decoding-results/fast_beam_search/log-decode-epoch-30-avg-9-beam-20.0-max-contexts-8-max-states-64-use-averaged-model-2023-10-17-18-59-01

2023-10-17 18:59:01,742 INFO [decode.py:847] Decoding started
2023-10-17 18:59:01,743 INFO [decode.py:853] Device: cuda:0
2023-10-17 18:59:01,747 INFO [decode.py:863] {'best_train_loss': inf, 'best_valid_loss': inf, 'best_train_epoch': -1, 'best_valid_epoch': -1, 'batch_idx_train': 0, 'log_interval': 500, 'reset_interval': 2000, 'valid_interval': 20000, 'feature_dim': 80, 'subsampling_factor': 4, 'warm_step': 2000, 'env_info': {'k2-version': '1.23.4', 'k2-build-type': 'Release', 'k2-with-cuda': True, 'k2-git-sha1': '0d7ef1a7867f70354ab5c59f2feb98c45558dcc7', 'k2-git-date': 'Sat Mar 18 12:59:04 2023', 'lhotse-version': '1.16.0', 'torch-version': '2.0.0', 'torch-cuda-available': True, 'torch-cuda-version': '11.8', 'python-version': '3.1', 'icefall-git-branch': 'master', 'icefall-git-sha1': '6f0f358-dirty', 'icefall-git-date': 'Tue Oct 17 17:58:28 2023', 'icefall-path': '/mnt/lustre/sjtu/home/yfy62/icefall-phone2', 'k2-path': '/home/yfy62/anaconda3/envs/icefall/lib/python3.10/site-packages/k2-1.23.4.dev20230319+cuda11.8.torch2.0.0-py3.10-linux-x86_64.egg/k2/__init__.py', 'lhotse-path': '/home/yfy62/anaconda3/envs/icefall/lib/python3.10/site-packages/lhotse/__init__.py', 'hostname': 'd3-hpc-sjtu-test-004', 'IP address': '10.11.11.11'}, 'epoch': 30, 'iter': 0, 'avg': 9, 'use_averaged_model': True, 'exp_dir': PosixPath('zipformer/exp'), 'bpe_model': 'data/lang_bpe_500/bpe.model', 'lang_dir': PosixPath('data/lang_bpe_500'), 'decoding_method': 'fast_beam_search', 'beam_size': 4, 'beam': 20.0, 'ngram_lm_scale': 0.01, 'max_contexts': 8, 'max_states': 64, 'context_size': 2, 'max_sym_per_frame': 1, 'num_paths': 200, 'nbest_scale': 0.5, 'use_shallow_fusion': False, 'lm_type': 'rnn', 'lm_scale': 0.3, 'tokens_ngram': 2, 'backoff_id': 500, 'context_score': 2, 'context_file': '', '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, 'causal': False, 'chunk_size': '16,32,64,-1', 'left_context_frames': '64,128,256,-1', 'use_transducer': True, 'use_ctc': False, 'manifest_dir': PosixPath('data/fbank'), 'max_duration': 1000, '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', 'subset': 'XL', 'small_dev': False, 'lm_vocab_size': 500, 'lm_epoch': 7, 'lm_avg': 1, 'lm_exp_dir': None, 'rnn_lm_embedding_dim': 2048, 'rnn_lm_hidden_dim': 2048, 'rnn_lm_num_layers': 3, 'rnn_lm_tie_weights': True, 'transformer_lm_exp_dir': None, 'transformer_lm_dim_feedforward': 2048, 'transformer_lm_encoder_dim': 768, 'transformer_lm_embedding_dim': 768, 'transformer_lm_nhead': 8, 'transformer_lm_num_layers': 16, 'transformer_lm_tie_weights': True, 'res_dir': PosixPath('zipformer/exp/fast_beam_search'), 'has_contexts': False, 'suffix': 'epoch-30-avg-9-beam-20.0-max-contexts-8-max-states-64-use-averaged-model', 'blank_id': 0, 'unk_id': 2, 'vocab_size': 500}
2023-10-17 18:59:01,747 INFO [decode.py:865] About to create model
2023-10-17 18:59:02,421 INFO [decode.py:932] Calculating the averaged model over epoch range from 21 (excluded) to 30
2023-10-17 18:59:07,070 INFO [decode.py:1026] Number of model parameters: 65549011
2023-10-17 18:59:07,070 INFO [asr_datamodule.py:420] About to get dev cuts
2023-10-17 18:59:07,072 INFO [asr_datamodule.py:431] About to get test cuts
2023-10-17 18:59:15,022 INFO [decode.py:719] batch 0/?, cuts processed until now is 90
2023-10-17 18:59:41,893 INFO [decode.py:719] batch 20/?, cuts processed until now is 2210
2023-10-17 19:00:07,626 INFO [decode.py:719] batch 40/?, cuts processed until now is 4827
2023-10-17 19:00:18,615 INFO [decode.py:736] The transcripts are stored in zipformer/exp/fast_beam_search/recogs-dev-beam_20.0_max_contexts_8_max_states_64-epoch-30-avg-9-beam-20.0-max-contexts-8-max-states-64-use-averaged-model.txt
2023-10-17 19:00:18,816 INFO [utils.py:565] [dev-beam_20.0_max_contexts_8_max_states_64] %WER 10.26% [13112 / 127790, 3135 ins, 3044 del, 6933 sub ]
2023-10-17 19:00:19,721 INFO [decode.py:749] Wrote detailed error stats to zipformer/exp/fast_beam_search/errs-dev-beam_20.0_max_contexts_8_max_states_64-epoch-30-avg-9-beam-20.0-max-contexts-8-max-states-64-use-averaged-model.txt
2023-10-17 19:00:19,723 INFO [decode.py:765] 
For dev, WER of different settings are:
beam_20.0_max_contexts_8_max_states_64	10.26	best for dev

