Add ctc prefix beam search results

decoding_results/ctc-prefix-beam-search/log-decode-epoch-50_avg-24_use-averaged-model-2024-09-27-10-50-32

@@ -0,0 +1,27 @@
+2024-09-27 10:50:32,892 INFO [ctc_decode.py:836] Decoding started
+2024-09-27 10:50:32,892 INFO [ctc_decode.py:844] Device: cuda:0
+2024-09-27 10:50:32,892 INFO [ctc_decode.py:845] {'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, 'ignore_id': -1, 'label_smoothing': 0.1, 'warm_step': 2000, 'env_info': {'k2-version': '1.24.4', 'k2-build-type': 'Release', 'k2-with-cuda': True, 'k2-git-sha1': 'ff1d435a8d3c4eaa15828a84a7240678a70539a7', 'k2-git-date': 'Fri Feb 23 01:48:38 2024', 'lhotse-version': '1.25.0.dev+git.da4d70d.clean', 'torch-version': '2.1.0+cu118', 'torch-cuda-available': True, 'torch-cuda-version': '11.8', 'python-version': '3.8', 'icefall-git-branch': 'cr-ctc', 'icefall-git-sha1': 'cef16574-dirty', 'icefall-git-date': 'Fri Sep 27 10:23:21 2024', 'icefall-path': '/star-kw/kangwei/code/icefall_cotrain', 'k2-path': '/star-kw/kangwei/envs/c5/lib/python3.8/site-packages/k2/__init__.py', 'lhotse-path': '/star-kw/kangwei/envs/c5/lib/python3.8/site-packages/lhotse/__init__.py', 'hostname': 'de-74279-k2-train-6-0905175136-74fb5b4b6f-p65lp', 'IP address': '10.30.18.160'}, 'frame_shift_ms': 10, 'search_beam': 20, 'output_beam': 8, 'min_active_states': 30, 'max_active_states': 10000, 'use_double_scores': True, 'epoch': 50, 'iter': 0, 'avg': 24, 'use_averaged_model': True, 'exp_dir': PosixPath('zipformer/exp_cr_ctc'), 'bpe_model': 'data/lang_bpe_500/bpe.model', 'lang_dir': PosixPath('data/lang_bpe_500'), 'context_size': 2, 'decoding_method': 'prefix-beam-search', 'num_paths': 100, 'nbest_scale': 1.0, 'lm_type': 'rnn', 'lm_scale': 0.3, 'hlg_scale': 0.6, 'lm_dir': PosixPath('data/lm'), 'skip_scoring': False, 'num_encoder_layers': '2,2,3,4,3,2', 'downsampling_factor': '1,2,4,8,4,2', 'feedforward_dim': '512,768,1024,1536,1024,768', 'num_heads': '4,4,4,8,4,4', 'encoder_dim': '192,256,384,512,384,256', 'query_head_dim': '32', 'value_head_dim': '12', 'pos_head_dim': '4', 'pos_dim': 48, 'encoder_unmasked_dim': '192,192,256,256,256,192', 'cnn_module_kernel': '31,31,15,15,15,31', 'decoder_dim': 512, 'joiner_dim': 512, 'attention_decoder_dim': 512, 'attention_decoder_num_layers': 6, 'attention_decoder_attention_dim': 512, 'attention_decoder_num_heads': 8, 'attention_decoder_feedforward_dim': 2048, 'causal': False, 'chunk_size': '16,32,64,-1', 'left_context_frames': '64,128,256,-1', 'use_transducer': False, 'use_ctc': True, 'use_attention_decoder': False, 'use_cr_ctc': False, 'full_libri': True, 'mini_libri': False, 'manifest_dir': PosixPath('data/fbank'), 'max_duration': 3000, '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', 'res_dir': PosixPath('zipformer/exp_cr_ctc/prefix-beam-search'), 'suffix': 'epoch-50_avg-24_use-averaged-model', 'device': device(type='cuda', index=0)}
+2024-09-27 10:50:33,225 INFO [lexicon.py:168] Loading pre-compiled data/lang_bpe_500/Linv.pt
+2024-09-27 10:50:33,279 INFO [ctc_decode.py:954] About to create model
+2024-09-27 10:50:33,995 INFO [ctc_decode.py:1021] Calculating the averaged model over epoch range from 26 (excluded) to 50
+2024-09-27 10:50:37,510 INFO [ctc_decode.py:1038] Number of model parameters: 64250603
+2024-09-27 10:50:37,510 INFO [asr_datamodule.py:467] About to get test-clean cuts
+2024-09-27 10:50:37,521 INFO [asr_datamodule.py:474] About to get test-other cuts
+2024-09-27 10:50:45,487 INFO [ctc_decode.py:718] batch 0/?, cuts processed until now is 132
+2024-09-27 10:51:48,613 INFO [ctc_decode.py:737] The transcripts are stored in zipformer/exp_cr_ctc/prefix-beam-search/recogs-test-clean-epoch-50_avg-24_use-averaged-model.txt
+2024-09-27 10:51:48,727 INFO [utils.py:657] [test-clean_prefix-beam-search] %WER 2.10% [1102 / 52576, 113 ins, 84 del, 905 sub ]
+2024-09-27 10:51:48,938 INFO [ctc_decode.py:765] Wrote detailed error stats to zipformer/exp_cr_ctc/prefix-beam-search/errs-test-clean-epoch-50_avg-24_use-averaged-model.txt
+2024-09-27 10:51:48,943 INFO [ctc_decode.py:781] 
+For test-clean, WER of different settings are:
+prefix-beam-search	2.1	best for test-clean
+
+2024-09-27 10:51:56,553 INFO [ctc_decode.py:718] batch 0/?, cuts processed until now is 86
+2024-09-27 10:52:56,687 INFO [zipformer.py:1858] name=None, attn_weights_entropy = tensor([5.1205, 3.8935, 4.8551, 4.6742], device='cuda:0')
+2024-09-27 10:53:01,966 INFO [ctc_decode.py:737] The transcripts are stored in zipformer/exp_cr_ctc/prefix-beam-search/recogs-test-other-epoch-50_avg-24_use-averaged-model.txt
+2024-09-27 10:53:02,064 INFO [utils.py:657] [test-other_prefix-beam-search] %WER 4.61% [2412 / 52343, 247 ins, 183 del, 1982 sub ]
+2024-09-27 10:53:02,280 INFO [ctc_decode.py:765] Wrote detailed error stats to zipformer/exp_cr_ctc/prefix-beam-search/errs-test-other-epoch-50_avg-24_use-averaged-model.txt
+2024-09-27 10:53:02,284 INFO [ctc_decode.py:781] 
+For test-other, WER of different settings are:
+prefix-beam-search	4.61	best for test-other
+
+2024-09-27 10:53:02,284 INFO [ctc_decode.py:1079] Done!

decoding_results/ctc-prefix-beam-search/wer-summary-test-clean-epoch-50_avg-24_use-averaged-model.txt

@@ -0,0 +1,2 @@
+settings	WER
+prefix-beam-search	2.1

decoding_results/ctc-prefix-beam-search/wer-summary-test-other-epoch-50_avg-24_use-averaged-model.txt

@@ -0,0 +1,2 @@
+settings	WER
+prefix-beam-search	4.61