3.997e+02, percent-clipped=0.0 +2023-03-21 13:14:16,447 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5736, 5.0210, 5.1152, 5.0549, 4.8807, 4.5490, 5.1313, 4.8621], + device='cuda:0'), covar=tensor([0.0437, 0.0407, 0.0418, 0.0487, 0.0351, 0.0440, 0.0342, 0.0511], + device='cuda:0'), in_proj_covar=tensor([0.0226, 0.0279, 0.0216, 0.0215, 0.0164, 0.0246, 0.0225, 0.0155], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:14:16,917 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 13:14:17,900 INFO [train.py:901] (0/2) Epoch 48, batch 1950, loss[loss=0.1193, simple_loss=0.1976, pruned_loss=0.02052, over 7319.00 frames. ], tot_loss[loss=0.1246, simple_loss=0.2058, pruned_loss=0.02171, over 1439847.82 frames. ], batch size: 59, lr: 3.43e-03, grad_scale: 8.0 +2023-03-21 13:14:28,143 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 13:14:28,968 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.46 vs. limit=2.0 +2023-03-21 13:14:29,193 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=134701.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:14:32,659 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 13:14:33,134 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 13:14:43,887 INFO [train.py:901] (0/2) Epoch 48, batch 2000, loss[loss=0.1155, simple_loss=0.1923, pruned_loss=0.01932, over 7148.00 frames. ], tot_loss[loss=0.1255, simple_loss=0.2071, pruned_loss=0.02195, over 1441079.14 frames. ], batch size: 39, lr: 3.43e-03, grad_scale: 8.0 +2023-03-21 13:14:45,749 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.76 vs. limit=2.0 +2023-03-21 13:14:49,687 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 13:15:00,382 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 13:15:03,348 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.176e+02 1.689e+02 2.059e+02 2.364e+02 4.863e+02, threshold=4.119e+02, percent-clipped=2.0 +2023-03-21 13:15:07,958 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 13:15:10,002 INFO [train.py:901] (0/2) Epoch 48, batch 2050, loss[loss=0.1214, simple_loss=0.2109, pruned_loss=0.01598, over 7136.00 frames. ], tot_loss[loss=0.1251, simple_loss=0.2068, pruned_loss=0.02173, over 1438430.11 frames. ], batch size: 98, lr: 3.43e-03, grad_scale: 8.0 +2023-03-21 13:15:12,133 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.1018, 4.6654, 4.4196, 5.0377, 4.8683, 4.9671, 4.2961, 4.6070], + device='cuda:0'), covar=tensor([0.0830, 0.2159, 0.2228, 0.1081, 0.0890, 0.1304, 0.0947, 0.1170], + device='cuda:0'), in_proj_covar=tensor([0.0273, 0.0404, 0.0305, 0.0320, 0.0236, 0.0380, 0.0238, 0.0288], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 13:15:13,831 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.15 vs. limit=2.0 +2023-03-21 13:15:36,860 INFO [train.py:901] (0/2) Epoch 48, batch 2100, loss[loss=0.1274, simple_loss=0.21, pruned_loss=0.02236, over 7260.00 frames. ], tot_loss[loss=0.1253, simple_loss=0.2069, pruned_loss=0.02184, over 1437407.23 frames. ], batch size: 89, lr: 3.43e-03, grad_scale: 8.0 +2023-03-21 13:15:41,995 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 13:15:45,103 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 13:15:53,986 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=134862.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:15:55,918 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.259e+02 1.812e+02 2.170e+02 2.535e+02 4.024e+02, threshold=4.341e+02, percent-clipped=0.0 +2023-03-21 13:15:57,055 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8292, 3.9123, 3.7270, 3.8574, 3.5603, 3.8374, 4.1940, 4.2081], + device='cuda:0'), covar=tensor([0.0236, 0.0190, 0.0277, 0.0213, 0.0332, 0.0379, 0.0260, 0.0212], + device='cuda:0'), in_proj_covar=tensor([0.0128, 0.0132, 0.0126, 0.0130, 0.0117, 0.0105, 0.0103, 0.0106], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 13:15:58,854 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.64 vs. limit=2.0 +2023-03-21 13:16:01,168 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3408, 3.2835, 3.2051, 3.4063, 3.0145, 2.8776, 3.5294, 2.3649], + device='cuda:0'), covar=tensor([0.0666, 0.0764, 0.1005, 0.0929, 0.0935, 0.1318, 0.0864, 0.3421], + device='cuda:0'), in_proj_covar=tensor([0.0329, 0.0337, 0.0269, 0.0348, 0.0279, 0.0284, 0.0343, 0.0235], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:16:02,499 INFO [train.py:901] (0/2) Epoch 48, batch 2150, loss[loss=0.1312, simple_loss=0.2225, pruned_loss=0.01997, over 7348.00 frames. ], tot_loss[loss=0.1255, simple_loss=0.2071, pruned_loss=0.02198, over 1439858.09 frames. ], batch size: 54, lr: 3.42e-03, grad_scale: 8.0 +2023-03-21 13:16:26,328 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=134923.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:16:27,137 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.93 vs. limit=2.0 +2023-03-21 13:16:29,184 INFO [train.py:901] (0/2) Epoch 48, batch 2200, loss[loss=0.1309, simple_loss=0.2068, pruned_loss=0.02752, over 7261.00 frames. ], tot_loss[loss=0.1261, simple_loss=0.2078, pruned_loss=0.02219, over 1442878.95 frames. ], batch size: 52, lr: 3.42e-03, grad_scale: 8.0 +2023-03-21 13:16:32,314 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 13:16:40,100 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3528, 2.0780, 2.2803, 3.3877, 1.8663, 3.1221, 1.3822, 3.0196], + device='cuda:0'), covar=tensor([0.0248, 0.1726, 0.2092, 0.0304, 0.4021, 0.0300, 0.1436, 0.0455], + device='cuda:0'), in_proj_covar=tensor([0.0208, 0.0241, 0.0252, 0.0210, 0.0246, 0.0219, 0.0216, 0.0228], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:16:45,165 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=134960.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:16:47,628 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=134965.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:16:48,034 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.290e+02 1.726e+02 2.022e+02 2.434e+02 4.958e+02, threshold=4.044e+02, percent-clipped=3.0 +2023-03-21 13:16:55,319 INFO [train.py:901] (0/2) Epoch 48, batch 2250, loss[loss=0.1401, simple_loss=0.2195, pruned_loss=0.03038, over 7325.00 frames. ], tot_loss[loss=0.1267, simple_loss=0.2083, pruned_loss=0.02252, over 1443549.58 frames. ], batch size: 75, lr: 3.42e-03, grad_scale: 8.0 +2023-03-21 13:17:06,987 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 13:17:07,464 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 13:17:11,151 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=135008.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:17:12,249 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6928, 3.4691, 3.3089, 3.6592, 3.2244, 3.0407, 3.8098, 2.5138], + device='cuda:0'), covar=tensor([0.0600, 0.0716, 0.1027, 0.0970, 0.0927, 0.1234, 0.0911, 0.3392], + device='cuda:0'), in_proj_covar=tensor([0.0328, 0.0336, 0.0269, 0.0347, 0.0278, 0.0283, 0.0342, 0.0234], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:17:13,648 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=135013.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:17:20,718 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 13:17:21,772 INFO [train.py:901] (0/2) Epoch 48, batch 2300, loss[loss=0.1114, simple_loss=0.195, pruned_loss=0.01392, over 7138.00 frames. ], tot_loss[loss=0.1266, simple_loss=0.2082, pruned_loss=0.02248, over 1444219.91 frames. ], batch size: 41, lr: 3.42e-03, grad_scale: 8.0 +2023-03-21 13:17:25,719 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.56 vs. limit=2.0 +2023-03-21 13:17:41,369 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.259e+02 1.741e+02 2.077e+02 2.361e+02 4.770e+02, threshold=4.154e+02, percent-clipped=2.0 +2023-03-21 13:17:47,944 INFO [train.py:901] (0/2) Epoch 48, batch 2350, loss[loss=0.1314, simple_loss=0.2133, pruned_loss=0.02474, over 7307.00 frames. ], tot_loss[loss=0.1268, simple_loss=0.2087, pruned_loss=0.02246, over 1445552.06 frames. ], batch size: 86, lr: 3.42e-03, grad_scale: 8.0 +2023-03-21 13:17:55,825 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7532, 3.2218, 2.8269, 3.0442, 2.9632, 2.6925, 2.8766, 2.8949], + device='cuda:0'), covar=tensor([0.0898, 0.0610, 0.0773, 0.0839, 0.0723, 0.0596, 0.1406, 0.1071], + device='cuda:0'), in_proj_covar=tensor([0.0061, 0.0062, 0.0070, 0.0062, 0.0059, 0.0067, 0.0059, 0.0056], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:18:03,105 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.64 vs. limit=2.0 +2023-03-21 13:18:08,375 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 13:18:12,170 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.76 vs. limit=2.0 +2023-03-21 13:18:13,321 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-03-21 13:18:13,933 INFO [train.py:901] (0/2) Epoch 48, batch 2400, loss[loss=0.1211, simple_loss=0.2071, pruned_loss=0.01758, over 7327.00 frames. ], tot_loss[loss=0.1268, simple_loss=0.2086, pruned_loss=0.0225, over 1445672.77 frames. ], batch size: 83, lr: 3.42e-03, grad_scale: 8.0 +2023-03-21 13:18:14,956 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 13:18:20,734 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3417, 3.8872, 3.7449, 4.3493, 4.1000, 4.2350, 3.6755, 3.8616], + device='cuda:0'), covar=tensor([0.0979, 0.2662, 0.2383, 0.1115, 0.1040, 0.1371, 0.1152, 0.1400], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0407, 0.0307, 0.0321, 0.0238, 0.0383, 0.0240, 0.0289], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 13:18:22,614 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.92 vs. limit=2.0 +2023-03-21 13:18:25,533 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2554, 2.5335, 2.0911, 3.0566, 2.7378, 2.9642, 2.7597, 2.5388], + device='cuda:0'), covar=tensor([0.2252, 0.1412, 0.3908, 0.0875, 0.0374, 0.0310, 0.0434, 0.0414], + device='cuda:0'), in_proj_covar=tensor([0.0245, 0.0230, 0.0240, 0.0253, 0.0202, 0.0204, 0.0217, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:18:26,367 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 13:18:29,435 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 13:18:33,398 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.237e+02 1.657e+02 2.114e+02 2.439e+02 4.545e+02, threshold=4.228e+02, percent-clipped=1.0 +2023-03-21 13:18:36,798 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=135171.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:18:40,683 INFO [train.py:901] (0/2) Epoch 48, batch 2450, loss[loss=0.1439, simple_loss=0.2275, pruned_loss=0.03018, over 6616.00 frames. ], tot_loss[loss=0.1265, simple_loss=0.2082, pruned_loss=0.02247, over 1443867.24 frames. ], batch size: 106, lr: 3.42e-03, grad_scale: 8.0 +2023-03-21 13:18:43,378 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.0261, 2.8677, 3.0234, 3.2304, 2.8638, 2.7825, 3.2778, 2.3114], + device='cuda:0'), covar=tensor([0.0627, 0.0623, 0.0982, 0.0808, 0.0748, 0.1072, 0.0758, 0.3341], + device='cuda:0'), in_proj_covar=tensor([0.0328, 0.0336, 0.0268, 0.0346, 0.0277, 0.0283, 0.0343, 0.0235], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:18:53,639 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.84 vs. limit=2.0 +2023-03-21 13:18:56,378 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 13:19:01,106 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=135218.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:19:06,646 INFO [train.py:901] (0/2) Epoch 48, batch 2500, loss[loss=0.1014, simple_loss=0.1759, pruned_loss=0.01343, over 6928.00 frames. ], tot_loss[loss=0.126, simple_loss=0.2075, pruned_loss=0.02218, over 1444006.90 frames. ], batch size: 35, lr: 3.42e-03, grad_scale: 8.0 +2023-03-21 13:19:08,329 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=135232.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:19:24,351 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 13:19:26,857 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.176e+02 1.741e+02 2.067e+02 2.417e+02 6.629e+02, threshold=4.134e+02, percent-clipped=1.0 +2023-03-21 13:19:33,501 INFO [train.py:901] (0/2) Epoch 48, batch 2550, loss[loss=0.1342, simple_loss=0.2228, pruned_loss=0.02286, over 7318.00 frames. ], tot_loss[loss=0.1259, simple_loss=0.2074, pruned_loss=0.02219, over 1444218.15 frames. ], batch size: 83, lr: 3.42e-03, grad_scale: 8.0 +2023-03-21 13:19:58,867 INFO [train.py:901] (0/2) Epoch 48, batch 2600, loss[loss=0.1262, simple_loss=0.2094, pruned_loss=0.0215, over 7321.00 frames. ], tot_loss[loss=0.1251, simple_loss=0.2065, pruned_loss=0.02186, over 1443565.65 frames. ], batch size: 83, lr: 3.42e-03, grad_scale: 8.0 +2023-03-21 13:20:18,010 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.259e+02 1.707e+02 2.022e+02 2.362e+02 4.542e+02, threshold=4.043e+02, percent-clipped=1.0 +2023-03-21 13:20:24,537 INFO [train.py:901] (0/2) Epoch 48, batch 2650, loss[loss=0.1347, simple_loss=0.2253, pruned_loss=0.02205, over 6715.00 frames. ], tot_loss[loss=0.126, simple_loss=0.2077, pruned_loss=0.02213, over 1442957.58 frames. ], batch size: 106, lr: 3.42e-03, grad_scale: 8.0 +2023-03-21 13:20:50,012 INFO [train.py:901] (0/2) Epoch 48, batch 2700, loss[loss=0.1316, simple_loss=0.2136, pruned_loss=0.02485, over 7341.00 frames. ], tot_loss[loss=0.1262, simple_loss=0.2079, pruned_loss=0.0223, over 1441378.00 frames. ], batch size: 54, lr: 3.42e-03, grad_scale: 8.0 +2023-03-21 13:20:50,605 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4385, 4.0010, 4.0255, 4.1322, 4.1039, 4.0183, 4.3906, 3.8896], + device='cuda:0'), covar=tensor([0.0166, 0.0182, 0.0133, 0.0172, 0.0470, 0.0130, 0.0136, 0.0176], + device='cuda:0'), in_proj_covar=tensor([0.0101, 0.0107, 0.0109, 0.0094, 0.0185, 0.0113, 0.0110, 0.0119], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0004, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 13:21:08,367 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.481e+02 1.782e+02 2.093e+02 2.517e+02 6.149e+02, threshold=4.186e+02, percent-clipped=2.0 +2023-03-21 13:21:14,688 INFO [train.py:901] (0/2) Epoch 48, batch 2750, loss[loss=0.1255, simple_loss=0.2141, pruned_loss=0.01843, over 7279.00 frames. ], tot_loss[loss=0.1268, simple_loss=0.2083, pruned_loss=0.02265, over 1443205.55 frames. ], batch size: 70, lr: 3.42e-03, grad_scale: 8.0 +2023-03-21 13:21:18,527 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.16 vs. limit=2.0 +2023-03-21 13:21:34,336 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=135518.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:21:38,628 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=135527.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:21:39,562 INFO [train.py:901] (0/2) Epoch 48, batch 2800, loss[loss=0.1201, simple_loss=0.2081, pruned_loss=0.01604, over 7215.00 frames. ], tot_loss[loss=0.1266, simple_loss=0.2079, pruned_loss=0.02263, over 1442066.78 frames. ], batch size: 93, lr: 3.42e-03, grad_scale: 8.0 +2023-03-21 13:21:52,005 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-48.pt +2023-03-21 13:22:04,764 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 13:22:04,821 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 13:22:04,847 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 13:22:04,864 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 13:22:05,044 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 13:22:05,059 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 13:22:05,071 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 13:22:05,529 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 13:22:05,605 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 13:22:05,620 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 13:22:05,638 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 13:22:05,806 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 13:22:05,812 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 13:22:05,820 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 13:22:05,970 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 13:22:05,973 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 13:22:06,032 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 13:22:08,196 INFO [train.py:901] (0/2) Epoch 49, batch 0, loss[loss=0.1223, simple_loss=0.2042, pruned_loss=0.02024, over 7271.00 frames. ], tot_loss[loss=0.1223, simple_loss=0.2042, pruned_loss=0.02024, over 7271.00 frames. ], batch size: 57, lr: 3.38e-03, grad_scale: 8.0 +2023-03-21 13:22:08,198 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 13:22:14,433 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9995, 3.7487, 4.1477, 4.2718, 4.2431, 4.0262, 4.5652, 4.1970], + device='cuda:0'), covar=tensor([0.0038, 0.0099, 0.0036, 0.0029, 0.0027, 0.0038, 0.0015, 0.0044], + device='cuda:0'), in_proj_covar=tensor([0.0052, 0.0074, 0.0062, 0.0060, 0.0057, 0.0063, 0.0049, 0.0082], + device='cuda:0'), out_proj_covar=tensor([8.5679e-05, 1.4646e-04, 1.0733e-04, 9.8756e-05, 9.2221e-05, 1.0694e-04, + 9.0068e-05, 1.4766e-04], device='cuda:0') +2023-03-21 13:22:21,311 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.1114, 4.6794, 4.4644, 5.1492, 4.8633, 5.1167, 4.4394, 4.9114], + device='cuda:0'), covar=tensor([0.0576, 0.1894, 0.1686, 0.0966, 0.0589, 0.0818, 0.0622, 0.0656], + device='cuda:0'), in_proj_covar=tensor([0.0273, 0.0405, 0.0305, 0.0318, 0.0238, 0.0382, 0.0238, 0.0287], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 13:22:25,976 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3276, 2.4434, 2.5109, 3.5376, 2.0086, 3.2367, 1.5204, 3.2571], + device='cuda:0'), covar=tensor([0.0179, 0.1499, 0.1865, 0.0241, 0.4184, 0.0288, 0.1375, 0.0442], + device='cuda:0'), in_proj_covar=tensor([0.0208, 0.0241, 0.0253, 0.0212, 0.0246, 0.0218, 0.0216, 0.0228], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:22:33,016 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7051, 2.8761, 2.6374, 2.8472, 2.8223, 2.5063, 2.8351, 2.7010], + device='cuda:0'), covar=tensor([0.0658, 0.0417, 0.0854, 0.0684, 0.0400, 0.0467, 0.0299, 0.0950], + device='cuda:0'), in_proj_covar=tensor([0.0061, 0.0062, 0.0070, 0.0062, 0.0058, 0.0066, 0.0059, 0.0056], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:22:33,728 INFO [train.py:935] (0/2) Epoch 49, validation: loss=0.1665, simple_loss=0.2593, pruned_loss=0.03685, over 1622729.00 frames. +2023-03-21 13:22:33,728 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 13:22:40,313 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.219e+02 1.655e+02 2.021e+02 2.392e+02 5.342e+02, threshold=4.041e+02, percent-clipped=1.0 +2023-03-21 13:22:40,338 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 13:22:40,382 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=135566.