2022-07-20 21:02:14,439 INFO [decode.py:653] Decoding started
2022-07-20 21:02:14,440 INFO [decode.py:654] {'subsampling_factor': 4, 'feature_dim': 80, 'nhead': 8, 'dim_feedforward': 2048, 'encoder_dim': 512, 'num_encoder_layers': 12, 'search_beam': 20, 'output_beam': 8, 'min_active_states': 30, 'max_active_states': 10000, 'use_double_scores': True, 'env_info': {'k2-version': '1.15.1', 'k2-build-type': 'Release', 'k2-with-cuda': True, 'k2-git-sha1': '350e449fe5d7b4231f77eb0add764782eed9f5d2', 'k2-git-date': 'Thu May 26 22:56:24 2022', 'lhotse-version': '1.3.0.dev+git.232f3eb.dirty', 'torch-version': '1.10.0+cu102', 'torch-cuda-available': True, 'torch-cuda-version': '10.2', 'python-version': '3.8', 'icefall-git-branch': 'conformer_ctc2', 'icefall-git-sha1': 'd812f82-dirty', 'icefall-git-date': 'Wed Jul 20 19:47:36 2022', 'icefall-path': '/home/storage23/wangquandong/experiment/open_source/icefall', 'k2-path': '/home/storage23/wangquandong/tools/k2/k2/python/k2/__init__.py', 'lhotse-path': '/home/storage23/wangquandong/tools/lhotse/lhotse/__init__.py', 'hostname': 'tj1-asr-train-v100-30.kscn', 'IP address': '10.38.22.183'}, 'epoch': 30, 'iter': 0, 'avg': 8, 'method': 'whole-lattice-rescoring', 'use_averaged_model': True, 'num_decoder_layers': 6, 'num_paths': 100, 'nbest_scale': 0.5, 'exp_dir': PosixPath('pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp'), 'lang_dir': PosixPath('data/lang_bpe_500'), 'lm_dir': PosixPath('data/lm'), 'full_libri': True, 'manifest_dir': PosixPath('data/fbank'), 'max_duration': 20, 'bucketing_sampler': True, 'num_buckets': 30, 'concatenate_cuts': False, 'duration_factor': 1.0, 'gap': 1.0, 'on_the_fly_feats': False, 'shuffle': True, 'drop_last': True, 'return_cuts': True, 'num_workers': 2, 'enable_spec_aug': True, 'spec_aug_time_warp_factor': 80, 'enable_musan': True, 'input_strategy': 'PrecomputedFeatures'}
2022-07-20 21:02:14,840 INFO [lexicon.py:176] Loading pre-compiled data/lang_bpe_500/Linv.pt
2022-07-20 21:02:14,895 INFO [decode.py:664] device: cuda:0
2022-07-20 21:02:21,292 INFO [decode.py:730] Loading pre-compiled G_4_gram.pt
2022-07-20 21:02:26,887 INFO [decode.py:821] Calculating the averaged model over epoch range from 22 (excluded) to 30
2022-07-20 21:02:30,017 INFO [decode.py:837] Number of model parameters: 103071035
2022-07-20 21:02:30,018 INFO [asr_datamodule.py:444] About to get test-clean cuts
2022-07-20 21:02:30,022 INFO [asr_datamodule.py:451] About to get test-other cuts
2022-07-20 21:02:31,504 INFO [decode.py:588] batch 0/?, cuts processed until now is 1
2022-07-20 21:03:17,856 INFO [decode.py:588] batch 100/?, cuts processed until now is 212
2022-07-20 21:03:58,002 INFO [decode.py:588] batch 200/?, cuts processed until now is 459
2022-07-20 21:04:39,673 INFO [decode.py:588] batch 300/?, cuts processed until now is 717
2022-07-20 21:05:25,026 INFO [decode.py:588] batch 400/?, cuts processed until now is 949
2022-07-20 21:06:10,220 INFO [decode.py:588] batch 500/?, cuts processed until now is 1163
2022-07-20 21:06:51,165 INFO [decode.py:588] batch 600/?, cuts processed until now is 1411
2022-07-20 21:07:30,465 INFO [decode.py:588] batch 700/?, cuts processed until now is 1652
2022-07-20 21:08:14,019 INFO [decode.py:588] batch 800/?, cuts processed until now is 1867
2022-07-20 21:08:52,156 INFO [decode.py:588] batch 900/?, cuts processed until now is 2122
2022-07-20 21:09:33,815 INFO [decode.py:588] batch 1000/?, cuts processed until now is 2321
2022-07-20 21:10:18,006 INFO [decode.py:588] batch 1100/?, cuts processed until now is 2468
2022-07-20 21:11:00,039 INFO [decode.py:588] batch 1200/?, cuts processed until now is 2601
2022-07-20 21:11:06,549 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_0.1.txt
2022-07-20 21:11:06,631 INFO [utils.py:416] [test-clean-lm_scale_0.1] %WER 2.82% [1482 / 52576, 214 ins, 100 del, 1168 sub ]
