2022-04-29 14:31:08,183 INFO [decode.py:475] Decoding started
2022-04-29 14:31:08,183 INFO [decode.py:481] Device: cuda:0
2022-04-29 14:31:08,185 INFO [decode.py:491] {'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, 'encoder_dim': 512, 'nhead': 8, 'dim_feedforward': 2048, 'num_encoder_layers': 12, 'decoder_dim': 512, 'joiner_dim': 512, 'model_warm_step': 3000, 'env_info': {'k2-version': '1.15.1', 'k2-build-type': 'Release', 'k2-with-cuda': True, 'k2-git-sha1': 'f8d2dba06c000ffee36aab5b66f24e7c9809f116', 'k2-git-date': 'Thu Apr 21 12:20:34 2022', 'lhotse-version': '1.1.0.dev+missing.version.file', 'torch-version': '1.10.0+cu102', 'torch-cuda-available': True, 'torch-cuda-version': '10.2', 'python-version': '3.8', 'icefall-git-branch': 'modified-conformer-with-multi-datasets', 'icefall-git-sha1': 'fb61e31-dirty', 'icefall-git-date': 'Fri Apr 29 14:15:19 2022', 'icefall-path': '/ceph-fj/fangjun/open-source-2/icefall-multi-4', 'k2-path': '/ceph-fj/fangjun/open-source-2/k2-multi-22/k2/python/k2/__init__.py', 'lhotse-path': '/ceph-fj/fangjun/open-source-2/lhotse-multi-3/lhotse/__init__.py', 'hostname': 'de-74279-k2-train-3-0307202051-57dc848959-8tmmp', 'IP address': '10.177.24.138'}, 'epoch': 38, 'iter': 0, 'avg': 10, 'exp_dir': PosixPath('pruned_transducer_stateless2/exp'), 'bpe_model': 'data/lang_bpe_500/bpe.model', 'decoding_method': 'greedy_search', 'beam_size': 4, 'beam': 4, 'max_contexts': 4, 'max_states': 8, 'context_size': 2, 'max_sym_per_frame': 1, 'full_libri': True, 'manifest_dir': PosixPath('data/fbank'), 'max_duration': 600, 'bucketing_sampler': True, 'num_buckets': 30, 'concatenate_cuts': False, 'duration_factor': 1.0, 'gap': 1.0, 'on_the_fly_feats': False, 'shuffle': True, 'return_cuts': True, 'num_workers': 2, 'enable_spec_aug': True, 'spec_aug_time_warp_factor': 80, 'enable_musan': True, 'res_dir': PosixPath('pruned_transducer_stateless2/exp/greedy_search'), 'suffix': 'epoch-38-avg-10-context-2-max-sym-per-frame-1', 'blank_id': 0, 'unk_id': 2, 'vocab_size': 500}
2022-04-29 14:31:08,185 INFO [decode.py:493] About to create model
2022-04-29 14:31:08,588 INFO [decode.py:521] averaging ['pruned_transducer_stateless2/exp/epoch-29.pt', 'pruned_transducer_stateless2/exp/epoch-30.pt', 'pruned_transducer_stateless2/exp/epoch-31.pt', 'pruned_transducer_stateless2/exp/epoch-32.pt', 'pruned_transducer_stateless2/exp/epoch-33.pt', 'pruned_transducer_stateless2/exp/epoch-34.pt', 'pruned_transducer_stateless2/exp/epoch-35.pt', 'pruned_transducer_stateless2/exp/epoch-36.pt', 'pruned_transducer_stateless2/exp/epoch-37.pt', 'pruned_transducer_stateless2/exp/epoch-38.pt']
2022-04-29 14:31:19,251 INFO [decode.py:535] Number of model parameters: 78648040
2022-04-29 14:31:19,251 INFO [asr_datamodule.py:422] About to get test-clean cuts
2022-04-29 14:31:19,353 INFO [asr_datamodule.py:427] About to get test-other cuts
2022-04-29 14:31:20,322 INFO [decode.py:389] batch 0/?, cuts processed until now is 123
2022-04-29 14:31:35,970 INFO [decode.py:406] The transcripts are stored in pruned_transducer_stateless2/exp/greedy_search/recogs-test-clean-greedy_search-epoch-38-avg-10-context-2-max-sym-per-frame-1.txt
2022-04-29 14:31:36,036 INFO [utils.py:405] [test-clean-greedy_search] %WER 2.62% [1379 / 52576, 141 ins, 116 del, 1122 sub ]
2022-04-29 14:31:36,203 INFO [decode.py:419] Wrote detailed error stats to pruned_transducer_stateless2/exp/greedy_search/errs-test-clean-greedy_search-epoch-38-avg-10-context-2-max-sym-per-frame-1.txt
2022-04-29 14:31:36,204 INFO [decode.py:436] 
For test-clean, WER of different settings are:
greedy_search	2.62	best for test-clean

2022-04-29 14:31:36,980 INFO [decode.py:389] batch 0/?, cuts processed until now is 138
2022-04-29 14:31:51,855 INFO [decode.py:406] The transcripts are stored in pruned_transducer_stateless2/exp/greedy_search/recogs-test-other-greedy_search-epoch-38-avg-10-context-2-max-sym-per-frame-1.txt
2022-04-29 14:31:51,923 INFO [utils.py:405] [test-other-greedy_search] %WER 6.02% [3153 / 52343, 309 ins, 320 del, 2524 sub ]
2022-04-29 14:31:52,091 INFO [decode.py:419] Wrote detailed error stats to pruned_transducer_stateless2/exp/greedy_search/errs-test-other-greedy_search-epoch-38-avg-10-context-2-max-sym-per-frame-1.txt
2022-04-29 14:31:52,092 INFO [decode.py:436] 
For test-other, WER of different settings are:
greedy_search	6.02	best for test-other

2022-04-29 14:31:52,092 INFO [decode.py:563] Done!