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:22:51,610 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 13:22:59,286 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 13:23:00,803 INFO [train.py:901] (0/2) Epoch 49, batch 50, loss[loss=0.1125, simple_loss=0.1919, pruned_loss=0.01652, over 7255.00 frames. ], tot_loss[loss=0.1238, simple_loss=0.2065, pruned_loss=0.02051, over 325602.83 frames. ], batch size: 64, lr: 3.38e-03, grad_scale: 8.0 +2023-03-21 13:23:01,879 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 13:23:04,940 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 13:23:14,658 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=135630.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:23:16,632 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5955, 3.9209, 2.9502, 4.0229, 3.4138, 3.7308, 1.9190, 3.0174], + device='cuda:0'), covar=tensor([0.0562, 0.0664, 0.2373, 0.0541, 0.0563, 0.1145, 0.4177, 0.1923], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0256, 0.0275, 0.0267, 0.0265, 0.0262, 0.0227, 0.0255], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:23:21,974 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 13:23:22,485 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 13:23:25,839 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.28 vs. limit=5.0 +2023-03-21 13:23:26,010 INFO [train.py:901] (0/2) Epoch 49, batch 100, loss[loss=0.1338, simple_loss=0.2216, pruned_loss=0.02304, over 6689.00 frames. ], tot_loss[loss=0.1251, simple_loss=0.2072, pruned_loss=0.02147, over 573485.10 frames. ], batch size: 106, lr: 3.38e-03, grad_scale: 8.0 +2023-03-21 13:23:32,728 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.252e+02 1.833e+02 2.096e+02 2.422e+02 4.356e+02, threshold=4.193e+02, percent-clipped=1.0 +2023-03-21 13:23:35,319 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.65 vs. limit=2.0 +2023-03-21 13:23:47,053 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=135691.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:23:52,951 INFO [train.py:901] (0/2) Epoch 49, batch 150, loss[loss=0.12, simple_loss=0.2125, pruned_loss=0.01377, over 7285.00 frames. ], tot_loss[loss=0.1254, simple_loss=0.2075, pruned_loss=0.02167, over 768574.89 frames. ], batch size: 66, lr: 3.38e-03, grad_scale: 8.0 +2023-03-21 13:24:06,690 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2427, 4.3900, 4.0812, 4.3466, 3.9229, 4.2565, 4.6645, 4.6444], + device='cuda:0'), covar=tensor([0.0195, 0.0128, 0.0219, 0.0165, 0.0348, 0.0323, 0.0198, 0.0161], + device='cuda:0'), in_proj_covar=tensor([0.0127, 0.0131, 0.0126, 0.0129, 0.0116, 0.0104, 0.0102, 0.0105], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 13:24:18,777 INFO [train.py:901] (0/2) Epoch 49, batch 200, loss[loss=0.1098, simple_loss=0.1904, pruned_loss=0.01461, over 6960.00 frames. ], tot_loss[loss=0.1252, simple_loss=0.2073, pruned_loss=0.02155, over 920584.75 frames. ], batch size: 35, lr: 3.38e-03, grad_scale: 8.0 +2023-03-21 13:24:18,892 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9781, 4.1192, 3.8185, 4.0751, 3.6764, 3.9923, 4.3484, 4.3385], + device='cuda:0'), covar=tensor([0.0221, 0.0148, 0.0243, 0.0192, 0.0387, 0.0354, 0.0223, 0.0178], + device='cuda:0'), in_proj_covar=tensor([0.0128, 0.0131, 0.0127, 0.0129, 0.0116, 0.0105, 0.0102, 0.0105], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 13:24:24,372 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 13:24:25,361 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.280e+02 1.715e+02 2.006e+02 2.248e+02 4.801e+02, threshold=4.012e+02, percent-clipped=1.0 +2023-03-21 13:24:29,523 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 13:24:35,029 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 13:24:45,020 INFO [train.py:901] (0/2) Epoch 49, batch 250, loss[loss=0.1248, simple_loss=0.2048, pruned_loss=0.02234, over 7304.00 frames. ], tot_loss[loss=0.1253, simple_loss=0.2072, pruned_loss=0.02166, over 1038676.72 frames. ], batch size: 80, lr: 3.38e-03, grad_scale: 4.0 +2023-03-21 13:24:48,089 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 13:24:57,247 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=135827.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:25:08,034 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 13:25:11,187 INFO [train.py:901] (0/2) Epoch 49, batch 300, loss[loss=0.1253, simple_loss=0.2082, pruned_loss=0.0212, over 7270.00 frames. ], tot_loss[loss=0.1253, simple_loss=0.2072, pruned_loss=0.02175, over 1128204.67 frames. ], batch size: 52, lr: 3.38e-03, grad_scale: 4.0 +2023-03-21 13:25:17,393 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 13:25:18,891 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.312e+02 1.783e+02 2.054e+02 2.391e+02 4.462e+02, threshold=4.108e+02, percent-clipped=1.0 +2023-03-21 13:25:23,070 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=135875.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:25:35,195 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.2758, 4.7986, 4.5874, 5.2776, 5.0410, 5.1735, 4.7171, 4.8416], + device='cuda:0'), covar=tensor([0.0804, 0.2478, 0.2341, 0.1086, 0.0907, 0.1204, 0.0750, 0.1230], + device='cuda:0'), in_proj_covar=tensor([0.0273, 0.0404, 0.0303, 0.0318, 0.0238, 0.0380, 0.0239, 0.0287], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 13:25:37,120 INFO [train.py:901] (0/2) Epoch 49, batch 350, loss[loss=0.143, simple_loss=0.223, pruned_loss=0.03145, over 7347.00 frames. ], tot_loss[loss=0.1259, simple_loss=0.2077, pruned_loss=0.02204, over 1199860.12 frames. ], batch size: 73, lr: 3.38e-03, grad_scale: 4.0 +2023-03-21 13:25:52,290 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 13:26:03,410 INFO [train.py:901] (0/2) Epoch 49, batch 400, loss[loss=0.1355, simple_loss=0.2202, pruned_loss=0.02542, over 7365.00 frames. ], tot_loss[loss=0.1251, simple_loss=0.2068, pruned_loss=0.02171, over 1252198.09 frames. ], batch size: 51, lr: 3.38e-03, grad_scale: 8.0 +2023-03-21 13:26:04,648 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5944, 1.8723, 1.5971, 1.7661, 1.8664, 1.8168, 1.7677, 1.5089], + device='cuda:0'), covar=tensor([0.0188, 0.0206, 0.0326, 0.0250, 0.0134, 0.0170, 0.0177, 0.0218], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0042, 0.0040, 0.0042, 0.0039, 0.0039, 0.0041, 0.0051], + device='cuda:0'), out_proj_covar=tensor([4.8412e-05, 4.6503e-05, 4.5218e-05, 4.5933e-05, 4.3597e-05, 4.3752e-05, + 4.5889e-05, 5.5443e-05], device='cuda:0') +2023-03-21 13:26:10,506 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.343e+02 1.801e+02 2.070e+02 2.366e+02 4.044e+02, threshold=4.140e+02, percent-clipped=0.0 +2023-03-21 13:26:18,126 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=135982.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:26:20,090 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=135986.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:26:20,297 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.90 vs. limit=5.0 +2023-03-21 13:26:24,147 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=135994.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:26:27,493 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-136000.pt +2023-03-21 13:26:32,950 INFO [train.py:901] (0/2) Epoch 49, batch 450, loss[loss=0.1323, simple_loss=0.2103, pruned_loss=0.02712, over 7333.00 frames. ], tot_loss[loss=0.1249, simple_loss=0.2061, pruned_loss=0.02186, over 1291325.85 frames. ], batch size: 54, lr: 3.38e-03, grad_scale: 8.0 +2023-03-21 13:26:40,081 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 13:26:40,563 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 13:26:54,441 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=136043.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:26:59,373 INFO [train.py:901] (0/2) Epoch 49, batch 500, loss[loss=0.1135, simple_loss=0.1929, pruned_loss=0.017, over 7241.00 frames. ], tot_loss[loss=0.1251, simple_loss=0.2064, pruned_loss=0.02193, over 1326320.68 frames. ], batch size: 55, lr: 3.37e-03, grad_scale: 8.0 +2023-03-21 13:27:00,504 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=136055.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:27:06,353 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.271e+02 1.831e+02 2.073e+02 2.387e+02 4.515e+02, threshold=4.145e+02, percent-clipped=2.0 +2023-03-21 13:27:12,937 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 13:27:13,952 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 13:27:14,945 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 13:27:16,964 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 13:27:21,511 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 13:27:25,082 INFO [train.py:901] (0/2) Epoch 49, batch 550, loss[loss=0.1377, simple_loss=0.2166, pruned_loss=0.02944, over 7320.00 frames. ], tot_loss[loss=0.1256, simple_loss=0.207, pruned_loss=0.0221, over 1351803.02 frames. ], batch size: 59, lr: 3.37e-03, grad_scale: 8.0 +2023-03-21 13:27:32,973 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 13:27:41,229 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 13:27:44,240 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 13:27:45,387 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=136142.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:27:50,829 INFO [train.py:901] (0/2) Epoch 49, batch 600, loss[loss=0.1209, simple_loss=0.2039, pruned_loss=0.01898, over 7282.00 frames. ], tot_loss[loss=0.1257, simple_loss=0.207, pruned_loss=0.02219, over 1367993.59 frames. ], batch size: 66, lr: 3.37e-03, grad_scale: 8.0 +2023-03-21 13:27:52,368 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 13:27:57,842 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.300e+02 1.816e+02 2.119e+02 2.512e+02 4.639e+02, threshold=4.239e+02, percent-clipped=1.0 +2023-03-21 13:28:09,619 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 13:28:10,900 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.95 vs. limit=2.0 +2023-03-21 13:28:13,242 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6145, 1.7976, 1.5510, 1.6979, 1.7884, 1.6894, 1.8105, 1.4187], + device='cuda:0'), covar=tensor([0.0198, 0.0229, 0.0380, 0.0235, 0.0184, 0.0186, 0.0165, 0.0266], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0042, 0.0040, 0.0042, 0.0039, 0.0039, 0.0041, 0.0051], + device='cuda:0'), out_proj_covar=tensor([4.8375e-05, 4.6251e-05, 4.5046e-05, 4.5661e-05, 4.3431e-05, 4.3432e-05, + 4.5720e-05, 5.5289e-05], device='cuda:0') +2023-03-21 13:28:17,040 INFO [train.py:901] (0/2) Epoch 49, batch 650, loss[loss=0.127, simple_loss=0.2109, pruned_loss=0.02153, over 7346.00 frames. ], tot_loss[loss=0.1256, simple_loss=0.2068, pruned_loss=0.02216, over 1384025.04 frames. ], batch size: 54, lr: 3.37e-03, grad_scale: 8.0 +2023-03-21 13:28:17,207 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=136203.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:28:18,179 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6575, 1.8498, 1.5148, 1.7362, 1.8818, 1.7759, 1.7789, 1.4636], + device='cuda:0'), covar=tensor([0.0252, 0.0228, 0.0346, 0.0256, 0.0195, 0.0196, 0.0263, 0.0314], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0042, 0.0040, 0.0042, 0.0039, 0.0039, 0.0041, 0.0050], + device='cuda:0'), out_proj_covar=tensor([4.8323e-05, 4.6200e-05, 4.5008e-05, 4.5599e-05, 4.3379e-05, 4.3397e-05, + 4.5687e-05, 5.5244e-05], device='cuda:0') +2023-03-21 13:28:19,201 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 13:28:35,816 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 13:28:42,969 INFO [train.py:901] (0/2) Epoch 49, batch 700, loss[loss=0.09834, simple_loss=0.1643, pruned_loss=0.01617, over 6993.00 frames. ], tot_loss[loss=0.1258, simple_loss=0.2071, pruned_loss=0.02225, over 1398247.40 frames. ], batch size: 35, lr: 3.37e-03, grad_scale: 8.0 +2023-03-21 13:28:44,597 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 13:28:50,199 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.282e+02 1.813e+02 2.063e+02 2.417e+02 3.262e+02, threshold=4.126e+02, percent-clipped=0.0 +2023-03-21 13:28:50,347 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.2861, 2.4689, 2.6567, 2.1835, 2.5592, 2.5816, 2.2555, 1.9988], + device='cuda:0'), covar=tensor([0.0674, 0.0420, 0.0360, 0.0357, 0.0484, 0.0465, 0.0461, 0.0371], + device='cuda:0'), in_proj_covar=tensor([0.0046, 0.0044, 0.0045, 0.0044, 0.0041, 0.0042, 0.0049, 0.0048], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 13:28:54,309 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.66 vs. limit=5.0 +2023-03-21 13:29:00,534 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=136286.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:29:03,340 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3567, 3.1008, 3.1696, 3.2852, 2.9513, 2.9044, 3.3321, 2.2917], + device='cuda:0'), covar=tensor([0.0552, 0.0721, 0.0996, 0.0854, 0.0877, 0.1285, 0.0929, 0.3231], + device='cuda:0'), in_proj_covar=tensor([0.0328, 0.0340, 0.0269, 0.0347, 0.0279, 0.0285, 0.0343, 0.0234], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:29:09,807 INFO [train.py:901] (0/2) Epoch 49, batch 750, loss[loss=0.1031, simple_loss=0.1741, pruned_loss=0.01606, over 6411.00 frames. ], tot_loss[loss=0.1258, simple_loss=0.2069, pruned_loss=0.02233, over 1404398.07 frames. ], batch size: 28, lr: 3.37e-03, grad_scale: 8.0 +2023-03-21 13:29:10,348 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 13:29:10,869 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 13:29:24,717 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 13:29:25,768 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=136334.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:29:26,342 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=136335.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:29:26,802 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9743, 4.3774, 4.4002, 4.3970, 4.3621, 4.0184, 4.4256, 4.2372], + device='cuda:0'), covar=tensor([0.0924, 0.0985, 0.0837, 0.0946, 0.0685, 0.0861, 0.0810, 0.0909], + device='cuda:0'), in_proj_covar=tensor([0.0220, 0.0271, 0.0211, 0.0210, 0.0160, 0.0238, 0.0220, 0.0150], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:29:27,775 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=136338.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:29:29,761 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 13:29:34,007 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=136350.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:29:35,492 INFO [train.py:901] (0/2) Epoch 49, batch 800, loss[loss=0.1352, simple_loss=0.2187, pruned_loss=0.02592, over 7337.00 frames. ], tot_loss[loss=0.1257, simple_loss=0.207, pruned_loss=0.02219, over 1414605.53 frames. ], batch size: 73, lr: 3.37e-03, grad_scale: 8.0 +2023-03-21 13:29:35,520 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 13:29:37,000 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 13:29:43,114 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.323e+02 1.743e+02 2.098e+02 2.420e+02 5.303e+02, threshold=4.195e+02, percent-clipped=1.0 +2023-03-21 13:29:44,322 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8445, 2.3755, 3.0795, 2.8574, 3.0456, 2.7799, 2.4678, 2.9059], + device='cuda:0'), covar=tensor([0.1007, 0.0998, 0.0945, 0.1140, 0.0706, 0.0981, 0.2002, 0.1294], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0073, 0.0055, 0.0053, 0.0054, 0.0053, 0.0071, 0.0054], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 13:29:47,704 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 13:29:49,005 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6424, 2.2352, 2.4182, 3.7963, 1.9976, 3.4848, 1.5131, 3.3743], + device='cuda:0'), covar=tensor([0.0242, 0.1804, 0.2112, 0.0267, 0.4615, 0.0411, 0.1649, 0.0558], + device='cuda:0'), in_proj_covar=tensor([0.0206, 0.0238, 0.0249, 0.0209, 0.0244, 0.0215, 0.0214, 0.0225], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:29:58,576 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=136396.