2022-07-20 21:11:06,864 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_0.1.txt
2022-07-20 21:11:06,893 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_0.2.txt
2022-07-20 21:11:06,973 INFO [utils.py:416] [test-clean-lm_scale_0.2] %WER 2.74% [1440 / 52576, 206 ins, 102 del, 1132 sub ]
2022-07-20 21:11:07,204 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_0.2.txt
2022-07-20 21:11:07,234 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_0.3.txt
2022-07-20 21:11:07,313 INFO [utils.py:416] [test-clean-lm_scale_0.3] %WER 2.70% [1421 / 52576, 200 ins, 105 del, 1116 sub ]
2022-07-20 21:11:07,543 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_0.3.txt
2022-07-20 21:11:07,572 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_0.4.txt
2022-07-20 21:11:07,649 INFO [utils.py:416] [test-clean-lm_scale_0.4] %WER 2.67% [1403 / 52576, 195 ins, 109 del, 1099 sub ]
2022-07-20 21:11:07,879 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_0.4.txt
2022-07-20 21:11:07,907 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_0.5.txt
2022-07-20 21:11:07,984 INFO [utils.py:416] [test-clean-lm_scale_0.5] %WER 2.66% [1400 / 52576, 188 ins, 112 del, 1100 sub ]
2022-07-20 21:11:08,213 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_0.5.txt
2022-07-20 21:11:08,241 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_0.6.txt
2022-07-20 21:11:08,318 INFO [utils.py:416] [test-clean-lm_scale_0.6] %WER 2.68% [1407 / 52576, 181 ins, 119 del, 1107 sub ]
2022-07-20 21:11:08,549 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_0.6.txt
2022-07-20 21:11:08,577 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_0.7.txt
2022-07-20 21:11:08,794 INFO [utils.py:416] [test-clean-lm_scale_0.7] %WER 2.68% [1409 / 52576, 173 ins, 130 del, 1106 sub ]
2022-07-20 21:11:09,022 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_0.7.txt
2022-07-20 21:11:09,050 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_0.8.txt
2022-07-20 21:11:09,127 INFO [utils.py:416] [test-clean-lm_scale_0.8] %WER 2.70% [1422 / 52576, 169 ins, 142 del, 1111 sub ]
2022-07-20 21:11:09,369 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_0.8.txt
2022-07-20 21:11:09,400 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_0.9.txt
2022-07-20 21:11:09,476 INFO [utils.py:416] [test-clean-lm_scale_0.9] %WER 2.74% [1443 / 52576, 156 ins, 167 del, 1120 sub ]
2022-07-20 21:11:09,705 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_0.9.txt
2022-07-20 21:11:09,735 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_1.0.txt
2022-07-20 21:11:09,811 INFO [utils.py:416] [test-clean-lm_scale_1.0] %WER 2.84% [1491 / 52576, 156 ins, 203 del, 1132 sub ]
2022-07-20 21:11:10,041 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_1.0.txt
2022-07-20 21:11:10,070 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_1.1.txt
2022-07-20 21:11:10,147 INFO [utils.py:416] [test-clean-lm_scale_1.1] %WER 2.94% [1545 / 52576, 151 ins, 246 del, 1148 sub ]
2022-07-20 21:11:10,377 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_1.1.txt
2022-07-20 21:11:10,407 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_1.2.txt
2022-07-20 21:11:10,485 INFO [utils.py:416] [test-clean-lm_scale_1.2] %WER 3.09% [1623 / 52576, 146 ins, 300 del, 1177 sub ]
2022-07-20 21:11:10,714 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_1.2.txt
2022-07-20 21:11:10,743 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_1.3.txt
2022-07-20 21:11:10,820 INFO [utils.py:416] [test-clean-lm_scale_1.3] %WER 3.30% [1734 / 52576, 149 ins, 376 del, 1209 sub ]
2022-07-20 21:11:11,203 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_1.3.txt
2022-07-20 21:11:11,232 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_1.4.txt
2022-07-20 21:11:11,310 INFO [utils.py:416] [test-clean-lm_scale_1.4] %WER 3.51% [1846 / 52576, 148 ins, 462 del, 1236 sub ]