2023-10-17 19:00:21,491 INFO [zipformer.py:1853] name=None, attn_weights_entropy = tensor([5.1283, 2.6171, 5.3495, 4.8337], device='cuda:0')
2023-10-17 19:00:22,696 INFO [decode.py:719] batch 0/?, cuts processed until now is 103
2023-10-17 19:00:48,854 INFO [decode.py:719] batch 20/?, cuts processed until now is 2381
2023-10-17 19:01:12,736 INFO [decode.py:719] batch 40/?, cuts processed until now is 4791
2023-10-17 19:01:34,848 INFO [decode.py:719] batch 60/?, cuts processed until now is 7544
2023-10-17 19:01:58,361 INFO [decode.py:719] batch 80/?, cuts processed until now is 9659
2023-10-17 19:02:12,045 INFO [zipformer.py:1853] name=None, attn_weights_entropy = tensor([3.0533, 2.9850, 3.9854, 4.3328], device='cuda:0')
2023-10-17 19:02:16,510 INFO [zipformer.py:1853] name=None, attn_weights_entropy = tensor([3.9255, 3.2101, 3.1162, 2.4563], device='cuda:0')
2023-10-17 19:02:21,320 INFO [decode.py:719] batch 100/?, cuts processed until now is 12300
2023-10-17 19:02:45,868 INFO [decode.py:719] batch 120/?, cuts processed until now is 14474
2023-10-17 19:02:47,533 INFO [zipformer.py:1853] name=None, attn_weights_entropy = tensor([2.2928, 2.7611, 3.0020, 3.4689, 3.6112, 3.4128, 3.7453, 2.2417],
       device='cuda:0')
2023-10-17 19:03:08,203 INFO [decode.py:719] batch 140/?, cuts processed until now is 18338
2023-10-17 19:03:11,724 INFO [zipformer.py:1853] name=None, attn_weights_entropy = tensor([2.4436, 3.6440, 3.7375, 2.7308, 3.3381, 3.6095, 2.9244, 3.1445],
       device='cuda:0')
2023-10-17 19:03:21,750 INFO [decode.py:736] The transcripts are stored in zipformer/exp/fast_beam_search/recogs-test-beam_20.0_max_contexts_8_max_states_64-epoch-30-avg-9-beam-20.0-max-contexts-8-max-states-64-use-averaged-model.txt
2023-10-17 19:03:22,733 INFO [utils.py:565] [test-beam_20.0_max_contexts_8_max_states_64] %WER 10.48% [40961 / 390744, 7114 ins, 10034 del, 23813 sub ]
2023-10-17 19:03:23,880 INFO [decode.py:749] Wrote detailed error stats to zipformer/exp/fast_beam_search/errs-test-beam_20.0_max_contexts_8_max_states_64-epoch-30-avg-9-beam-20.0-max-contexts-8-max-states-64-use-averaged-model.txt
2023-10-17 19:03:23,882 INFO [decode.py:765] 
For test, WER of different settings are:
beam_20.0_max_contexts_8_max_states_64	10.48	best for test

2023-10-17 19:03:23,917 INFO [decode.py:1061] Done!