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:30:02,309 INFO [train.py:901] (0/2) Epoch 49, batch 850, loss[loss=0.1304, simple_loss=0.2125, pruned_loss=0.02416, over 7363.00 frames. ], tot_loss[loss=0.1259, simple_loss=0.2073, pruned_loss=0.02227, over 1423298.05 frames. ], batch size: 63, lr: 3.37e-03, grad_scale: 8.0 +2023-03-21 13:30:07,291 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 13:30:07,299 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 13:30:10,925 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7681, 3.8899, 3.6187, 3.8570, 3.6263, 3.8022, 4.1744, 4.1658], + device='cuda:0'), covar=tensor([0.0248, 0.0172, 0.0285, 0.0187, 0.0347, 0.0415, 0.0253, 0.0190], + device='cuda:0'), in_proj_covar=tensor([0.0127, 0.0131, 0.0126, 0.0129, 0.0115, 0.0104, 0.0102, 0.0105], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 13:30:12,376 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 13:30:15,421 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 13:30:22,650 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7943, 2.3041, 2.9043, 2.7668, 2.9018, 2.7628, 2.5301, 2.8191], + device='cuda:0'), covar=tensor([0.1214, 0.1072, 0.1146, 0.1400, 0.0839, 0.0898, 0.1549, 0.1350], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0073, 0.0054, 0.0053, 0.0054, 0.0052, 0.0071, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 13:30:26,691 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3304, 4.7663, 4.8405, 4.7888, 4.7355, 4.3669, 4.8485, 4.6839], + device='cuda:0'), covar=tensor([0.0430, 0.0388, 0.0348, 0.0487, 0.0334, 0.0394, 0.0348, 0.0447], + device='cuda:0'), in_proj_covar=tensor([0.0221, 0.0271, 0.0211, 0.0211, 0.0161, 0.0238, 0.0221, 0.0150], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:30:28,110 INFO [train.py:901] (0/2) Epoch 49, batch 900, loss[loss=0.143, simple_loss=0.222, pruned_loss=0.03196, over 7298.00 frames. ], tot_loss[loss=0.1259, simple_loss=0.2072, pruned_loss=0.02226, over 1425833.36 frames. ], batch size: 68, lr: 3.37e-03, grad_scale: 8.0 +2023-03-21 13:30:35,851 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.260e+02 1.732e+02 2.021e+02 2.349e+02 3.234e+02, threshold=4.042e+02, percent-clipped=0.0 +2023-03-21 13:30:50,029 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.0491, 4.6119, 4.3898, 5.0473, 4.7994, 4.9358, 4.4375, 4.5877], + device='cuda:0'), covar=tensor([0.0830, 0.2141, 0.2288, 0.0952, 0.0953, 0.1156, 0.0778, 0.1057], + device='cuda:0'), in_proj_covar=tensor([0.0273, 0.0403, 0.0302, 0.0317, 0.0238, 0.0378, 0.0238, 0.0286], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 13:30:51,553 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=136498.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:30:53,934 INFO [train.py:901] (0/2) Epoch 49, batch 950, loss[loss=0.1363, simple_loss=0.2161, pruned_loss=0.02826, over 7253.00 frames. ], tot_loss[loss=0.1262, simple_loss=0.2073, pruned_loss=0.02252, over 1427937.55 frames. ], batch size: 47, lr: 3.37e-03, grad_scale: 8.0 +2023-03-21 13:30:55,003 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 13:31:08,723 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6698, 2.8168, 3.6775, 3.6698, 3.7736, 3.7733, 3.6376, 3.6000], + device='cuda:0'), covar=tensor([0.0034, 0.0162, 0.0036, 0.0031, 0.0028, 0.0030, 0.0056, 0.0053], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0076, 0.0063, 0.0061, 0.0058, 0.0065, 0.0051, 0.0084], + device='cuda:0'), out_proj_covar=tensor([8.8292e-05, 1.4988e-04, 1.0913e-04, 1.0156e-04, 9.4737e-05, 1.0888e-04, + 9.2517e-05, 1.5139e-04], device='cuda:0') +2023-03-21 13:31:18,901 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 13:31:19,883 INFO [train.py:901] (0/2) Epoch 49, batch 1000, loss[loss=0.1169, simple_loss=0.2046, pruned_loss=0.01462, over 7318.00 frames. ], tot_loss[loss=0.1255, simple_loss=0.2067, pruned_loss=0.02216, over 1429678.18 frames. ], batch size: 83, lr: 3.37e-03, grad_scale: 8.0 +2023-03-21 13:31:27,256 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4871, 2.6471, 2.7123, 2.4251, 2.8027, 2.6426, 2.4397, 2.2308], + device='cuda:0'), covar=tensor([0.0581, 0.0638, 0.0473, 0.0377, 0.0822, 0.0568, 0.0347, 0.0451], + device='cuda:0'), in_proj_covar=tensor([0.0045, 0.0044, 0.0044, 0.0044, 0.0041, 0.0042, 0.0047, 0.0048], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 13:31:27,614 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.406e+02 1.764e+02 1.959e+02 2.305e+02 3.488e+02, threshold=3.918e+02, percent-clipped=0.0 +2023-03-21 13:31:40,814 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 13:31:46,171 INFO [train.py:901] (0/2) Epoch 49, batch 1050, loss[loss=0.1315, simple_loss=0.1994, pruned_loss=0.03173, over 7231.00 frames. ], tot_loss[loss=0.1251, simple_loss=0.2062, pruned_loss=0.02202, over 1432176.24 frames. ], batch size: 45, lr: 3.37e-03, grad_scale: 8.0 +2023-03-21 13:32:02,292 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.65 vs. limit=2.0 +2023-03-21 13:32:02,489 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 13:32:04,576 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=136638.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:32:07,601 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 13:32:07,732 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=136643.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:32:11,324 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=136650.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:32:12,724 INFO [train.py:901] (0/2) Epoch 49, batch 1100, loss[loss=0.1511, simple_loss=0.2241, pruned_loss=0.03907, over 7311.00 frames. ], tot_loss[loss=0.1254, simple_loss=0.2068, pruned_loss=0.02204, over 1435536.20 frames. ], batch size: 49, lr: 3.37e-03, grad_scale: 8.0 +2023-03-21 13:32:19,843 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 8.990e+01 1.716e+02 2.031e+02 2.257e+02 3.350e+02, threshold=4.062e+02, percent-clipped=0.0 +2023-03-21 13:32:29,506 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=136686.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:32:31,933 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=136691.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:32:35,509 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=136698.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:32:35,988 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 13:32:36,495 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 13:32:38,610 INFO [train.py:901] (0/2) Epoch 49, batch 1150, loss[loss=0.09437, simple_loss=0.1553, pruned_loss=0.01671, over 6555.00 frames. ], tot_loss[loss=0.1259, simple_loss=0.2071, pruned_loss=0.02231, over 1436050.64 frames. ], batch size: 27, lr: 3.37e-03, grad_scale: 8.0 +2023-03-21 13:32:39,291 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=136704.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:32:40,412 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.24 vs. limit=5.0 +2023-03-21 13:32:48,858 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 13:32:49,875 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 13:32:55,814 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1221, 3.0470, 3.0102, 3.2125, 2.7887, 2.8586, 3.3400, 2.2548], + device='cuda:0'), covar=tensor([0.0623, 0.0740, 0.1141, 0.0844, 0.0930, 0.1437, 0.0914, 0.3271], + device='cuda:0'), in_proj_covar=tensor([0.0328, 0.0338, 0.0269, 0.0347, 0.0279, 0.0284, 0.0343, 0.0234], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:33:01,302 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9579, 4.1435, 3.8137, 4.1410, 3.7485, 4.1004, 4.4173, 4.3930], + device='cuda:0'), covar=tensor([0.0204, 0.0149, 0.0258, 0.0163, 0.0327, 0.0257, 0.0200, 0.0163], + device='cuda:0'), in_proj_covar=tensor([0.0128, 0.0131, 0.0125, 0.0128, 0.0116, 0.0104, 0.0102, 0.0105], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 13:33:04,801 INFO [train.py:901] (0/2) Epoch 49, batch 1200, loss[loss=0.1517, simple_loss=0.2322, pruned_loss=0.03559, over 7260.00 frames. ], tot_loss[loss=0.1256, simple_loss=0.207, pruned_loss=0.02213, over 1437208.57 frames. ], batch size: 55, lr: 3.37e-03, grad_scale: 8.0 +2023-03-21 13:33:11,773 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.244e+02 1.907e+02 2.162e+02 2.451e+02 4.461e+02, threshold=4.324e+02, percent-clipped=2.0 +2023-03-21 13:33:12,962 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1362, 3.4514, 2.4868, 3.8169, 2.9234, 3.4215, 1.6603, 2.6981], + device='cuda:0'), covar=tensor([0.0460, 0.0736, 0.2800, 0.0635, 0.0503, 0.0634, 0.4499, 0.1796], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0257, 0.0275, 0.0267, 0.0264, 0.0262, 0.0226, 0.0254], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:33:22,660 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 13:33:28,204 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=136798.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:33:30,889 INFO [train.py:901] (0/2) Epoch 49, batch 1250, loss[loss=0.1265, simple_loss=0.2122, pruned_loss=0.02036, over 7284.00 frames. ], tot_loss[loss=0.1264, simple_loss=0.2077, pruned_loss=0.02251, over 1439188.90 frames. ], batch size: 68, lr: 3.37e-03, grad_scale: 8.0 +2023-03-21 13:33:32,522 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=136806.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:33:46,104 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.19 vs. limit=2.0 +2023-03-21 13:33:47,279 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 13:33:47,378 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5316, 2.7447, 3.4474, 3.5661, 3.6008, 3.6206, 3.3223, 3.4387], + device='cuda:0'), covar=tensor([0.0034, 0.0162, 0.0037, 0.0032, 0.0027, 0.0028, 0.0065, 0.0057], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0076, 0.0063, 0.0060, 0.0058, 0.0064, 0.0050, 0.0084], + device='cuda:0'), out_proj_covar=tensor([8.7977e-05, 1.4869e-04, 1.0852e-04, 1.0042e-04, 9.4465e-05, 1.0780e-04, + 9.1647e-05, 1.5072e-04], device='cuda:0') +2023-03-21 13:33:51,274 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 13:33:52,752 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 13:33:53,310 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=136846.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:33:56,777 INFO [train.py:901] (0/2) Epoch 49, batch 1300, loss[loss=0.1184, simple_loss=0.2038, pruned_loss=0.01655, over 7137.00 frames. ], tot_loss[loss=0.1263, simple_loss=0.2077, pruned_loss=0.02242, over 1440925.20 frames. ], batch size: 41, lr: 3.36e-03, grad_scale: 8.0 +2023-03-21 13:34:03,864 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.273e+02 1.719e+02 1.984e+02 2.382e+02 4.139e+02, threshold=3.968e+02, percent-clipped=0.0 +2023-03-21 13:34:04,044 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=136867.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:34:04,485 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.7331, 4.2813, 4.1243, 4.7545, 4.5385, 4.6263, 4.1218, 4.2604], + device='cuda:0'), covar=tensor([0.0903, 0.2347, 0.2442, 0.0852, 0.1019, 0.1179, 0.0806, 0.1180], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0404, 0.0304, 0.0319, 0.0240, 0.0380, 0.0239, 0.0288], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 13:34:12,115 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7137, 2.9143, 3.6534, 3.7255, 3.7366, 3.7618, 3.5940, 3.6074], + device='cuda:0'), covar=tensor([0.0033, 0.0146, 0.0033, 0.0030, 0.0029, 0.0032, 0.0050, 0.0053], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0075, 0.0062, 0.0060, 0.0058, 0.0064, 0.0050, 0.0083], + device='cuda:0'), out_proj_covar=tensor([8.7670e-05, 1.4823e-04, 1.0813e-04, 9.9929e-05, 9.4148e-05, 1.0740e-04, + 9.1025e-05, 1.4992e-04], device='cuda:0') +2023-03-21 13:34:16,097 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 13:34:18,633 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 13:34:22,155 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 13:34:22,634 INFO [train.py:901] (0/2) Epoch 49, batch 1350, loss[loss=0.1311, simple_loss=0.2079, pruned_loss=0.0271, over 7270.00 frames. ], tot_loss[loss=0.1262, simple_loss=0.2076, pruned_loss=0.02239, over 1443250.54 frames. ], batch size: 47, lr: 3.36e-03, grad_scale: 8.0 +2023-03-21 13:34:33,353 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 13:34:34,944 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2167, 4.6656, 4.7194, 4.6601, 4.6562, 4.2095, 4.7344, 4.5350], + device='cuda:0'), covar=tensor([0.0471, 0.0430, 0.0457, 0.0521, 0.0343, 0.0433, 0.0395, 0.0456], + device='cuda:0'), in_proj_covar=tensor([0.0223, 0.0274, 0.0215, 0.0214, 0.0163, 0.0240, 0.0223, 0.0152], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:34:48,549 INFO [train.py:901] (0/2) Epoch 49, batch 1400, loss[loss=0.1286, simple_loss=0.2074, pruned_loss=0.02493, over 7348.00 frames. ], tot_loss[loss=0.1266, simple_loss=0.2077, pruned_loss=0.02276, over 1441011.30 frames. ], batch size: 63, lr: 3.36e-03, grad_scale: 8.0 +2023-03-21 13:34:56,148 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.332e+02 1.767e+02 1.984e+02 2.313e+02 6.157e+02, threshold=3.967e+02, percent-clipped=1.0 +2023-03-21 13:35:05,319 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 13:35:08,418 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=136991.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:35:13,078 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=136999.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:35:15,393 INFO [train.py:901] (0/2) Epoch 49, batch 1450, loss[loss=0.1311, simple_loss=0.2229, pruned_loss=0.01969, over 7226.00 frames. ], tot_loss[loss=0.1266, simple_loss=0.2077, pruned_loss=0.02278, over 1439065.88 frames. ], batch size: 93, lr: 3.36e-03, grad_scale: 8.0 +2023-03-21 13:35:16,047 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.3310, 4.1350, 3.3595, 3.8485, 3.1757, 2.2004, 1.9316, 4.2013], + device='cuda:0'), covar=tensor([0.0047, 0.0074, 0.0148, 0.0071, 0.0184, 0.0646, 0.0728, 0.0052], + device='cuda:0'), in_proj_covar=tensor([0.0119, 0.0094, 0.0117, 0.0100, 0.0134, 0.0135, 0.0130, 0.0107], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 13:35:16,599 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4729, 1.9646, 2.2160, 3.5937, 1.9042, 3.4100, 1.3511, 3.2669], + device='cuda:0'), covar=tensor([0.0263, 0.2075, 0.2360, 0.0264, 0.4772, 0.0354, 0.1691, 0.0478], + device='cuda:0'), in_proj_covar=tensor([0.0208, 0.0240, 0.0250, 0.0210, 0.0245, 0.0217, 0.0215, 0.0227], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:35:24,056 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.3873, 2.0269, 2.3047, 3.5631, 1.9193, 3.4088, 1.3510, 3.2546], + device='cuda:0'), covar=tensor([0.0212, 0.1897, 0.2070, 0.0294, 0.4137, 0.0357, 0.1453, 0.0473], + device='cuda:0'), in_proj_covar=tensor([0.0207, 0.0240, 0.0250, 0.0210, 0.0245, 0.0217, 0.0214, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:35:29,937 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 13:35:33,541 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=137039.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:35:40,604 INFO [train.py:901] (0/2) Epoch 49, batch 1500, loss[loss=0.1369, simple_loss=0.2156, pruned_loss=0.02903, over 7255.00 frames. ], tot_loss[loss=0.1266, simple_loss=0.2075, pruned_loss=0.02281, over 1440596.02 frames. ], batch size: 89, lr: 3.36e-03, grad_scale: 8.0 +2023-03-21 13:35:46,921 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 13:35:48,435 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.149e+02 1.668e+02 1.982e+02 2.289e+02 3.254e+02, threshold=3.964e+02, percent-clipped=0.0 +2023-03-21 13:36:07,421 INFO [train.py:901] (0/2) Epoch 49, batch 1550, loss[loss=0.134, simple_loss=0.219, pruned_loss=0.02448, over 7347.00 frames. ], tot_loss[loss=0.1263, simple_loss=0.207, pruned_loss=0.02283, over 1439581.37 frames. ], batch size: 51, lr: 3.36e-03, grad_scale: 8.0 +2023-03-21 13:36:11,984 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 13:36:33,346 INFO [train.py:901] (0/2) Epoch 49, batch 1600, loss[loss=0.1439, simple_loss=0.2273, pruned_loss=0.03019, over 6826.00 frames. ], tot_loss[loss=0.1261, simple_loss=0.2067, pruned_loss=0.02278, over 1438117.77 frames. ], batch size: 107, lr: 3.36e-03, grad_scale: 8.0 +2023-03-21 13:36:34,000 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2559, 1.9958, 2.4024, 3.2906, 1.8442, 3.2908, 1.4706, 3.1052], + device='cuda:0'), covar=tensor([0.0281, 0.1730, 0.1863, 0.0388, 0.4473, 0.0366, 0.1392, 0.0557], + device='cuda:0'), in_proj_covar=tensor([0.0207, 0.0238, 0.0250, 0.0210, 0.0244, 0.0216, 0.0214, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:36:37,939 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=137162.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:36:40,361 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.158e+02 1.807e+02 2.128e+02 2.612e+02 4.654e+02, threshold=4.257e+02, percent-clipped=2.0 +2023-03-21 13:36:44,114 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 13:36:44,709 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9050, 3.0219, 3.8622, 3.8489, 3.8954, 3.9418, 3.8801, 3.8351], + device='cuda:0'), covar=tensor([0.0034, 0.0148, 0.0034, 0.0033, 0.0032, 0.0031, 0.0042, 0.0049], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0076, 0.0062, 0.0060, 0.0058, 0.0064, 0.0050, 0.0083], + device='cuda:0'), out_proj_covar=tensor([8.7851e-05, 1.4852e-04, 1.0809e-04, 9.9913e-05, 9.4269e-05, 1.0753e-04, + 9.1018e-05, 1.4922e-04], device='cuda:0') +2023-03-21 13:36:45,113 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 13:36:48,179 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 13:36:56,524 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.97 vs. limit=2.0 +2023-03-21 13:36:58,070 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 13:36:59,564 INFO [train.py:901] (0/2) Epoch 49, batch 1650, loss[loss=0.1324, simple_loss=0.2087, pruned_loss=0.02808, over 7277.00 frames. ], tot_loss[loss=0.1264, simple_loss=0.2071, pruned_loss=0.02282, over 1439751.13 frames. ], batch size: 57, lr: 3.36e-03, grad_scale: 8.0 +2023-03-21 13:37:02,107 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 13:37:04,245 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8176, 2.3385, 2.9147, 2.8635, 2.8559, 2.6922, 2.3760, 2.8529], + device='cuda:0'), covar=tensor([0.1534, 0.1130, 0.0973, 0.0901, 0.1012, 0.0990, 0.2105, 0.1200], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0073, 0.0054, 0.0053, 0.0054, 0.0052, 0.0071, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 13:37:10,657 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 13:37:25,331 INFO [train.py:901] (0/2) Epoch 49, batch 1700, loss[loss=0.1456, simple_loss=0.2141, pruned_loss=0.03852, over 7225.00 frames. ], tot_loss[loss=0.1267, simple_loss=0.2075, pruned_loss=0.02295, over 1441031.19 frames. ], batch size: 45, lr: 3.36e-03, grad_scale: 8.0 +2023-03-21 13:37:27,603 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0255, 4.2445, 3.2543, 4.3178, 3.9278, 4.1021, 2.4206, 3.1914], + device='cuda:0'), covar=tensor([0.0619, 0.0834, 0.2353, 0.0479, 0.0582, 0.0866, 0.3593, 0.2032], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0257, 0.0273, 0.0266, 0.0264, 0.0262, 0.0225, 0.0252], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:37:28,488 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 13:37:33,087 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.251e+02 1.788e+02 2.140e+02 2.383e+02 3.723e+02, threshold=4.279e+02, percent-clipped=0.0 +2023-03-21 13:37:33,635 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 13:37:43,304 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 13:37:49,451 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=137299.