2022-07-20 21:11:11,539 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_1.4.txt
2022-07-20 21:11:11,567 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_1.5.txt
2022-07-20 21:11:11,644 INFO [utils.py:416] [test-clean-lm_scale_1.5] %WER 3.79% [1992 / 52576, 144 ins, 567 del, 1281 sub ]
2022-07-20 21:11:11,873 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_1.5.txt
2022-07-20 21:11:11,901 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_1.6.txt
2022-07-20 21:11:11,978 INFO [utils.py:416] [test-clean-lm_scale_1.6] %WER 4.07% [2140 / 52576, 131 ins, 698 del, 1311 sub ]
2022-07-20 21:11:12,209 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_1.6.txt
2022-07-20 21:11:12,237 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_1.7.txt
2022-07-20 21:11:12,315 INFO [utils.py:416] [test-clean-lm_scale_1.7] %WER 4.38% [2303 / 52576, 131 ins, 826 del, 1346 sub ]
2022-07-20 21:11:12,544 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_1.7.txt
2022-07-20 21:11:12,572 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_1.8.txt
2022-07-20 21:11:12,649 INFO [utils.py:416] [test-clean-lm_scale_1.8] %WER 4.69% [2465 / 52576, 131 ins, 955 del, 1379 sub ]
2022-07-20 21:11:12,881 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_1.8.txt
2022-07-20 21:11:12,910 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_1.9.txt
2022-07-20 21:11:12,989 INFO [utils.py:416] [test-clean-lm_scale_1.9] %WER 4.99% [2625 / 52576, 133 ins, 1068 del, 1424 sub ]
2022-07-20 21:11:13,366 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_1.9.txt
2022-07-20 21:11:13,405 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-clean-lm_scale_2.0.txt
2022-07-20 21:11:13,494 INFO [utils.py:416] [test-clean-lm_scale_2.0] %WER 5.31% [2794 / 52576, 130 ins, 1197 del, 1467 sub ]
2022-07-20 21:11:13,727 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-clean-lm_scale_2.0.txt
2022-07-20 21:11:13,728 INFO [decode.py:637] 
For test-clean, WER of different settings are:
lm_scale_0.5	2.66	best for test-clean
lm_scale_0.4	2.67
lm_scale_0.6	2.68
lm_scale_0.7	2.68
lm_scale_0.3	2.7
lm_scale_0.8	2.7
lm_scale_0.2	2.74
lm_scale_0.9	2.74
lm_scale_0.1	2.82
lm_scale_1.0	2.84
lm_scale_1.1	2.94
lm_scale_1.2	3.09
lm_scale_1.3	3.3
lm_scale_1.4	3.51
lm_scale_1.5	3.79
lm_scale_1.6	4.07
lm_scale_1.7	4.38
lm_scale_1.8	4.69
lm_scale_1.9	4.99
lm_scale_2.0	5.31

2022-07-20 21:11:14,658 INFO [decode.py:588] batch 0/?, cuts processed until now is 2
2022-07-20 21:11:54,814 INFO [decode.py:588] batch 100/?, cuts processed until now is 249
2022-07-20 21:12:32,635 INFO [decode.py:588] batch 200/?, cuts processed until now is 540
2022-07-20 21:13:09,226 INFO [decode.py:588] batch 300/?, cuts processed until now is 834
2022-07-20 21:13:44,899 INFO [decode.py:588] batch 400/?, cuts processed until now is 1104
2022-07-20 21:14:26,260 INFO [decode.py:588] batch 500/?, cuts processed until now is 1350
2022-07-20 21:15:05,158 INFO [decode.py:588] batch 600/?, cuts processed until now is 1640
2022-07-20 21:15:40,242 INFO [decode.py:588] batch 700/?, cuts processed until now is 1933
2022-07-20 21:16:18,132 INFO [decode.py:588] batch 800/?, cuts processed until now is 2175
2022-07-20 21:16:56,866 INFO [decode.py:588] batch 900/?, cuts processed until now is 2440
2022-07-20 21:17:04,052 INFO [decode.py:783] Caught exception:
CUDA out of memory. Tried to allocate 740.00 MiB (GPU 0; 31.75 GiB total capacity; 28.28 GiB already allocated; 445.50 MiB free; 29.93 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation.  See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF

2022-07-20 21:17:04,052 INFO [decode.py:789] num_arcs before pruning: 349313
2022-07-20 21:17:04,052 INFO [decode.py:792] This OOM is not an error. You can ignore it. If your model does not converge well, or --max-duration is too large, or the input sound file is difficult to decode, you will meet this exception.