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:37:51,369 INFO [train.py:901] (0/2) Epoch 49, batch 1750, loss[loss=0.12, simple_loss=0.211, pruned_loss=0.01445, over 7222.00 frames. ], tot_loss[loss=0.1272, simple_loss=0.2083, pruned_loss=0.02299, over 1441955.51 frames. ], batch size: 93, lr: 3.36e-03, grad_scale: 8.0 +2023-03-21 13:38:02,519 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8675, 5.3648, 5.4443, 5.3749, 5.1871, 4.8925, 5.4457, 5.2583], + device='cuda:0'), covar=tensor([0.0423, 0.0322, 0.0339, 0.0452, 0.0343, 0.0361, 0.0294, 0.0396], + device='cuda:0'), in_proj_covar=tensor([0.0223, 0.0273, 0.0214, 0.0213, 0.0164, 0.0241, 0.0223, 0.0153], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:38:07,025 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 13:38:08,582 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 13:38:13,862 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.4585, 4.8813, 4.9767, 4.9008, 4.8261, 4.4327, 4.9664, 4.7638], + device='cuda:0'), covar=tensor([0.0431, 0.0337, 0.0342, 0.0452, 0.0330, 0.0390, 0.0308, 0.0490], + device='cuda:0'), in_proj_covar=tensor([0.0224, 0.0273, 0.0214, 0.0213, 0.0164, 0.0242, 0.0223, 0.0153], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:38:14,845 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=137347.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:38:17,835 INFO [train.py:901] (0/2) Epoch 49, batch 1800, loss[loss=0.1309, simple_loss=0.2204, pruned_loss=0.02071, over 7313.00 frames. ], tot_loss[loss=0.1265, simple_loss=0.2081, pruned_loss=0.02247, over 1440340.67 frames. ], batch size: 80, lr: 3.36e-03, grad_scale: 8.0 +2023-03-21 13:38:25,038 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.223e+02 1.722e+02 1.981e+02 2.424e+02 4.564e+02, threshold=3.962e+02, percent-clipped=1.0 +2023-03-21 13:38:29,554 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 13:38:42,620 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 13:38:44,223 INFO [train.py:901] (0/2) Epoch 49, batch 1850, loss[loss=0.1307, simple_loss=0.2088, pruned_loss=0.02626, over 7322.00 frames. ], tot_loss[loss=0.1271, simple_loss=0.2087, pruned_loss=0.02275, over 1443053.80 frames. ], batch size: 59, lr: 3.36e-03, grad_scale: 8.0 +2023-03-21 13:38:52,773 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 13:38:53,435 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.5849, 1.6551, 1.3712, 1.5724, 1.6235, 1.6873, 1.5911, 1.4224], + device='cuda:0'), covar=tensor([0.0239, 0.0216, 0.0324, 0.0243, 0.0199, 0.0244, 0.0197, 0.0213], + device='cuda:0'), in_proj_covar=tensor([0.0042, 0.0041, 0.0040, 0.0040, 0.0039, 0.0038, 0.0040, 0.0050], + device='cuda:0'), out_proj_covar=tensor([4.6999e-05, 4.5629e-05, 4.4526e-05, 4.4385e-05, 4.2562e-05, 4.2374e-05, + 4.4716e-05, 5.4316e-05], device='cuda:0') +2023-03-21 13:39:07,673 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6984, 3.8934, 3.6945, 3.8371, 3.5020, 3.8423, 4.1382, 4.1451], + device='cuda:0'), covar=tensor([0.0221, 0.0153, 0.0233, 0.0176, 0.0327, 0.0347, 0.0220, 0.0181], + device='cuda:0'), in_proj_covar=tensor([0.0128, 0.0130, 0.0125, 0.0128, 0.0116, 0.0104, 0.0101, 0.0105], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 13:39:09,622 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 13:39:10,089 INFO [train.py:901] (0/2) Epoch 49, batch 1900, loss[loss=0.1225, simple_loss=0.2079, pruned_loss=0.01854, over 7233.00 frames. ], tot_loss[loss=0.1272, simple_loss=0.2091, pruned_loss=0.02267, over 1443531.06 frames. ], batch size: 93, lr: 3.36e-03, grad_scale: 8.0 +2023-03-21 13:39:14,738 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=137462.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:39:17,243 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.231e+02 1.672e+02 1.883e+02 2.145e+02 4.351e+02, threshold=3.765e+02, percent-clipped=1.0 +2023-03-21 13:39:35,130 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 13:39:36,105 INFO [train.py:901] (0/2) Epoch 49, batch 1950, loss[loss=0.1222, simple_loss=0.2079, pruned_loss=0.01827, over 7294.00 frames. ], tot_loss[loss=0.127, simple_loss=0.2084, pruned_loss=0.02274, over 1441736.16 frames. ], batch size: 49, lr: 3.36e-03, grad_scale: 8.0 +2023-03-21 13:39:39,821 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=137510.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:39:46,584 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 13:39:51,262 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 13:39:51,774 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 13:40:02,189 INFO [train.py:901] (0/2) Epoch 49, batch 2000, loss[loss=0.1055, simple_loss=0.1804, pruned_loss=0.01525, over 6995.00 frames. ], tot_loss[loss=0.1277, simple_loss=0.2094, pruned_loss=0.02302, over 1442562.81 frames. ], batch size: 35, lr: 3.36e-03, grad_scale: 8.0 +2023-03-21 13:40:07,131 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 13:40:09,123 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.480e+02 1.776e+02 2.003e+02 2.327e+02 4.090e+02, threshold=4.007e+02, percent-clipped=2.0 +2023-03-21 13:40:18,396 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 13:40:20,210 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.58 vs. limit=2.0 +2023-03-21 13:40:26,626 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7783, 3.1033, 3.7285, 3.9166, 3.9333, 3.8544, 3.7896, 3.6950], + device='cuda:0'), covar=tensor([0.0035, 0.0135, 0.0037, 0.0027, 0.0027, 0.0031, 0.0059, 0.0056], + device='cuda:0'), in_proj_covar=tensor([0.0053, 0.0076, 0.0063, 0.0060, 0.0058, 0.0064, 0.0050, 0.0084], + device='cuda:0'), out_proj_covar=tensor([8.8091e-05, 1.4960e-04, 1.0864e-04, 1.0049e-04, 9.4006e-05, 1.0761e-04, + 9.1934e-05, 1.5017e-04], device='cuda:0') +2023-03-21 13:40:27,411 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 13:40:28,911 INFO [train.py:901] (0/2) Epoch 49, batch 2050, loss[loss=0.1189, simple_loss=0.2034, pruned_loss=0.0172, over 7307.00 frames. ], tot_loss[loss=0.1275, simple_loss=0.2088, pruned_loss=0.0231, over 1443461.31 frames. ], batch size: 49, lr: 3.36e-03, grad_scale: 8.0 +2023-03-21 13:40:54,342 INFO [train.py:901] (0/2) Epoch 49, batch 2100, loss[loss=0.1271, simple_loss=0.2109, pruned_loss=0.02167, over 7299.00 frames. ], tot_loss[loss=0.127, simple_loss=0.2085, pruned_loss=0.0228, over 1445297.23 frames. ], batch size: 66, lr: 3.35e-03, grad_scale: 8.0 +2023-03-21 13:41:00,128 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 13:41:00,659 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.9704, 4.5515, 4.3100, 4.9651, 4.7365, 4.8701, 4.3927, 4.5447], + device='cuda:0'), covar=tensor([0.0862, 0.2579, 0.2194, 0.0978, 0.0803, 0.1089, 0.0778, 0.1034], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0409, 0.0305, 0.0319, 0.0241, 0.0382, 0.0240, 0.0290], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 13:41:02,079 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.230e+02 1.716e+02 1.929e+02 2.374e+02 1.114e+03, threshold=3.858e+02, percent-clipped=2.0 +2023-03-21 13:41:02,904 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.40 vs. limit=2.0 +2023-03-21 13:41:03,104 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 13:41:18,445 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5700, 3.8253, 3.6325, 3.8001, 3.4406, 3.7348, 4.0493, 4.0659], + device='cuda:0'), covar=tensor([0.0258, 0.0168, 0.0234, 0.0157, 0.0336, 0.0383, 0.0227, 0.0183], + device='cuda:0'), in_proj_covar=tensor([0.0128, 0.0131, 0.0125, 0.0128, 0.0116, 0.0105, 0.0102, 0.0105], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 13:41:20,852 INFO [train.py:901] (0/2) Epoch 49, batch 2150, loss[loss=0.1237, simple_loss=0.2097, pruned_loss=0.01892, over 7271.00 frames. ], tot_loss[loss=0.1273, simple_loss=0.209, pruned_loss=0.02285, over 1447846.08 frames. ], batch size: 77, lr: 3.35e-03, grad_scale: 8.0 +2023-03-21 13:41:46,881 INFO [train.py:901] (0/2) Epoch 49, batch 2200, loss[loss=0.1407, simple_loss=0.2283, pruned_loss=0.02654, over 7321.00 frames. ], tot_loss[loss=0.1268, simple_loss=0.2082, pruned_loss=0.02267, over 1445750.50 frames. ], batch size: 59, lr: 3.35e-03, grad_scale: 8.0 +2023-03-21 13:41:49,967 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 13:41:54,586 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.272e+02 1.769e+02 2.125e+02 2.372e+02 3.697e+02, threshold=4.249e+02, percent-clipped=0.0 +2023-03-21 13:42:13,231 INFO [train.py:901] (0/2) Epoch 49, batch 2250, loss[loss=0.1194, simple_loss=0.2055, pruned_loss=0.01661, over 7295.00 frames. ], tot_loss[loss=0.1261, simple_loss=0.2076, pruned_loss=0.02227, over 1445579.27 frames. ], batch size: 80, lr: 3.35e-03, grad_scale: 16.0 +2023-03-21 13:42:22,192 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.56 vs. limit=5.0 +2023-03-21 13:42:22,948 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.0677, 4.2604, 4.1178, 4.2670, 3.9115, 4.2543, 4.6657, 4.6361], + device='cuda:0'), covar=tensor([0.0233, 0.0155, 0.0198, 0.0164, 0.0358, 0.0285, 0.0182, 0.0161], + device='cuda:0'), in_proj_covar=tensor([0.0128, 0.0131, 0.0125, 0.0129, 0.0116, 0.0105, 0.0102, 0.0104], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 13:42:24,919 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 13:42:25,428 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 13:42:32,340 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7790, 2.3522, 2.5904, 3.8466, 1.9811, 3.6580, 1.4580, 3.3849], + device='cuda:0'), covar=tensor([0.0267, 0.1747, 0.1967, 0.0273, 0.4748, 0.0316, 0.1562, 0.0501], + device='cuda:0'), in_proj_covar=tensor([0.0211, 0.0243, 0.0254, 0.0213, 0.0248, 0.0219, 0.0217, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:42:38,757 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 13:42:39,225 INFO [train.py:901] (0/2) Epoch 49, batch 2300, loss[loss=0.1469, simple_loss=0.2212, pruned_loss=0.03635, over 7218.00 frames. ], tot_loss[loss=0.126, simple_loss=0.2077, pruned_loss=0.02219, over 1444282.67 frames. ], batch size: 50, lr: 3.35e-03, grad_scale: 8.0 +2023-03-21 13:42:43,583 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8009, 4.0003, 3.8485, 3.9966, 3.6584, 3.9955, 4.3095, 4.3070], + device='cuda:0'), covar=tensor([0.0244, 0.0165, 0.0216, 0.0165, 0.0375, 0.0303, 0.0245, 0.0184], + device='cuda:0'), in_proj_covar=tensor([0.0128, 0.0131, 0.0125, 0.0129, 0.0116, 0.0105, 0.0102, 0.0105], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 13:42:47,627 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.190e+02 1.772e+02 2.010e+02 2.358e+02 5.104e+02, threshold=4.021e+02, percent-clipped=1.0 +2023-03-21 13:43:00,107 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=137892.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:43:01,562 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=137895.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:43:02,062 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7782, 1.9658, 1.6592, 1.8222, 1.9367, 1.8757, 1.8355, 1.5639], + device='cuda:0'), covar=tensor([0.0196, 0.0191, 0.0359, 0.0261, 0.0260, 0.0140, 0.0233, 0.0255], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0042, 0.0041, 0.0042, 0.0039, 0.0039, 0.0041, 0.0051], + device='cuda:0'), out_proj_covar=tensor([4.8032e-05, 4.6362e-05, 4.5563e-05, 4.5536e-05, 4.3462e-05, 4.3062e-05, + 4.5569e-05, 5.5379e-05], device='cuda:0') +2023-03-21 13:43:05,464 INFO [train.py:901] (0/2) Epoch 49, batch 2350, loss[loss=0.1088, simple_loss=0.1814, pruned_loss=0.0181, over 7024.00 frames. ], tot_loss[loss=0.1255, simple_loss=0.207, pruned_loss=0.02198, over 1442666.30 frames. ], batch size: 35, lr: 3.35e-03, grad_scale: 8.0 +2023-03-21 13:43:23,052 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2368, 1.9856, 2.2850, 3.3012, 1.8178, 3.2704, 1.3585, 3.0479], + device='cuda:0'), covar=tensor([0.0295, 0.1796, 0.2084, 0.0477, 0.4511, 0.0404, 0.1487, 0.0514], + device='cuda:0'), in_proj_covar=tensor([0.0211, 0.0243, 0.0255, 0.0213, 0.0248, 0.0220, 0.0217, 0.0230], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:43:23,223 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.05 vs. limit=2.0 +2023-03-21 13:43:28,199 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 13:43:32,245 INFO [train.py:901] (0/2) Epoch 49, batch 2400, loss[loss=0.1311, simple_loss=0.2201, pruned_loss=0.02103, over 7313.00 frames. ], tot_loss[loss=0.1255, simple_loss=0.2071, pruned_loss=0.02195, over 1444235.50 frames. ], batch size: 80, lr: 3.35e-03, grad_scale: 8.0 +2023-03-21 13:43:32,381 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=137953.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:43:33,822 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=137956.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:43:34,178 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 13:43:39,739 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.383e+02 1.856e+02 2.125e+02 2.479e+02 3.936e+02, threshold=4.250e+02, percent-clipped=0.0 +2023-03-21 13:43:44,376 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 13:43:47,505 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 13:43:58,066 INFO [train.py:901] (0/2) Epoch 49, batch 2450, loss[loss=0.1341, simple_loss=0.2101, pruned_loss=0.02898, over 7258.00 frames. ], tot_loss[loss=0.1255, simple_loss=0.2072, pruned_loss=0.02195, over 1444531.62 frames. ], batch size: 64, lr: 3.35e-03, grad_scale: 8.0 +2023-03-21 13:44:14,549 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 13:44:15,703 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7516, 2.5748, 2.6495, 3.7989, 2.1181, 3.6134, 1.4717, 3.2686], + device='cuda:0'), covar=tensor([0.0258, 0.1678, 0.1967, 0.0279, 0.4459, 0.0352, 0.1519, 0.0714], + device='cuda:0'), in_proj_covar=tensor([0.0212, 0.0244, 0.0255, 0.0213, 0.0249, 0.0220, 0.0218, 0.0231], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:44:24,607 INFO [train.py:901] (0/2) Epoch 49, batch 2500, loss[loss=0.1011, simple_loss=0.1712, pruned_loss=0.01553, over 6115.00 frames. ], tot_loss[loss=0.1253, simple_loss=0.2068, pruned_loss=0.02193, over 1442363.52 frames. ], batch size: 26, lr: 3.35e-03, grad_scale: 8.0 +2023-03-21 13:44:32,197 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.157e+02 1.701e+02 1.944e+02 2.363e+02 6.757e+02, threshold=3.888e+02, percent-clipped=1.0 +2023-03-21 13:44:39,663 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 13:44:50,363 INFO [train.py:901] (0/2) Epoch 49, batch 2550, loss[loss=0.1471, simple_loss=0.2218, pruned_loss=0.03616, over 7151.00 frames. ], tot_loss[loss=0.1253, simple_loss=0.2069, pruned_loss=0.02187, over 1442744.50 frames. ], batch size: 98, lr: 3.35e-03, grad_scale: 8.0 +2023-03-21 13:45:03,927 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=2.87 vs. limit=5.0 +2023-03-21 13:45:05,817 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6392, 1.7862, 1.5507, 1.7183, 1.7607, 1.7875, 1.6573, 1.5277], + device='cuda:0'), covar=tensor([0.0221, 0.0214, 0.0292, 0.0224, 0.0150, 0.0163, 0.0255, 0.0200], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0042, 0.0040, 0.0041, 0.0039, 0.0039, 0.0041, 0.0050], + device='cuda:0'), out_proj_covar=tensor([4.7626e-05, 4.5939e-05, 4.5282e-05, 4.5216e-05, 4.2906e-05, 4.2840e-05, + 4.5657e-05, 5.5004e-05], device='cuda:0') +2023-03-21 13:45:15,029 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.04 vs. limit=5.0 +2023-03-21 13:45:16,217 INFO [train.py:901] (0/2) Epoch 49, batch 2600, loss[loss=0.1361, simple_loss=0.2085, pruned_loss=0.03183, over 7222.00 frames. ], tot_loss[loss=0.1259, simple_loss=0.2074, pruned_loss=0.0222, over 1444046.77 frames. ], batch size: 50, lr: 3.35e-03, grad_scale: 8.0 +2023-03-21 13:45:22,976 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.58 vs. limit=2.0 +2023-03-21 13:45:23,532 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.329e+02 1.730e+02 2.027e+02 2.296e+02 4.292e+02, threshold=4.053e+02, percent-clipped=1.0 +2023-03-21 13:45:41,145 INFO [train.py:901] (0/2) Epoch 49, batch 2650, loss[loss=0.1249, simple_loss=0.2125, pruned_loss=0.01867, over 7215.00 frames. ], tot_loss[loss=0.1258, simple_loss=0.2075, pruned_loss=0.022, over 1445375.97 frames. ], batch size: 93, lr: 3.35e-03, grad_scale: 8.0 +2023-03-21 13:46:04,086 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=138248.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:46:05,607 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=138251.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:46:06,534 INFO [train.py:901] (0/2) Epoch 49, batch 2700, loss[loss=0.12, simple_loss=0.2078, pruned_loss=0.01613, over 7278.00 frames. ], tot_loss[loss=0.1253, simple_loss=0.2068, pruned_loss=0.02188, over 1442822.20 frames. ], batch size: 70, lr: 3.35e-03, grad_scale: 8.0 +2023-03-21 13:46:09,093 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.4355, 2.1907, 2.4162, 3.5299, 1.9009, 3.2785, 1.4931, 3.2165], + device='cuda:0'), covar=tensor([0.0243, 0.1752, 0.1937, 0.0309, 0.4190, 0.0254, 0.1402, 0.0552], + device='cuda:0'), in_proj_covar=tensor([0.0210, 0.0243, 0.0254, 0.0212, 0.0247, 0.0218, 0.0216, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:46:13,887 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.373e+02 1.750e+02 2.007e+02 2.310e+02 3.758e+02, threshold=4.014e+02, percent-clipped=0.0 +2023-03-21 13:46:15,015 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=138270.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:46:31,349 INFO [train.py:901] (0/2) Epoch 49, batch 2750, loss[loss=0.1345, simple_loss=0.2108, pruned_loss=0.02911, over 7321.00 frames. ], tot_loss[loss=0.1254, simple_loss=0.207, pruned_loss=0.02195, over 1444677.69 frames. ], batch size: 49, lr: 3.35e-03, grad_scale: 8.0 +2023-03-21 13:46:40,816 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=138322.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:46:45,779 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=138331.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:46:56,602 INFO [train.py:901] (0/2) Epoch 49, batch 2800, loss[loss=0.1231, simple_loss=0.2038, pruned_loss=0.02117, over 7276.00 frames. ], tot_loss[loss=0.1261, simple_loss=0.2077, pruned_loss=0.02221, over 1445089.32 frames. ], batch size: 70, lr: 3.35e-03, grad_scale: 8.0 +2023-03-21 13:47:04,003 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.135e+02 1.824e+02 2.150e+02 2.654e+02 5.945e+02, threshold=4.299e+02, percent-clipped=1.0 +2023-03-21 13:47:09,281 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-49.pt +2023-03-21 13:47:21,655 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0445-102620-0_sp0.9 from training. Duration: 13.3300625 +2023-03-21 13:47:21,711 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0343W0353-107668-0_sp0.9 from training. Duration: 12.0068125 +2023-03-21 13:47:21,735 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0_sp0.9 from training. Duration: 13.7855625 +2023-03-21 13:47:21,753 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0322-35834-0_sp0.9 from training. Duration: 12.7411875 +2023-03-21 13:47:21,933 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp1.1 from training. Duration: 13.21025 +2023-03-21 13:47:21,948 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0174W0255-47639-0_sp0.9 from training. Duration: 12.394375 +2023-03-21 13:47:21,960 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0431-52838-0_sp0.9 from training. Duration: 12.390125 +2023-03-21 13:47:22,059 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0123-40756-0_sp0.9 from training. Duration: 12.1100625 +2023-03-21 13:47:22,140 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0233-86222-0_sp0.9 from training. Duration: 13.103375 +2023-03-21 13:47:22,155 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0 from training. Duration: 12.021 +2023-03-21 13:47:22,172 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0495-41823-0_sp0.9 from training. Duration: 12.204875 +2023-03-21 13:47:22,693 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0107W0142-64458-0_sp0.9 from training. Duration: 12.61225 +2023-03-21 13:47:22,699 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0 from training. Duration: 12.1599375 +2023-03-21 13:47:22,707 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0466-31670-0_sp0.9 from training. Duration: 12.250125 +2023-03-21 13:47:22,851 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0126W0130-17071-0_sp0.9 from training. Duration: 12.6745 +2023-03-21 13:47:22,855 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0355-15040-0_sp0.9 from training. Duration: 13.2076875 +2023-03-21 13:47:22,914 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0 from training. Duration: 12.3199375 +2023-03-21 13:47:25,321 INFO [train.py:901] (0/2) Epoch 50, batch 0, loss[loss=0.1386, simple_loss=0.2248, pruned_loss=0.02621, over 7227.00 frames. ], tot_loss[loss=0.1386, simple_loss=0.2248, pruned_loss=0.02621, over 7227.00 frames. ], batch size: 93, lr: 3.31e-03, grad_scale: 8.0 +2023-03-21 13:47:25,322 INFO [train.py:926] (0/2) Computing validation loss +2023-03-21 13:47:32,530 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7456, 2.9619, 2.3205, 3.2742, 3.0653, 3.1675, 2.8715, 2.7188], + device='cuda:0'), covar=tensor([0.2024, 0.1201, 0.4281, 0.0718, 0.0374, 0.0296, 0.0415, 0.0438], + device='cuda:0'), in_proj_covar=tensor([0.0246, 0.0230, 0.0242, 0.0252, 0.0203, 0.0204, 0.0219, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:47:35,791 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.7845, 2.9107, 2.3130, 3.1960, 2.9168, 3.1380, 2.7527, 2.6060], + device='cuda:0'), covar=tensor([0.1914, 0.1187, 0.4138, 0.0589, 0.0362, 0.0273, 0.0378, 0.0388], + device='cuda:0'), in_proj_covar=tensor([0.0246, 0.0230, 0.0242, 0.0252, 0.0203, 0.0204, 0.0219, 0.0226], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:47:38,102 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5729, 4.2459, 3.9003, 4.6545, 4.3092, 4.6599, 4.4470, 4.3718], + device='cuda:0'), covar=tensor([0.0809, 0.2371, 0.2017, 0.1193, 0.0795, 0.1004, 0.0513, 0.0956], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0409, 0.0304, 0.0321, 0.0243, 0.0382, 0.0240, 0.0291], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 13:47:51,871 INFO [train.py:935] (0/2) Epoch 50, validation: loss=0.1661, simple_loss=0.2589, pruned_loss=0.03663, over 1622729.00 frames. +2023-03-21 13:47:51,872 INFO [train.py:936] (0/2) Maximum memory allocated so far is 12538MB +2023-03-21 13:47:55,103 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=138383.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:47:58,484 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0472-52669-0_sp0.9 from training. Duration: 12.9869375 +2023-03-21 13:48:09,979 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0377-52839-0_sp0.9 from training. Duration: 12.7031875 +2023-03-21 13:48:17,091 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0214-86176-0_sp0.9 from training. Duration: 13.6888125 +2023-03-21 13:48:17,612 INFO [train.py:901] (0/2) Epoch 50, batch 50, loss[loss=0.1396, simple_loss=0.2097, pruned_loss=0.03478, over 7310.00 frames. ], tot_loss[loss=0.1265, simple_loss=0.2074, pruned_loss=0.02281, over 322990.35 frames. ], batch size: 49, lr: 3.31e-03, grad_scale: 8.0 +2023-03-21 13:48:19,168 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0712W0259-53832-0_sp0.9 from training. Duration: 12.17325 +2023-03-21 13:48:22,158 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0468-86267-0_sp0.9 from training. Duration: 13.0943125 +2023-03-21 13:48:30,508 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9694, 3.1470, 2.1884, 3.3111, 2.6356, 3.0541, 1.4670, 2.4169], + device='cuda:0'), covar=tensor([0.0723, 0.1150, 0.3445, 0.0835, 0.0632, 0.0883, 0.4674, 0.2132], + device='cuda:0'), in_proj_covar=tensor([0.0273, 0.0255, 0.0272, 0.0265, 0.0264, 0.0261, 0.0224, 0.0250], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:48:39,556 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.230e+02 1.699e+02 2.026e+02 2.461e+02 4.880e+02, threshold=4.052e+02, percent-clipped=2.0 +2023-03-21 13:48:40,610 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0127-86359-0_sp0.9 from training. Duration: 12.76 +2023-03-21 13:48:41,106 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0363-35919-0 from training. Duration: 12.407 +2023-03-21 13:48:44,161 INFO [train.py:901] (0/2) Epoch 50, batch 100, loss[loss=0.1357, simple_loss=0.2168, pruned_loss=0.02726, over 7281.00 frames. ], tot_loss[loss=0.1263, simple_loss=0.2071, pruned_loss=0.02278, over 572803.26 frames. ], batch size: 77, lr: 3.31e-03, grad_scale: 8.0 +2023-03-21 13:48:56,019 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.5048, 2.6447, 2.7941, 2.5753, 2.8723, 2.5641, 2.5242, 2.1148], + device='cuda:0'), covar=tensor([0.0565, 0.0528, 0.0277, 0.0327, 0.0429, 0.0411, 0.0312, 0.0497], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0043, 0.0043, 0.0042, 0.0040, 0.0041, 0.0046, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 13:49:10,099 INFO [train.py:901] (0/2) Epoch 50, batch 150, loss[loss=0.1191, simple_loss=0.2054, pruned_loss=0.01636, over 7240.00 frames. ], tot_loss[loss=0.1253, simple_loss=0.2069, pruned_loss=0.02186, over 765475.90 frames. ], batch size: 93, lr: 3.31e-03, grad_scale: 8.0 +2023-03-21 13:49:20,152 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6186, 2.5136, 2.7397, 2.4961, 2.7191, 2.5151, 2.4630, 2.2707], + device='cuda:0'), covar=tensor([0.0494, 0.0737, 0.0406, 0.0423, 0.0627, 0.0672, 0.0417, 0.0306], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0043, 0.0043, 0.0042, 0.0040, 0.0041, 0.0046, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 13:49:20,597 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=138548.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:49:22,671 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=138551.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:49:31,278 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.281e+02 1.689e+02 1.944e+02 2.232e+02 3.987e+02, threshold=3.888e+02, percent-clipped=0.0 +2023-03-21 13:49:35,861 INFO [train.py:901] (0/2) Epoch 50, batch 200, loss[loss=0.1415, simple_loss=0.2118, pruned_loss=0.0356, over 7220.00 frames. ], tot_loss[loss=0.1258, simple_loss=0.2076, pruned_loss=0.02197, over 915483.00 frames. ], batch size: 45, lr: 3.31e-03, grad_scale: 8.0 +2023-03-21 13:49:39,405 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0_sp0.9 from training. Duration: 13.503375 +2023-03-21 13:49:43,931 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0480-86356-0_sp0.9 from training. Duration: 12.22225 +2023-03-21 13:49:45,502 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=138596.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:49:46,935 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=138599.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:49:50,279 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0 from training. Duration: 12.3249375 +2023-03-21 13:50:01,368 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=138626.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:50:01,803 INFO [train.py:901] (0/2) Epoch 50, batch 250, loss[loss=0.1322, simple_loss=0.2147, pruned_loss=0.0249, over 7326.00 frames. ], tot_loss[loss=0.1269, simple_loss=0.2087, pruned_loss=0.0225, over 1034821.87 frames. ], batch size: 75, lr: 3.31e-03, grad_scale: 8.0 +2023-03-21 13:50:03,338 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0 from training. Duration: 12.185 +2023-03-21 13:50:22,569 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8738, 3.8513, 3.0733, 3.5810, 3.0667, 2.1450, 1.9190, 3.9481], + device='cuda:0'), covar=tensor([0.0087, 0.0085, 0.0251, 0.0104, 0.0270, 0.0788, 0.0836, 0.0085], + device='cuda:0'), in_proj_covar=tensor([0.0121, 0.0096, 0.0118, 0.0102, 0.0136, 0.0137, 0.0133, 0.0109], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 13:50:22,912 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.259e+02 1.763e+02 2.018e+02 2.436e+02 6.098e+02, threshold=4.036e+02, percent-clipped=4.0 +2023-03-21 13:50:24,949 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0284-35956-0_sp0.9 from training. Duration: 12.4655625 +2023-03-21 13:50:27,517 INFO [train.py:901] (0/2) Epoch 50, batch 300, loss[loss=0.1138, simple_loss=0.2021, pruned_loss=0.01276, over 7284.00 frames. ], tot_loss[loss=0.1263, simple_loss=0.2079, pruned_loss=0.02231, over 1125491.23 frames. ], batch size: 66, lr: 3.31e-03, grad_scale: 8.0 +2023-03-21 13:50:28,106 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=138678.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:50:29,656 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=138681.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:50:32,132 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=138686.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:50:34,066 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0339W0266-38957-0_sp0.9 from training. Duration: 12.5200625 +2023-03-21 13:50:35,815 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.39 vs. limit=2.0 +2023-03-21 13:50:38,254 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=138697.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:50:53,987 INFO [train.py:901] (0/2) Epoch 50, batch 350, loss[loss=0.1333, simple_loss=0.2154, pruned_loss=0.02562, over 7302.00 frames. ], tot_loss[loss=0.1263, simple_loss=0.2078, pruned_loss=0.02241, over 1196449.47 frames. ], batch size: 80, lr: 3.31e-03, grad_scale: 8.0 +2023-03-21 13:51:01,717 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=138742.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:51:04,313 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=138747.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:51:08,201 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0249W0458-40379-0_sp0.9 from training. Duration: 12.348875 +2023-03-21 13:51:09,840 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=138758.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:51:14,668 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.024e+02 1.762e+02 2.045e+02 2.495e+02 3.731e+02, threshold=4.091e+02, percent-clipped=0.0 +2023-03-21 13:51:19,719 INFO [train.py:901] (0/2) Epoch 50, batch 400, loss[loss=0.1118, simple_loss=0.1879, pruned_loss=0.01785, over 7109.00 frames. ], tot_loss[loss=0.1262, simple_loss=0.2075, pruned_loss=0.02241, over 1250081.43 frames. ], batch size: 41, lr: 3.31e-03, grad_scale: 8.0 +2023-03-21 13:51:22,911 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9983, 3.2239, 2.3789, 3.5708, 2.6274, 3.0869, 1.6068, 2.6000], + device='cuda:0'), covar=tensor([0.0575, 0.1192, 0.3046, 0.0611, 0.0491, 0.0776, 0.4418, 0.1893], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0256, 0.0274, 0.0265, 0.0264, 0.0263, 0.0224, 0.0251], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:51:26,334 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.5939, 4.2853, 4.1635, 4.0429, 3.7126, 2.5420, 2.2830, 4.5469], + device='cuda:0'), covar=tensor([0.0045, 0.0088, 0.0072, 0.0070, 0.0115, 0.0541, 0.0574, 0.0046], + device='cuda:0'), in_proj_covar=tensor([0.0122, 0.0095, 0.0119, 0.0102, 0.0136, 0.0137, 0.0133, 0.0109], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 13:51:34,878 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4318, 2.3309, 2.5473, 2.3860, 2.6890, 2.3948, 2.3908, 2.0589], + device='cuda:0'), covar=tensor([0.0658, 0.0692, 0.0386, 0.0384, 0.0504, 0.0484, 0.0322, 0.0373], + device='cuda:0'), in_proj_covar=tensor([0.0045, 0.0044, 0.0044, 0.0043, 0.0041, 0.0041, 0.0047, 0.0048], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 13:51:45,917 INFO [train.py:901] (0/2) Epoch 50, batch 450, loss[loss=0.1502, simple_loss=0.2211, pruned_loss=0.03968, over 7233.00 frames. ], tot_loss[loss=0.126, simple_loss=0.2073, pruned_loss=0.02238, over 1292989.75 frames. ], batch size: 45, lr: 3.31e-03, grad_scale: 8.0 +2023-03-21 13:51:47,630 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.6588, 2.9552, 2.1760, 3.2052, 2.2521, 2.7187, 1.4257, 2.3459], + device='cuda:0'), covar=tensor([0.0534, 0.1025, 0.2956, 0.0778, 0.0657, 0.0701, 0.4456, 0.1907], + device='cuda:0'), in_proj_covar=tensor([0.0275, 0.0256, 0.0274, 0.0266, 0.0265, 0.0263, 0.0225, 0.0251], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:51:50,430 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0657W0266-27035-0_sp0.9 from training. Duration: 12.0923125 +2023-03-21 13:51:50,447 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0_sp0.9 from training. Duration: 13.955625 +2023-03-21 13:51:54,807 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.91 vs. limit=2.0 +2023-03-21 13:52:04,791 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8711, 1.5166, 2.1890, 2.3729, 2.1115, 2.2821, 1.9236, 2.2276], + device='cuda:0'), covar=tensor([0.2242, 0.5727, 0.1169, 0.0977, 0.1556, 0.2022, 0.2233, 0.2376], + device='cuda:0'), in_proj_covar=tensor([0.0087, 0.0089, 0.0081, 0.0071, 0.0070, 0.0070, 0.0113, 0.0072], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:52:07,176 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.180e+02 1.662e+02 2.041e+02 2.387e+02 3.807e+02, threshold=4.081e+02, percent-clipped=0.0 +2023-03-21 13:52:11,761 INFO [train.py:901] (0/2) Epoch 50, batch 500, loss[loss=0.1317, simple_loss=0.2189, pruned_loss=0.02228, over 7249.00 frames. ], tot_loss[loss=0.1255, simple_loss=0.2072, pruned_loss=0.02187, over 1325821.58 frames. ], batch size: 89, lr: 3.31e-03, grad_scale: 8.0 +2023-03-21 13:52:14,873 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.8135, 2.4831, 3.0773, 2.8426, 3.0478, 2.7799, 2.5230, 3.0135], + device='cuda:0'), covar=tensor([0.1596, 0.0838, 0.0886, 0.1400, 0.0647, 0.0946, 0.1998, 0.1030], + device='cuda:0'), in_proj_covar=tensor([0.0055, 0.0073, 0.0055, 0.0054, 0.0054, 0.0053, 0.0071, 0.0053], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 13:52:24,931 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0 from training. Duration: 13.031 +2023-03-21 13:52:26,409 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0269-85831-0_sp0.9 from training. Duration: 13.694375 +2023-03-21 13:52:26,919 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0421W0442-98380-0_sp0.9 from training. Duration: 13.187625 +2023-03-21 13:52:28,492 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=138909.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:52:29,422 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0178W0358-85850-0_sp0.9 from training. Duration: 13.2424375 +2023-03-21 13:52:32,954 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0 from training. Duration: 14.53125 +2023-03-21 13:52:36,999 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=138926.