2022-07-20 21:17:04,072 INFO [decode.py:803] num_arcs after pruning: 8796
2022-07-20 21:17:41,206 INFO [decode.py:588] batch 1000/?, cuts processed until now is 2656
2022-07-20 21:18:21,669 INFO [decode.py:588] batch 1100/?, cuts processed until now is 2839
2022-07-20 21:18:59,437 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_0.1.txt
2022-07-20 21:18:59,529 INFO [utils.py:416] [test-other-lm_scale_0.1] %WER 6.24% [3266 / 52343, 440 ins, 240 del, 2586 sub ]
2022-07-20 21:18:59,774 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_0.1.txt
2022-07-20 21:18:59,807 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_0.2.txt
2022-07-20 21:18:59,890 INFO [utils.py:416] [test-other-lm_scale_0.2] %WER 6.04% [3161 / 52343, 414 ins, 245 del, 2502 sub ]
2022-07-20 21:19:00,128 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_0.2.txt
2022-07-20 21:19:00,161 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_0.3.txt
2022-07-20 21:19:00,243 INFO [utils.py:416] [test-other-lm_scale_0.3] %WER 5.94% [3107 / 52343, 396 ins, 249 del, 2462 sub ]
2022-07-20 21:19:00,480 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_0.3.txt
2022-07-20 21:19:00,512 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_0.4.txt
2022-07-20 21:19:00,594 INFO [utils.py:416] [test-other-lm_scale_0.4] %WER 5.85% [3060 / 52343, 379 ins, 268 del, 2413 sub ]
2022-07-20 21:19:01,006 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_0.4.txt
2022-07-20 21:19:01,035 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_0.5.txt
2022-07-20 21:19:01,118 INFO [utils.py:416] [test-other-lm_scale_0.5] %WER 5.77% [3022 / 52343, 351 ins, 286 del, 2385 sub ]
2022-07-20 21:19:01,356 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_0.5.txt
2022-07-20 21:19:01,385 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_0.6.txt
2022-07-20 21:19:01,467 INFO [utils.py:416] [test-other-lm_scale_0.6] %WER 5.76% [3017 / 52343, 332 ins, 320 del, 2365 sub ]
2022-07-20 21:19:01,708 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_0.6.txt
2022-07-20 21:19:01,736 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_0.7.txt
2022-07-20 21:19:01,816 INFO [utils.py:416] [test-other-lm_scale_0.7] %WER 5.77% [3020 / 52343, 318 ins, 348 del, 2354 sub ]
2022-07-20 21:19:02,052 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_0.7.txt
2022-07-20 21:19:02,081 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_0.8.txt
2022-07-20 21:19:02,162 INFO [utils.py:416] [test-other-lm_scale_0.8] %WER 5.78% [3028 / 52343, 294 ins, 398 del, 2336 sub ]
2022-07-20 21:19:02,398 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_0.8.txt
2022-07-20 21:19:02,427 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_0.9.txt
2022-07-20 21:19:02,507 INFO [utils.py:416] [test-other-lm_scale_0.9] %WER 5.86% [3066 / 52343, 278 ins, 464 del, 2324 sub ]
2022-07-20 21:19:02,743 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_0.9.txt
2022-07-20 21:19:02,771 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_1.0.txt
2022-07-20 21:19:02,852 INFO [utils.py:416] [test-other-lm_scale_1.0] %WER 6.02% [3149 / 52343, 268 ins, 540 del, 2341 sub ]
2022-07-20 21:19:03,087 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_1.0.txt
2022-07-20 21:19:03,115 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_1.1.txt
2022-07-20 21:19:03,196 INFO [utils.py:416] [test-other-lm_scale_1.1] %WER 6.20% [3247 / 52343, 258 ins, 636 del, 2353 sub ]