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:52:37,408 INFO [train.py:901] (0/2) Epoch 50, batch 550, loss[loss=0.1174, simple_loss=0.2047, pruned_loss=0.01507, over 7286.00 frames. ], tot_loss[loss=0.1258, simple_loss=0.2077, pruned_loss=0.02196, over 1353430.66 frames. ], batch size: 77, lr: 3.31e-03, grad_scale: 8.0 +2023-03-21 13:52:44,485 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0 from training. Duration: 12.062 +2023-03-21 13:52:52,422 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.51 vs. limit=2.0 +2023-03-21 13:52:52,534 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0723W0457-18562-0_sp0.9 from training. Duration: 12.26775 +2023-03-21 13:52:56,141 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0127W0333-18886-0_sp0.9 from training. Duration: 12.32125 +2023-03-21 13:52:56,988 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.86 vs. limit=2.0 +2023-03-21 13:52:58,590 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.346e+02 1.767e+02 2.083e+02 2.460e+02 6.205e+02, threshold=4.166e+02, percent-clipped=1.0 +2023-03-21 13:52:59,793 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=138970.0, num_to_drop=1, layers_to_drop={2} +2023-03-21 13:53:01,753 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=138974.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:53:03,248 INFO [train.py:901] (0/2) Epoch 50, batch 600, loss[loss=0.1137, simple_loss=0.1959, pruned_loss=0.01576, over 7335.00 frames. ], tot_loss[loss=0.1252, simple_loss=0.2069, pruned_loss=0.02174, over 1374199.60 frames. ], batch size: 75, lr: 3.31e-03, grad_scale: 8.0 +2023-03-21 13:53:03,263 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0227-102578-0_sp0.9 from training. Duration: 12.063375 +2023-03-21 13:53:03,852 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=138978.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:53:21,343 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0147-14735-0_sp0.9 from training. Duration: 13.0755 +2023-03-21 13:53:28,998 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=139026.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:53:29,453 INFO [train.py:901] (0/2) Epoch 50, batch 650, loss[loss=0.1281, simple_loss=0.2081, pruned_loss=0.02409, over 7338.00 frames. ], tot_loss[loss=0.125, simple_loss=0.2063, pruned_loss=0.02183, over 1389414.03 frames. ], batch size: 75, lr: 3.30e-03, grad_scale: 8.0 +2023-03-21 13:53:29,953 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0465-86257-0_sp0.9 from training. Duration: 12.8344375 +2023-03-21 13:53:34,556 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=139037.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:53:37,057 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=139042.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:53:43,234 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=139053.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:53:47,824 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0704W0234-59743-0_sp0.9 from training. Duration: 12.6989375 +2023-03-21 13:53:50,875 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.276e+02 1.832e+02 2.103e+02 2.420e+02 4.796e+02, threshold=4.205e+02, percent-clipped=1.0 +2023-03-21 13:53:56,184 INFO [train.py:901] (0/2) Epoch 50, batch 700, loss[loss=0.1246, simple_loss=0.2151, pruned_loss=0.01704, over 7121.00 frames. ], tot_loss[loss=0.1249, simple_loss=0.2063, pruned_loss=0.0217, over 1401051.64 frames. ], batch size: 98, lr: 3.30e-03, grad_scale: 8.0 +2023-03-21 13:53:56,200 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0312-45768-0_sp0.9 from training. Duration: 12.422125 +2023-03-21 13:54:20,593 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0173-32335-0_sp0.9 from training. Duration: 13.2133125 +2023-03-21 13:54:21,059 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0247-36021-0_sp0.9 from training. Duration: 13.40225 +2023-03-21 13:54:21,540 INFO [train.py:901] (0/2) Epoch 50, batch 750, loss[loss=0.1267, simple_loss=0.2039, pruned_loss=0.02476, over 7229.00 frames. ], tot_loss[loss=0.1252, simple_loss=0.2068, pruned_loss=0.02177, over 1411671.64 frames. ], batch size: 45, lr: 3.30e-03, grad_scale: 8.0 +2023-03-21 13:54:28,584 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.55 vs. limit=2.0 +2023-03-21 13:54:34,864 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0064W0442-15905-0_sp0.9 from training. Duration: 12.0798125 +2023-03-21 13:54:38,887 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0062W0390-112800-0_sp0.9 from training. Duration: 12.335625 +2023-03-21 13:54:41,508 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.49 vs. limit=2.0 +2023-03-21 13:54:43,655 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.310e+02 1.826e+02 2.099e+02 2.412e+02 4.667e+02, threshold=4.199e+02, percent-clipped=1.0 +2023-03-21 13:54:43,755 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2735, 4.7155, 4.8073, 4.7254, 4.6519, 4.2288, 4.7996, 4.6359], + device='cuda:0'), covar=tensor([0.0443, 0.0348, 0.0303, 0.0437, 0.0369, 0.0467, 0.0289, 0.0436], + device='cuda:0'), in_proj_covar=tensor([0.0224, 0.0275, 0.0214, 0.0210, 0.0164, 0.0241, 0.0222, 0.0151], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:54:45,232 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0 from training. Duration: 12.6340625 +2023-03-21 13:54:46,614 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp0.9 from training. Duration: 14.8566875 +2023-03-21 13:54:48,072 INFO [train.py:901] (0/2) Epoch 50, batch 800, loss[loss=0.1243, simple_loss=0.2014, pruned_loss=0.02361, over 7265.00 frames. ], tot_loss[loss=0.1256, simple_loss=0.2073, pruned_loss=0.02199, over 1417596.21 frames. ], batch size: 47, lr: 3.30e-03, grad_scale: 8.0 +2023-03-21 13:54:56,958 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.01 vs. limit=5.0 +2023-03-21 13:54:57,674 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0596W0136-102670-0_sp0.9 from training. Duration: 13.538875 +2023-03-21 13:55:14,433 INFO [train.py:901] (0/2) Epoch 50, batch 850, loss[loss=0.1259, simple_loss=0.2066, pruned_loss=0.02264, over 7279.00 frames. ], tot_loss[loss=0.1262, simple_loss=0.2079, pruned_loss=0.02223, over 1425577.10 frames. ], batch size: 70, lr: 3.30e-03, grad_scale: 8.0 +2023-03-21 13:55:14,518 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.8739, 5.3122, 5.3745, 5.3133, 5.1215, 4.8457, 5.4103, 5.2280], + device='cuda:0'), covar=tensor([0.0377, 0.0339, 0.0335, 0.0440, 0.0364, 0.0388, 0.0285, 0.0428], + device='cuda:0'), in_proj_covar=tensor([0.0224, 0.0275, 0.0215, 0.0211, 0.0164, 0.0241, 0.0222, 0.0152], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:55:16,908 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0_sp0.9 from training. Duration: 13.7823125 +2023-03-21 13:55:16,913 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0386-52709-0_sp0.9 from training. Duration: 12.77775 +2023-03-21 13:55:20,089 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=139238.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:55:22,562 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0078W0423-41503-0_sp0.9 from training. Duration: 12.9845 +2023-03-21 13:55:26,087 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0462-112684-0_sp0.9 from training. Duration: 12.15225 +2023-03-21 13:55:34,307 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=139265.0, num_to_drop=1, layers_to_drop={3} +2023-03-21 13:55:35,729 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.240e+02 1.836e+02 2.170e+02 2.399e+02 3.753e+02, threshold=4.340e+02, percent-clipped=0.0 +2023-03-21 13:55:39,953 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=139276.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:55:40,325 INFO [train.py:901] (0/2) Epoch 50, batch 900, loss[loss=0.1212, simple_loss=0.2055, pruned_loss=0.01848, over 7294.00 frames. ], tot_loss[loss=0.1259, simple_loss=0.2075, pruned_loss=0.02213, over 1428516.25 frames. ], batch size: 57, lr: 3.30e-03, grad_scale: 8.0 +2023-03-21 13:55:46,577 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=139289.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:55:51,661 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8853, 1.5268, 2.0823, 2.2086, 1.9728, 2.2184, 1.7064, 2.1172], + device='cuda:0'), covar=tensor([0.2784, 0.5633, 0.0958, 0.0926, 0.1132, 0.1630, 0.2265, 0.2399], + device='cuda:0'), in_proj_covar=tensor([0.0087, 0.0087, 0.0080, 0.0070, 0.0070, 0.0070, 0.0112, 0.0071], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:55:51,672 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=139299.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:56:03,812 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0119W0370-86573-0_sp0.9 from training. Duration: 13.3566875 +2023-03-21 13:56:06,356 INFO [train.py:901] (0/2) Epoch 50, batch 950, loss[loss=0.1442, simple_loss=0.2249, pruned_loss=0.03177, over 7341.00 frames. ], tot_loss[loss=0.1262, simple_loss=0.2076, pruned_loss=0.02242, over 1433206.38 frames. ], batch size: 54, lr: 3.30e-03, grad_scale: 8.0 +2023-03-21 13:56:11,573 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=139337.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:56:11,606 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=139337.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 13:56:14,685 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=139342.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:56:18,665 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=139350.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:56:20,199 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=139353.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:56:23,791 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.5565, 3.3850, 3.2118, 3.5155, 2.9721, 2.9365, 3.7330, 2.3990], + device='cuda:0'), covar=tensor([0.0687, 0.0867, 0.1114, 0.1049, 0.1053, 0.1473, 0.0911, 0.3667], + device='cuda:0'), in_proj_covar=tensor([0.0329, 0.0336, 0.0269, 0.0347, 0.0278, 0.0283, 0.0343, 0.0234], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:56:27,647 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.176e+02 1.735e+02 1.986e+02 2.308e+02 3.681e+02, threshold=3.972e+02, percent-clipped=0.0 +2023-03-21 13:56:28,719 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0_sp0.9 from training. Duration: 14.10775 +2023-03-21 13:56:32,159 INFO [train.py:901] (0/2) Epoch 50, batch 1000, loss[loss=0.1117, simple_loss=0.1977, pruned_loss=0.01287, over 7240.00 frames. ], tot_loss[loss=0.1258, simple_loss=0.2072, pruned_loss=0.02218, over 1434108.32 frames. ], batch size: 55, lr: 3.30e-03, grad_scale: 8.0 +2023-03-21 13:56:36,241 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=139385.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:56:38,761 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=139390.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:56:45,302 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=139401.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:56:49,803 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0027W0202-6154-0_sp0.9 from training. Duration: 12.862125 +2023-03-21 13:56:58,108 INFO [train.py:901] (0/2) Epoch 50, batch 1050, loss[loss=0.132, simple_loss=0.2152, pruned_loss=0.02443, over 7261.00 frames. ], tot_loss[loss=0.1262, simple_loss=0.2075, pruned_loss=0.02243, over 1435609.49 frames. ], batch size: 64, lr: 3.30e-03, grad_scale: 8.0 +2023-03-21 13:57:11,469 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0480-79636-0_sp0.9 from training. Duration: 14.478875 +2023-03-21 13:57:15,514 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0600W0231-469-0_sp0.9 from training. Duration: 12.9744375 +2023-03-21 13:57:19,433 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.213e+02 1.721e+02 1.925e+02 2.210e+02 3.439e+02, threshold=3.850e+02, percent-clipped=0.0 +2023-03-21 13:57:23,961 INFO [train.py:901] (0/2) Epoch 50, batch 1100, loss[loss=0.1266, simple_loss=0.2122, pruned_loss=0.02045, over 7244.00 frames. ], tot_loss[loss=0.1269, simple_loss=0.2082, pruned_loss=0.02277, over 1437290.65 frames. ], batch size: 89, lr: 3.30e-03, grad_scale: 8.0 +2023-03-21 13:57:38,982 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.1493, 3.4183, 2.9518, 3.4094, 3.3839, 3.1574, 3.2834, 3.4729], + device='cuda:0'), covar=tensor([0.0756, 0.0508, 0.1321, 0.1046, 0.0948, 0.0522, 0.0935, 0.0555], + device='cuda:0'), in_proj_covar=tensor([0.0062, 0.0063, 0.0071, 0.0063, 0.0058, 0.0067, 0.0060, 0.0057], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:57:45,154 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0007W0369-32499-0_sp0.9 from training. Duration: 12.25775 +2023-03-21 13:57:45,171 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0478-86278-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 13:57:50,647 INFO [train.py:901] (0/2) Epoch 50, batch 1150, loss[loss=0.1262, simple_loss=0.2017, pruned_loss=0.0254, over 7276.00 frames. ], tot_loss[loss=0.1269, simple_loss=0.2081, pruned_loss=0.02279, over 1438840.15 frames. ], batch size: 52, lr: 3.30e-03, grad_scale: 8.0 +2023-03-21 13:57:52,269 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.7098, 2.4452, 2.5921, 3.8151, 2.1322, 3.5435, 1.6823, 3.3583], + device='cuda:0'), covar=tensor([0.0239, 0.1770, 0.1965, 0.0289, 0.3884, 0.0297, 0.1414, 0.0430], + device='cuda:0'), in_proj_covar=tensor([0.0212, 0.0245, 0.0255, 0.0215, 0.0250, 0.0220, 0.0218, 0.0231], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 13:57:58,654 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0459-40844-0_sp0.9 from training. Duration: 13.11575 +2023-03-21 13:57:59,158 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0094W0460-59915-0_sp0.9 from training. Duration: 12.979125 +2023-03-21 13:58:09,869 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=139565.0, num_to_drop=1, layers_to_drop={1} +2023-03-21 13:58:11,261 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.066e+02 1.822e+02 2.028e+02 2.488e+02 3.805e+02, threshold=4.056e+02, percent-clipped=0.0 +2023-03-21 13:58:12,947 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=139571.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:58:15,000 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([5.7447, 5.3442, 5.0794, 5.7187, 5.5019, 5.6678, 5.1520, 5.3067], + device='cuda:0'), covar=tensor([0.0562, 0.1864, 0.1786, 0.0714, 0.0783, 0.0899, 0.0676, 0.1086], + device='cuda:0'), in_proj_covar=tensor([0.0271, 0.0403, 0.0303, 0.0316, 0.0240, 0.0379, 0.0238, 0.0286], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0003, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 13:58:16,481 INFO [train.py:901] (0/2) Epoch 50, batch 1200, loss[loss=0.1409, simple_loss=0.2224, pruned_loss=0.02974, over 7243.00 frames. ], tot_loss[loss=0.1269, simple_loss=0.208, pruned_loss=0.02287, over 1439682.83 frames. ], batch size: 55, lr: 3.30e-03, grad_scale: 8.0 +2023-03-21 13:58:19,202 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=139582.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:58:25,252 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=139594.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:58:33,676 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0361W0207-97658-0_sp0.9 from training. Duration: 12.3268125 +2023-03-21 13:58:35,789 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=139613.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:58:42,762 INFO [train.py:901] (0/2) Epoch 50, batch 1250, loss[loss=0.1304, simple_loss=0.2188, pruned_loss=0.02096, over 7289.00 frames. ], tot_loss[loss=0.1264, simple_loss=0.2078, pruned_loss=0.02249, over 1440215.48 frames. ], batch size: 68, lr: 3.30e-03, grad_scale: 8.0 +2023-03-21 13:58:45,392 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=139632.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 13:58:45,440 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=139632.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:58:51,068 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=139643.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:58:51,978 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=139645.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:58:55,952 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0251-86481-0_sp0.9 from training. Duration: 13.1033125 +2023-03-21 13:59:01,058 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0039W0307-49457-0 from training. Duration: 12.4040625 +2023-03-21 13:59:02,067 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0121-45607-0_sp0.9 from training. Duration: 14.037875 +2023-03-21 13:59:03,972 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.064e+02 1.670e+02 1.886e+02 2.283e+02 6.089e+02, threshold=3.771e+02, percent-clipped=1.0 +2023-03-21 13:59:08,574 INFO [train.py:901] (0/2) Epoch 50, batch 1300, loss[loss=0.1021, simple_loss=0.1825, pruned_loss=0.01084, over 7181.00 frames. ], tot_loss[loss=0.1266, simple_loss=0.2079, pruned_loss=0.02269, over 1439897.23 frames. ], batch size: 39, lr: 3.30e-03, grad_scale: 8.0 +2023-03-21 13:59:24,816 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0013W0486-15004-0_sp0.9 from training. Duration: 12.0100625 +2023-03-21 13:59:27,681 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0703W0471-112404-0_sp0.9 from training. Duration: 12.43325 +2023-03-21 13:59:31,177 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0715W0442-101321-0 from training. Duration: 12.697 +2023-03-21 13:59:34,262 INFO [train.py:901] (0/2) Epoch 50, batch 1350, loss[loss=0.1052, simple_loss=0.1796, pruned_loss=0.01541, over 7011.00 frames. ], tot_loss[loss=0.1255, simple_loss=0.2066, pruned_loss=0.0222, over 1439201.03 frames. ], batch size: 35, lr: 3.30e-03, grad_scale: 8.0 +2023-03-21 13:59:35,436 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9375, 3.5462, 3.3916, 3.8072, 3.2617, 3.1937, 3.8575, 2.6093], + device='cuda:0'), covar=tensor([0.0532, 0.0661, 0.0867, 0.0714, 0.0887, 0.1169, 0.0920, 0.3011], + device='cuda:0'), in_proj_covar=tensor([0.0327, 0.0335, 0.0268, 0.0346, 0.0277, 0.0283, 0.0344, 0.0234], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 13:59:36,952 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3601, 2.3563, 2.5049, 2.3313, 2.5624, 2.4763, 2.1771, 1.8327], + device='cuda:0'), covar=tensor([0.0382, 0.0382, 0.0252, 0.0292, 0.0366, 0.0468, 0.0417, 0.0480], + device='cuda:0'), in_proj_covar=tensor([0.0045, 0.0043, 0.0044, 0.0044, 0.0041, 0.0041, 0.0048, 0.0048], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 13:59:41,936 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0_sp0.9 from training. Duration: 13.930125 +2023-03-21 13:59:55,539 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.225e+02 1.698e+02 1.919e+02 2.403e+02 4.718e+02, threshold=3.837e+02, percent-clipped=1.0 +2023-03-21 13:59:56,140 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=139769.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 13:59:59,412 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=2.05 vs. limit=2.0 +2023-03-21 13:59:59,738 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.2418, 4.1735, 3.3106, 3.8293, 3.3418, 2.2516, 1.9260, 4.1943], + device='cuda:0'), covar=tensor([0.0051, 0.0065, 0.0178, 0.0093, 0.0177, 0.0650, 0.0761, 0.0060], + device='cuda:0'), in_proj_covar=tensor([0.0124, 0.0097, 0.0121, 0.0104, 0.0139, 0.0139, 0.0136, 0.0112], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0002, 0.0002, 0.0002, 0.0001], + device='cuda:0') +2023-03-21 14:00:00,123 INFO [train.py:901] (0/2) Epoch 50, batch 1400, loss[loss=0.1274, simple_loss=0.2064, pruned_loss=0.02421, over 7350.00 frames. ], tot_loss[loss=0.1251, simple_loss=0.2061, pruned_loss=0.02204, over 1440351.36 frames. ], batch size: 54, lr: 3.30e-03, grad_scale: 8.0 +2023-03-21 14:00:06,994 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=139789.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:00:13,988 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.8434, 4.0034, 3.6777, 3.9787, 3.5169, 3.8591, 4.2358, 4.2641], + device='cuda:0'), covar=tensor([0.0244, 0.0173, 0.0282, 0.0181, 0.0453, 0.0393, 0.0238, 0.0182], + device='cuda:0'), in_proj_covar=tensor([0.0128, 0.0130, 0.0125, 0.0129, 0.0116, 0.0104, 0.0101, 0.0104], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 14:00:15,043 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=139804.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:00:16,040 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=139806.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:00:16,432 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0149-113139-0_sp0.9 from training. Duration: 12.431125 +2023-03-21 14:00:21,317 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.52 vs. limit=2.0 +2023-03-21 14:00:26,412 INFO [train.py:901] (0/2) Epoch 50, batch 1450, loss[loss=0.1334, simple_loss=0.2174, pruned_loss=0.02475, over 7224.00 frames. ], tot_loss[loss=0.1252, simple_loss=0.2062, pruned_loss=0.02212, over 1440066.11 frames. ], batch size: 93, lr: 3.30e-03, grad_scale: 16.0 +2023-03-21 14:00:28,113 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=139830.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:00:38,795 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=139850.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:00:38,809 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=139850.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:00:40,271 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6987, 2.2487, 2.4865, 3.8142, 2.1123, 3.4492, 1.4863, 3.3235], + device='cuda:0'), covar=tensor([0.0289, 0.1925, 0.2311, 0.0311, 0.4163, 0.0385, 0.1602, 0.0594], + device='cuda:0'), in_proj_covar=tensor([0.0210, 0.0243, 0.0253, 0.0213, 0.0248, 0.0219, 0.0216, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 14:00:40,609 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0491-31830-0_sp0.9 from training. Duration: 13.241125 +2023-03-21 14:00:46,225 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=139865.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:00:47,226 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=139867.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:00:48,101 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.097e+02 1.906e+02 2.192e+02 2.677e+02 5.370e+02, threshold=4.384e+02, percent-clipped=2.0 +2023-03-21 14:00:52,793 INFO [train.py:901] (0/2) Epoch 50, batch 1500, loss[loss=0.1201, simple_loss=0.1925, pruned_loss=0.02386, over 7224.00 frames. ], tot_loss[loss=0.1257, simple_loss=0.2064, pruned_loss=0.02245, over 1437992.26 frames. ], batch size: 45, lr: 3.29e-03, grad_scale: 8.0 +2023-03-21 14:00:55,840 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp1.1 from training. Duration: 12.7053125 +2023-03-21 14:01:01,433 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=139894.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:01:10,146 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=139911.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:01:17,568 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.70 vs. limit=5.0 +2023-03-21 14:01:18,747 INFO [train.py:901] (0/2) Epoch 50, batch 1550, loss[loss=0.1175, simple_loss=0.1999, pruned_loss=0.01755, over 7288.00 frames. ], tot_loss[loss=0.126, simple_loss=0.207, pruned_loss=0.0225, over 1437537.70 frames. ], batch size: 77, lr: 3.29e-03, grad_scale: 8.0 +2023-03-21 14:01:18,819 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=139927.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:01:19,811 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0 from training. Duration: 13.371 +2023-03-21 14:01:21,415 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=139932.0, num_to_drop=1, layers_to_drop={0} +2023-03-21 14:01:24,318 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=139938.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:01:26,370 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=139942.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:01:27,888 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=139945.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:01:40,574 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.234e+02 1.753e+02 2.129e+02 2.590e+02 7.618e+02, threshold=4.259e+02, percent-clipped=2.0 +2023-03-21 14:01:44,743 INFO [train.py:901] (0/2) Epoch 50, batch 1600, loss[loss=0.128, simple_loss=0.2094, pruned_loss=0.02329, over 7278.00 frames. ], tot_loss[loss=0.1255, simple_loss=0.2065, pruned_loss=0.02229, over 1435401.91 frames. ], batch size: 77, lr: 3.29e-03, grad_scale: 8.0 +2023-03-21 14:01:46,301 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=139980.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:01:50,247 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0360W0244-86520-0_sp0.9 from training. Duration: 12.86675 +2023-03-21 14:01:50,735 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0434-40735-0 from training. Duration: 12.5600625 +2023-03-21 14:01:52,791 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=139993.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:01:54,708 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0_sp0.9 from training. Duration: 15.5286875 +2023-03-21 14:01:56,636 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/checkpoint-140000.pt +2023-03-21 14:02:07,480 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.6569, 2.9911, 3.2982, 3.5025, 2.9924, 3.1064, 3.6558, 2.4255], + device='cuda:0'), covar=tensor([0.0639, 0.0634, 0.1124, 0.0802, 0.0877, 0.1288, 0.0803, 0.3475], + device='cuda:0'), in_proj_covar=tensor([0.0328, 0.0336, 0.0270, 0.0347, 0.0278, 0.0285, 0.0345, 0.0234], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 14:02:08,734 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0 from training. Duration: 13.074 +2023-03-21 14:02:11,807 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0162W0314-19617-0 from training. Duration: 12.1530625 +2023-03-21 14:02:14,265 INFO [train.py:901] (0/2) Epoch 50, batch 1650, loss[loss=0.1318, simple_loss=0.2091, pruned_loss=0.02728, over 7246.00 frames. ], tot_loss[loss=0.1266, simple_loss=0.2078, pruned_loss=0.02276, over 1437947.16 frames. ], batch size: 55, lr: 3.29e-03, grad_scale: 8.0 +2023-03-21 14:02:20,767 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0023W0395-77944-0_sp0.9 from training. Duration: 13.130125 +2023-03-21 14:02:35,783 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.1010, 4.2727, 3.9954, 4.2281, 3.8634, 4.1193, 4.4910, 4.5422], + device='cuda:0'), covar=tensor([0.0209, 0.0144, 0.0225, 0.0150, 0.0343, 0.0312, 0.0210, 0.0166], + device='cuda:0'), in_proj_covar=tensor([0.0127, 0.0129, 0.0124, 0.0128, 0.0116, 0.0103, 0.0101, 0.0104], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 14:02:36,151 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.257e+02 1.886e+02 2.154e+02 2.450e+02 4.461e+02, threshold=4.308e+02, percent-clipped=1.0 +2023-03-21 14:02:37,719 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0713W0320-44227-0_sp0.9 from training. Duration: 12.868875 +2023-03-21 14:02:40,215 INFO [train.py:901] (0/2) Epoch 50, batch 1700, loss[loss=0.1218, simple_loss=0.2051, pruned_loss=0.01932, over 7272.00 frames. ], tot_loss[loss=0.1262, simple_loss=0.2074, pruned_loss=0.02246, over 1439902.93 frames. ], batch size: 70, lr: 3.29e-03, grad_scale: 8.0 +2023-03-21 14:02:43,325 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0012W0478-21379-0_sp0.9 from training. Duration: 12.137875 +2023-03-21 14:02:53,596 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0_sp0.9 from training. Duration: 14.19225 +2023-03-21 14:02:54,217 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140103.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:02:57,797 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140110.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:03:06,002 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140125.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:03:06,943 INFO [train.py:901] (0/2) Epoch 50, batch 1750, loss[loss=0.1344, simple_loss=0.2155, pruned_loss=0.02669, over 7229.00 frames. ], tot_loss[loss=0.1264, simple_loss=0.2075, pruned_loss=0.02261, over 1437677.73 frames. ], batch size: 93, lr: 3.29e-03, grad_scale: 8.0 +2023-03-21 14:03:12,639 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9141, 4.0488, 3.8866, 3.9508, 3.7935, 3.5422, 4.0612, 4.1661], + device='cuda:0'), covar=tensor([0.0340, 0.0210, 0.0275, 0.0281, 0.0369, 0.0495, 0.0372, 0.0282], + device='cuda:0'), in_proj_covar=tensor([0.0126, 0.0128, 0.0123, 0.0126, 0.0115, 0.0103, 0.0100, 0.0103], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002, 0.0002], + device='cuda:0') +2023-03-21 14:03:16,126 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140145.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:03:18,017 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0025W0444-36635-0_sp0.9 from training. Duration: 13.15675 +2023-03-21 14:03:19,070 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0063W0488-49084-0_sp0.9 from training. Duration: 12.4045 +2023-03-21 14:03:22,027 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.55 vs. limit=2.0 +2023-03-21 14:03:23,717 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140160.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:03:24,737 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140162.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:03:25,797 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140164.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:03:28,147 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.418e+02 1.791e+02 2.014e+02 2.461e+02 5.476e+02, threshold=4.028e+02, percent-clipped=1.0 +2023-03-21 14:03:29,338 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140171.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:03:32,264 INFO [train.py:901] (0/2) Epoch 50, batch 1800, loss[loss=0.1342, simple_loss=0.2175, pruned_loss=0.02544, over 7340.00 frames. ], tot_loss[loss=0.1268, simple_loss=0.2082, pruned_loss=0.02267, over 1439565.77 frames. ], batch size: 54, lr: 3.29e-03, grad_scale: 8.0 +2023-03-21 14:03:41,518 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0242W0130-46331-0_sp0.9 from training. Duration: 13.184375 +2023-03-21 14:03:42,134 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2824, 1.9967, 2.2341, 3.3994, 1.9456, 3.2509, 1.3560, 3.1153], + device='cuda:0'), covar=tensor([0.0258, 0.2097, 0.2185, 0.0328, 0.4433, 0.0360, 0.1514, 0.0531], + device='cuda:0'), in_proj_covar=tensor([0.0211, 0.0243, 0.0255, 0.0214, 0.0248, 0.0220, 0.0216, 0.0229], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 14:03:48,149 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140206.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:03:55,919 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0170W0469-46827-0_sp0.9 from training. Duration: 12.9201875 +2023-03-21 14:03:56,538 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4180, 2.5728, 2.3109, 2.4770, 2.6062, 2.3199, 2.5465, 2.5001], + device='cuda:0'), covar=tensor([0.0993, 0.0656, 0.1061, 0.0765, 0.0394, 0.0663, 0.0811, 0.0769], + device='cuda:0'), in_proj_covar=tensor([0.0062, 0.0064, 0.0071, 0.0063, 0.0059, 0.0068, 0.0060, 0.0057], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 14:03:59,433 INFO [train.py:901] (0/2) Epoch 50, batch 1850, loss[loss=0.1202, simple_loss=0.203, pruned_loss=0.01871, over 7276.00 frames. ], tot_loss[loss=0.1268, simple_loss=0.2083, pruned_loss=0.02262, over 1440426.12 frames. ], batch size: 66, lr: 3.29e-03, grad_scale: 8.0 +2023-03-21 14:03:59,547 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140227.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:04:05,190 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140238.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:04:06,152 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0469-52699-0_sp0.9 from training. Duration: 12.1343125 +2023-03-21 14:04:09,906 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.16 vs. limit=2.0 +2023-03-21 14:04:15,845 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140259.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:04:21,423 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.209e+02 1.696e+02 1.990e+02 2.449e+02 3.903e+02, threshold=3.979e+02, percent-clipped=0.0 +2023-03-21 14:04:22,936 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0208-31683-0_sp0.9 from training. Duration: 12.273375 +2023-03-21 14:04:24,536 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=140275.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:04:25,549 INFO [train.py:901] (0/2) Epoch 50, batch 1900, loss[loss=0.1255, simple_loss=0.1996, pruned_loss=0.02571, over 7302.00 frames. ], tot_loss[loss=0.1261, simple_loss=0.2074, pruned_loss=0.02239, over 1438739.39 frames. ], batch size: 83, lr: 3.29e-03, grad_scale: 8.0 +2023-03-21 14:04:26,676 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.9893, 3.3115, 2.8863, 3.1616, 3.2468, 2.9582, 3.0711, 3.1084], + device='cuda:0'), covar=tensor([0.0642, 0.0549, 0.0707, 0.1076, 0.0555, 0.0588, 0.1299, 0.0657], + device='cuda:0'), in_proj_covar=tensor([0.0062, 0.0064, 0.0072, 0.0064, 0.0060, 0.0068, 0.0060, 0.0057], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 14:04:27,940 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.82 vs. limit=5.0 +2023-03-21 14:04:30,127 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=140286.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:04:48,135 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140320.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:04:48,516 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0702W0289-75727-0_sp0.9 from training. Duration: 12.8045 +2023-03-21 14:04:51,573 INFO [train.py:901] (0/2) Epoch 50, batch 1950, loss[loss=0.1411, simple_loss=0.2205, pruned_loss=0.03083, over 7281.00 frames. ], tot_loss[loss=0.1257, simple_loss=0.2071, pruned_loss=0.02218, over 1439481.38 frames. ], batch size: 47, lr: 3.29e-03, grad_scale: 8.0 +2023-03-21 14:04:55,290 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2579, 3.6805, 2.4309, 3.8929, 3.1799, 3.5559, 1.7637, 2.6341], + device='cuda:0'), covar=tensor([0.0493, 0.0776, 0.3240, 0.0620, 0.0523, 0.0618, 0.4523, 0.2117], + device='cuda:0'), in_proj_covar=tensor([0.0272, 0.0253, 0.0270, 0.0262, 0.0262, 0.0259, 0.0222, 0.0249], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 14:05:00,129 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0480-35915-0_sp0.9 from training. Duration: 12.51025 +2023-03-21 14:05:04,666 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0233W0325-47406-0_sp0.9 from training. Duration: 13.155375 +2023-03-21 14:05:05,242 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0708W0216-113208-0_sp1.1 from training. Duration: 12.1554375 +2023-03-21 14:05:13,387 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.395e+02 1.726e+02 1.971e+02 2.392e+02 3.404e+02, threshold=3.943e+02, percent-clipped=0.0 +2023-03-21 14:05:17,439 INFO [train.py:901] (0/2) Epoch 50, batch 2000, loss[loss=0.1181, simple_loss=0.2008, pruned_loss=0.01768, over 7304.00 frames. ], tot_loss[loss=0.126, simple_loss=0.2073, pruned_loss=0.02239, over 1439441.85 frames. ], batch size: 49, lr: 3.29e-03, grad_scale: 8.0 +2023-03-21 14:05:22,439 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0212W0468-40835-0 from training. Duration: 13.9758125 +2023-03-21 14:05:22,548 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140387.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:05:32,523 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([4.6871, 4.2070, 4.0372, 4.6429, 4.4741, 4.4852, 4.0893, 4.1252], + device='cuda:0'), covar=tensor([0.0882, 0.2719, 0.2540, 0.1045, 0.1040, 0.1457, 0.0802, 0.1359], + device='cuda:0'), in_proj_covar=tensor([0.0272, 0.0404, 0.0304, 0.0314, 0.0243, 0.0382, 0.0239, 0.0288], + device='cuda:0'), out_proj_covar=tensor([0.0002, 0.0004, 0.0003, 0.0003, 0.0002, 0.0003, 0.0002, 0.0003], + device='cuda:0') +2023-03-21 14:05:33,987 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0472-41898-0_sp0.9 from training. Duration: 16.1458125 +2023-03-21 14:05:42,002 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0348W0263-105296-0_sp0.9 from training. Duration: 14.5266875 +2023-03-21 14:05:42,576 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140425.