2022-07-20 21:19:03,431 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_1.1.txt
2022-07-20 21:19:03,460 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_1.2.txt
2022-07-20 21:19:03,686 INFO [utils.py:416] [test-other-lm_scale_1.2] %WER 6.40% [3351 / 52343, 248 ins, 749 del, 2354 sub ]
2022-07-20 21:19:03,923 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_1.2.txt
2022-07-20 21:19:03,951 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_1.3.txt
2022-07-20 21:19:04,036 INFO [utils.py:416] [test-other-lm_scale_1.3] %WER 6.67% [3489 / 52343, 241 ins, 891 del, 2357 sub ]
2022-07-20 21:19:04,280 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_1.3.txt
2022-07-20 21:19:04,309 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_1.4.txt
2022-07-20 21:19:04,392 INFO [utils.py:416] [test-other-lm_scale_1.4] %WER 7.08% [3705 / 52343, 234 ins, 1071 del, 2400 sub ]
2022-07-20 21:19:04,628 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_1.4.txt
2022-07-20 21:19:04,656 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_1.5.txt
2022-07-20 21:19:04,737 INFO [utils.py:416] [test-other-lm_scale_1.5] %WER 7.42% [3886 / 52343, 224 ins, 1253 del, 2409 sub ]
2022-07-20 21:19:04,973 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_1.5.txt
2022-07-20 21:19:05,001 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_1.6.txt
2022-07-20 21:19:05,081 INFO [utils.py:416] [test-other-lm_scale_1.6] %WER 7.84% [4102 / 52343, 214 ins, 1437 del, 2451 sub ]
2022-07-20 21:19:05,317 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_1.6.txt
2022-07-20 21:19:05,346 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_1.7.txt
2022-07-20 21:19:05,426 INFO [utils.py:416] [test-other-lm_scale_1.7] %WER 8.29% [4338 / 52343, 213 ins, 1641 del, 2484 sub ]
2022-07-20 21:19:05,667 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_1.7.txt
2022-07-20 21:19:05,696 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_1.8.txt
2022-07-20 21:19:05,777 INFO [utils.py:416] [test-other-lm_scale_1.8] %WER 8.78% [4596 / 52343, 209 ins, 1859 del, 2528 sub ]
2022-07-20 21:19:06,031 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_1.8.txt
2022-07-20 21:19:06,082 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_1.9.txt
2022-07-20 21:19:06,310 INFO [utils.py:416] [test-other-lm_scale_1.9] %WER 9.27% [4854 / 52343, 205 ins, 2085 del, 2564 sub ]
2022-07-20 21:19:06,549 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_1.9.txt
2022-07-20 21:19:06,578 INFO [decode.py:609] The transcripts are stored in pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/recogs-test-other-lm_scale_2.0.txt
2022-07-20 21:19:06,659 INFO [utils.py:416] [test-other-lm_scale_2.0] %WER 9.72% [5090 / 52343, 204 ins, 2272 del, 2614 sub ]
2022-07-20 21:19:06,896 INFO [decode.py:621] Wrote detailed error stats to pruned_transducer_stateless_smooth/quandongtest_8gpu_ctc_attnetion2-exp/errs-test-other-lm_scale_2.0.txt
2022-07-20 21:19:06,897 INFO [decode.py:637] 
For test-other, WER of different settings are:
lm_scale_0.6	5.76	best for test-other
lm_scale_0.5	5.77
lm_scale_0.7	5.77
lm_scale_0.8	5.78
lm_scale_0.4	5.85
lm_scale_0.9	5.86
lm_scale_0.3	5.94
lm_scale_1.0	6.02
lm_scale_0.2	6.04
lm_scale_1.1	6.2
lm_scale_0.1	6.24
lm_scale_1.2	6.4
lm_scale_1.3	6.67
lm_scale_1.4	7.08
lm_scale_1.5	7.42
lm_scale_1.6	7.84
lm_scale_1.7	8.29
lm_scale_1.8	8.78
lm_scale_1.9	9.27
lm_scale_2.0	9.72

2022-07-20 21:19:06,897 INFO [decode.py:868] Done!