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:05:43,545 INFO [train.py:901] (0/2) Epoch 50, batch 2050, loss[loss=0.1275, simple_loss=0.2016, pruned_loss=0.02669, over 7313.00 frames. ], tot_loss[loss=0.1265, simple_loss=0.2078, pruned_loss=0.02255, over 1441697.37 frames. ], batch size: 80, lr: 3.29e-03, grad_scale: 8.0 +2023-03-21 14:05:52,764 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140445.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:05:54,800 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140448.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:06:00,279 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140459.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:06:00,860 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140460.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:06:01,868 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140462.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:06:03,822 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140466.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:06:05,186 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.175e+02 1.778e+02 2.115e+02 2.474e+02 4.579e+02, threshold=4.229e+02, percent-clipped=2.0 +2023-03-21 14:06:07,204 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=140473.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:06:09,171 INFO [train.py:901] (0/2) Epoch 50, batch 2100, loss[loss=0.1276, simple_loss=0.2121, pruned_loss=0.02152, over 7253.00 frames. ], tot_loss[loss=0.1261, simple_loss=0.2074, pruned_loss=0.02239, over 1441311.04 frames. ], batch size: 64, lr: 3.29e-03, grad_scale: 8.0 +2023-03-21 14:06:13,400 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.6189, 1.7991, 1.6274, 1.6528, 1.6741, 1.6761, 1.7173, 1.5929], + device='cuda:0'), covar=tensor([0.0223, 0.0183, 0.0188, 0.0221, 0.0153, 0.0205, 0.0151, 0.0171], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0043, 0.0040, 0.0042, 0.0039, 0.0040, 0.0041, 0.0051], + device='cuda:0'), out_proj_covar=tensor([4.8712e-05, 4.7111e-05, 4.5161e-05, 4.6408e-05, 4.3267e-05, 4.4075e-05, + 4.6232e-05, 5.6077e-05], device='cuda:0') +2023-03-21 14:06:15,912 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140489.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:06:16,811 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0190W0196-79669-0_sp0.9 from training. Duration: 12.3110625 +2023-03-21 14:06:17,862 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=140493.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:06:19,371 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0364-52727-0 from training. Duration: 12.537125 +2023-03-21 14:06:24,464 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140506.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:06:25,391 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=140508.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:06:26,032 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140509.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:06:26,215 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=192, metric=1.88 vs. limit=2.0 +2023-03-21 14:06:26,428 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=140510.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:06:34,797 INFO [train.py:901] (0/2) Epoch 50, batch 2150, loss[loss=0.1438, simple_loss=0.2255, pruned_loss=0.03103, over 7278.00 frames. ], tot_loss[loss=0.1263, simple_loss=0.2075, pruned_loss=0.02254, over 1443063.11 frames. ], batch size: 57, lr: 3.29e-03, grad_scale: 8.0 +2023-03-21 14:06:47,334 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140550.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:06:49,234 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=140554.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:06:50,925 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=3.17 vs. limit=5.0 +2023-03-21 14:06:57,451 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.260e+02 1.798e+02 2.077e+02 2.497e+02 4.046e+02, threshold=4.154e+02, percent-clipped=0.0 +2023-03-21 14:06:58,139 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140570.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:07:01,577 INFO [train.py:901] (0/2) Epoch 50, batch 2200, loss[loss=0.1266, simple_loss=0.2072, pruned_loss=0.02297, over 7260.00 frames. ], tot_loss[loss=0.1264, simple_loss=0.2074, pruned_loss=0.02266, over 1439714.08 frames. ], batch size: 47, lr: 3.29e-03, grad_scale: 8.0 +2023-03-21 14:07:05,163 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0286-35767-0_sp0.9 from training. Duration: 13.03575 +2023-03-21 14:07:21,246 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140615.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:07:27,884 INFO [train.py:901] (0/2) Epoch 50, batch 2250, loss[loss=0.1346, simple_loss=0.2208, pruned_loss=0.02423, over 7357.00 frames. ], tot_loss[loss=0.1263, simple_loss=0.2075, pruned_loss=0.02256, over 1438340.85 frames. ], batch size: 51, lr: 3.29e-03, grad_scale: 4.0 +2023-03-21 14:07:35,040 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.9061, 3.1692, 3.9230, 3.7682, 4.0624, 3.8831, 3.9799, 3.7758], + device='cuda:0'), covar=tensor([0.0034, 0.0149, 0.0036, 0.0045, 0.0029, 0.0039, 0.0048, 0.0064], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0076, 0.0064, 0.0062, 0.0059, 0.0065, 0.0051, 0.0085], + device='cuda:0'), out_proj_covar=tensor([8.8336e-05, 1.4992e-04, 1.1098e-04, 1.0173e-04, 9.6333e-05, 1.0864e-04, + 9.3313e-05, 1.5362e-04], device='cuda:0') +2023-03-21 14:07:39,390 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0169W0400-1289-0_sp0.9 from training. Duration: 13.5110625 +2023-03-21 14:07:40,553 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0101W0466-14509-0_sp0.9 from training. Duration: 12.485625 +2023-03-21 14:07:50,017 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.217e+02 1.844e+02 2.128e+02 2.525e+02 5.922e+02, threshold=4.256e+02, percent-clipped=1.0 +2023-03-21 14:07:52,634 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0711W0333-38122-0_sp0.9 from training. Duration: 13.312125 +2023-03-21 14:07:53,650 INFO [train.py:901] (0/2) Epoch 50, batch 2300, loss[loss=0.1329, simple_loss=0.2151, pruned_loss=0.02529, over 7364.00 frames. ], tot_loss[loss=0.1264, simple_loss=0.2076, pruned_loss=0.02259, over 1441513.58 frames. ], batch size: 51, lr: 3.29e-03, grad_scale: 4.0 +2023-03-21 14:08:17,497 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140723.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:08:19,369 INFO [train.py:901] (0/2) Epoch 50, batch 2350, loss[loss=0.1001, simple_loss=0.1768, pruned_loss=0.01174, over 7161.00 frames. ], tot_loss[loss=0.1255, simple_loss=0.2068, pruned_loss=0.02212, over 1440846.60 frames. ], batch size: 39, lr: 3.28e-03, grad_scale: 4.0 +2023-03-21 14:08:28,117 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140743.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:08:29,372 INFO [scaling.py:679] (0/2) Whitening: num_groups=1, num_channels=384, metric=4.98 vs. limit=5.0 +2023-03-21 14:08:36,215 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140759.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:08:39,763 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140766.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:08:40,178 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0014W0402-35914-0_sp0.9 from training. Duration: 13.08 +2023-03-21 14:08:41,686 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.344e+02 1.750e+02 2.041e+02 2.523e+02 5.512e+02, threshold=4.082e+02, percent-clipped=1.0 +2023-03-21 14:08:45,237 INFO [train.py:901] (0/2) Epoch 50, batch 2400, loss[loss=0.1369, simple_loss=0.2236, pruned_loss=0.0251, over 7313.00 frames. ], tot_loss[loss=0.1261, simple_loss=0.2075, pruned_loss=0.02234, over 1440940.86 frames. ], batch size: 59, lr: 3.28e-03, grad_scale: 8.0 +2023-03-21 14:08:46,285 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0337W0459-31834-0_sp0.9 from training. Duration: 13.051 +2023-03-21 14:08:48,963 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140784.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:08:58,362 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0215W0439-96772-0_sp0.9 from training. Duration: 12.086625 +2023-03-21 14:09:01,367 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0016W0371-52743-0_sp0.9 from training. Duration: 12.3335 +2023-03-21 14:09:01,403 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=140807.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:09:02,501 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140809.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:09:05,005 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=140814.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:09:12,236 INFO [train.py:901] (0/2) Epoch 50, batch 2450, loss[loss=0.1286, simple_loss=0.209, pruned_loss=0.02408, over 7260.00 frames. ], tot_loss[loss=0.1255, simple_loss=0.207, pruned_loss=0.02201, over 1441177.67 frames. ], batch size: 89, lr: 3.28e-03, grad_scale: 8.0 +2023-03-21 14:09:21,497 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140845.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:09:21,567 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.3474, 2.5434, 2.6157, 2.2605, 2.7604, 2.5332, 2.3297, 1.9757], + device='cuda:0'), covar=tensor([0.0534, 0.0501, 0.0449, 0.0373, 0.0426, 0.0476, 0.0433, 0.0388], + device='cuda:0'), in_proj_covar=tensor([0.0044, 0.0043, 0.0043, 0.0043, 0.0040, 0.0040, 0.0047, 0.0047], + device='cuda:0'), out_proj_covar=tensor([0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001, 0.0001], + device='cuda:0') +2023-03-21 14:09:28,620 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0521W0124-45606-0 from training. Duration: 12.773 +2023-03-21 14:09:31,781 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=140865.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:09:34,208 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.235e+02 1.733e+02 1.983e+02 2.294e+02 8.300e+02, threshold=3.966e+02, percent-clipped=1.0 +2023-03-21 14:09:34,362 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140870.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:09:37,787 INFO [train.py:901] (0/2) Epoch 50, batch 2500, loss[loss=0.1239, simple_loss=0.2069, pruned_loss=0.02043, over 7307.00 frames. ], tot_loss[loss=0.126, simple_loss=0.207, pruned_loss=0.02248, over 1440286.42 frames. ], batch size: 80, lr: 3.28e-03, grad_scale: 8.0 +2023-03-21 14:09:54,472 WARNING [train.py:1061] (0/2) Exclude cut with ID BAC009S0658W0494-41924-0_sp0.9 from training. Duration: 12.070375 +2023-03-21 14:09:57,622 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=140915.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:09:58,654 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140917.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:10:00,403 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2921, 3.7016, 2.9507, 3.8740, 3.0453, 3.5166, 1.9394, 3.0115], + device='cuda:0'), covar=tensor([0.0405, 0.0741, 0.2027, 0.0505, 0.0426, 0.0494, 0.4206, 0.1903], + device='cuda:0'), in_proj_covar=tensor([0.0274, 0.0256, 0.0273, 0.0265, 0.0264, 0.0261, 0.0225, 0.0251], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0003, 0.0004, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 14:10:04,311 INFO [train.py:901] (0/2) Epoch 50, batch 2550, loss[loss=0.1444, simple_loss=0.2259, pruned_loss=0.03139, over 7323.00 frames. ], tot_loss[loss=0.1253, simple_loss=0.2061, pruned_loss=0.02222, over 1439036.33 frames. ], batch size: 59, lr: 3.28e-03, grad_scale: 8.0 +2023-03-21 14:10:14,653 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.4717, 2.6993, 2.4210, 2.6081, 2.6214, 2.2976, 2.4568, 2.5277], + device='cuda:0'), covar=tensor([0.0723, 0.0478, 0.0825, 0.0857, 0.0792, 0.0642, 0.0668, 0.0717], + device='cuda:0'), in_proj_covar=tensor([0.0063, 0.0064, 0.0072, 0.0064, 0.0060, 0.0069, 0.0061, 0.0057], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 14:10:14,657 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=140947.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:10:22,685 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=140963.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:10:26,237 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.104e+02 1.777e+02 2.057e+02 2.354e+02 4.545e+02, threshold=4.114e+02, percent-clipped=1.0 +2023-03-21 14:10:30,297 INFO [train.py:901] (0/2) Epoch 50, batch 2600, loss[loss=0.1217, simple_loss=0.1995, pruned_loss=0.02197, over 7156.00 frames. ], tot_loss[loss=0.1255, simple_loss=0.2062, pruned_loss=0.02239, over 1439561.98 frames. ], batch size: 39, lr: 3.28e-03, grad_scale: 8.0 +2023-03-21 14:10:30,984 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=140978.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:10:46,228 INFO [zipformer.py:625] (0/2) warmup_begin=3333.3, warmup_end=4000.0, batch_count=141008.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:10:55,563 INFO [train.py:901] (0/2) Epoch 50, batch 2650, loss[loss=0.126, simple_loss=0.2063, pruned_loss=0.02292, over 7289.00 frames. ], tot_loss[loss=0.1257, simple_loss=0.2061, pruned_loss=0.02263, over 1438334.81 frames. ], batch size: 77, lr: 3.28e-03, grad_scale: 8.0 +2023-03-21 14:11:03,567 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=141043.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:11:11,172 INFO [scaling.py:679] (0/2) Whitening: num_groups=8, num_channels=96, metric=1.44 vs. limit=2.0 +2023-03-21 14:11:12,446 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.7786, 1.8361, 2.2849, 2.2592, 2.0871, 2.2262, 2.0086, 2.3612], + device='cuda:0'), covar=tensor([0.2134, 0.2357, 0.1925, 0.1770, 0.2696, 0.1978, 0.3038, 0.2217], + device='cuda:0'), in_proj_covar=tensor([0.0089, 0.0087, 0.0082, 0.0072, 0.0072, 0.0072, 0.0113, 0.0072], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0002, 0.0003, 0.0002], + device='cuda:0') +2023-03-21 14:11:16,710 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.285e+02 1.781e+02 2.065e+02 2.401e+02 3.842e+02, threshold=4.131e+02, percent-clipped=0.0 +2023-03-21 14:11:17,882 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([1.8992, 2.0848, 1.7250, 2.1528, 2.1348, 1.8022, 1.7830, 1.4112], + device='cuda:0'), covar=tensor([0.0178, 0.0226, 0.0393, 0.0174, 0.0144, 0.0202, 0.0224, 0.0295], + device='cuda:0'), in_proj_covar=tensor([0.0043, 0.0042, 0.0040, 0.0041, 0.0039, 0.0039, 0.0041, 0.0050], + device='cuda:0'), out_proj_covar=tensor([4.7879e-05, 4.6709e-05, 4.4870e-05, 4.5573e-05, 4.2709e-05, 4.3206e-05, + 4.6008e-05, 5.5179e-05], device='cuda:0') +2023-03-21 14:11:19,339 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([3.2562, 2.9342, 3.3998, 3.1956, 3.4628, 3.0795, 2.8173, 3.1778], + device='cuda:0'), covar=tensor([0.1135, 0.0699, 0.1212, 0.1140, 0.0624, 0.0930, 0.1816, 0.1664], + device='cuda:0'), in_proj_covar=tensor([0.0054, 0.0073, 0.0055, 0.0053, 0.0054, 0.0053, 0.0071, 0.0054], + device='cuda:0'), out_proj_covar=tensor([0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004, 0.0004], + device='cuda:0') +2023-03-21 14:11:20,201 INFO [train.py:901] (0/2) Epoch 50, batch 2700, loss[loss=0.1244, simple_loss=0.2048, pruned_loss=0.02203, over 7311.00 frames. ], tot_loss[loss=0.1261, simple_loss=0.2067, pruned_loss=0.02276, over 1440578.42 frames. ], batch size: 49, lr: 3.28e-03, grad_scale: 8.0 +2023-03-21 14:11:21,243 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=141079.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:11:27,667 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=141091.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:11:45,203 INFO [train.py:901] (0/2) Epoch 50, batch 2750, loss[loss=0.1295, simple_loss=0.21, pruned_loss=0.0245, over 7343.00 frames. ], tot_loss[loss=0.1257, simple_loss=0.2066, pruned_loss=0.02237, over 1441451.82 frames. ], batch size: 63, lr: 3.28e-03, grad_scale: 8.0 +2023-03-21 14:11:53,266 INFO [zipformer.py:625] (0/2) warmup_begin=2666.7, warmup_end=3333.3, batch_count=141143.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:11:54,265 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=141145.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:12:03,916 INFO [zipformer.py:625] (0/2) warmup_begin=1333.3, warmup_end=2000.0, batch_count=141165.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:12:03,962 INFO [zipformer.py:625] (0/2) warmup_begin=2000.0, warmup_end=2666.7, batch_count=141165.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:12:06,326 INFO [optim.py:369] (0/2) Clipping_scale=2.0, grad-norm quartiles 1.171e+02 1.653e+02 1.885e+02 2.359e+02 4.895e+02, threshold=3.769e+02, percent-clipped=3.0 +2023-03-21 14:12:09,771 INFO [train.py:901] (0/2) Epoch 50, batch 2800, loss[loss=0.1323, simple_loss=0.2154, pruned_loss=0.02455, over 7230.00 frames. ], tot_loss[loss=0.1258, simple_loss=0.207, pruned_loss=0.02234, over 1441573.01 frames. ], batch size: 47, lr: 3.28e-03, grad_scale: 8.0 +2023-03-21 14:12:16,704 INFO [zipformer.py:1455] (0/2) attn_weights_entropy = tensor([2.1266, 2.5163, 1.9678, 2.7086, 2.8550, 2.8854, 2.7312, 2.4797], + device='cuda:0'), covar=tensor([0.2488, 0.1341, 0.4288, 0.0794, 0.0438, 0.0352, 0.0509, 0.0580], + device='cuda:0'), in_proj_covar=tensor([0.0244, 0.0230, 0.0239, 0.0250, 0.0204, 0.0204, 0.0219, 0.0224], + device='cuda:0'), out_proj_covar=tensor([0.0003, 0.0003, 0.0003, 0.0003, 0.0002, 0.0002, 0.0003, 0.0003], + device='cuda:0') +2023-03-21 14:12:17,601 INFO [zipformer.py:625] (0/2) warmup_begin=666.7, warmup_end=1333.3, batch_count=141193.0, num_to_drop=0, layers_to_drop=set() +2023-03-21 14:12:21,867 INFO [checkpoint.py:75] (0/2) Saving checkpoint to pruned_transducer_stateless7/exp/epoch-50.pt +2023-03-21 14:12:25,359 INFO [train.py:1170] (0/2) Done!