decoding_results/ctc-prefix-beam-search/log-decode-epoch-50_avg-26_beam-4_use-averaged-model-2024-10-01-17-27-53

2024-10-01 17:27:53,949 INFO [ctc_decode.py:844] Decoding started
2024-10-01 17:27:53,949 INFO [ctc_decode.py:852] Device: cuda:0
2024-10-01 17:27:53,949 INFO [ctc_decode.py:853] {
  "attention_decoder_attention_dim": 512,
  "attention_decoder_dim": 512,
  "attention_decoder_feedforward_dim": 2048,
  "attention_decoder_num_heads": 8,
  "attention_decoder_num_layers": 6,
  "avg": 26,
  "batch_idx_train": 0,
  "beam": 4,
  "best_train_epoch": -1,
  "best_train_loss": Infinity,
  "best_valid_epoch": -1,
  "best_valid_loss": Infinity,
  "bpe_model": "data/lang_bpe_500/bpe.model",
  "bucketing_sampler": true,
  "causal": false,
  "chunk_size": "16,32,64,-1",
  "cnn_module_kernel": "31,31,15,15,15,31",
  "concatenate_cuts": false,
  "context_size": 2,
  "decoder_dim": 512,
  "decoding_method": "prefix-beam-search",
  "device": "cuda:0",
  "downsampling_factor": "1,2,4,8,4,2",
  "drop_last": true,
  "duration_factor": 1.0,
  "enable_musan": true,
  "enable_spec_aug": true,
  "encoder_dim": "192,256,512,768,512,256",
  "encoder_unmasked_dim": "192,192,256,320,256,192",
  "env_info": {
    "IP address": "10.30.18.160",
    "hostname": "de-74279-k2-train-6-0905175136-74fb5b4b6f-p65lp",
    "icefall-git-branch": "cr-ctc",
    "icefall-git-date": "Sun Sep 29 12:00:45 2024",
    "icefall-git-sha1": "33fa9e8b-dirty",
    "icefall-path": "/star-kw/kangwei/code/icefall_cotrain",
    "k2-build-type": "Release",
    "k2-git-date": "Fri Feb 23 01:48:38 2024",
    "k2-git-sha1": "ff1d435a8d3c4eaa15828a84a7240678a70539a7",
    "k2-path": "/star-kw/kangwei/envs/c5/lib/python3.8/site-packages/k2/__init__.py",
    "k2-version": "1.24.4",
    "k2-with-cuda": true,
    "lhotse-path": "/star-kw/kangwei/envs/c5/lib/python3.8/site-packages/lhotse/__init__.py",
    "lhotse-version": "1.25.0.dev+git.da4d70d.clean",
    "python-version": "3.8",
    "torch-cuda-available": true,
    "torch-cuda-version": "11.8",
    "torch-version": "2.1.0+cu118"
  },
  "epoch": 50,
  "exp_dir": "zipformer/exp_cr_large",
  "feature_dim": 80,
  "feedforward_dim": "512,768,1536,2048,1536,768",
  "frame_shift_ms": 10,
  "full_libri": true,
  "gap": 1.0,
  "hlg_scale": 0.6,
  "ignore_id": -1,
  "input_strategy": "PrecomputedFeatures",
  "iter": 0,
  "joiner_dim": 512,
  "label_smoothing": 0.1,
  "lang_dir": "data/lang_bpe_500",
  "left_context_frames": "64,128,256,-1",
  "lm_avg": 1,
  "lm_dir": "data/lm",
  "lm_epoch": 7,
  "lm_exp_dir": null,
  "lm_scale": 0.3,
  "lm_type": "rnn",
  "lm_vocab_size": 500,
  "log_interval": 50,
  "manifest_dir": "data/fbank",
  "max_active_states": 10000,
  "max_duration": 2000,
  "min_active_states": 30,
  "mini_libri": false,
  "nbest_scale": 1.0,
  "num_buckets": 30,
  "num_encoder_layers": "2,2,4,5,4,2",
  "num_heads": "4,4,4,8,4,4",
  "num_paths": 100,
  "num_workers": 2,
  "on_the_fly_feats": false,
  "output_beam": 8,
  "pos_dim": 48,
  "pos_head_dim": "4",
  "query_head_dim": "32",
  "res_dir": "zipformer/exp_cr_large/prefix-beam-search",
  "reset_interval": 200,
  "return_cuts": true,
  "rnn_lm_embedding_dim": 2048,
  "rnn_lm_hidden_dim": 2048,
  "rnn_lm_num_layers": 3,
  "rnn_lm_tie_weights": true,
  "search_beam": 20,
  "shuffle": true,
  "skip_scoring": false,
  "spec_aug_time_warp_factor": 80,
  "subsampling_factor": 4,
  "suffix": "epoch-50_avg-26_beam-4_use-averaged-model",
  "transformer_lm_dim_feedforward": 2048,
  "transformer_lm_embedding_dim": 768,
  "transformer_lm_encoder_dim": 768,
  "transformer_lm_exp_dir": null,
  "transformer_lm_nhead": 8,
  "transformer_lm_num_layers": 16,
  "transformer_lm_tie_weights": true,
  "use_attention_decoder": false,
  "use_averaged_model": true,
  "use_cr_ctc": false,
  "use_ctc": true,
  "use_double_scores": true,
  "use_transducer": false,
  "valid_interval": 3000,
  "value_head_dim": "12",
  "warm_step": 2000
}
2024-10-01 17:27:54,226 INFO [lexicon.py:168] Loading pre-compiled data/lang_bpe_500/Linv.pt
2024-10-01 17:27:54,279 INFO [ctc_decode.py:962] About to create model
2024-10-01 17:27:55,310 INFO [ctc_decode.py:1029] Calculating the averaged model over epoch range from 24 (excluded) to 50
2024-10-01 17:28:04,596 INFO [ctc_decode.py:1046] Number of model parameters: 147010094
2024-10-01 17:28:04,596 INFO [asr_datamodule.py:467] About to get test-clean cuts
2024-10-01 17:28:04,610 INFO [asr_datamodule.py:474] About to get test-other cuts
2024-10-01 17:28:11,489 INFO [ctc_decode.py:720] batch 0/?, cuts processed until now is 123
2024-10-01 17:29:46,243 INFO [ctc_decode.py:739] The transcripts are stored in zipformer/exp_cr_large/prefix-beam-search/recogs-test-clean-epoch-50_avg-26_beam-4_use-averaged-model.txt
2024-10-01 17:29:46,335 INFO [utils.py:668] [test-clean_prefix-beam-search] %WER 2.02% [1060 / 52576, 114 ins, 82 del, 864 sub ]
2024-10-01 17:29:46,542 INFO [ctc_decode.py:767] Wrote detailed error stats to zipformer/exp_cr_large/prefix-beam-search/errs-test-clean-epoch-50_avg-26_beam-4_use-averaged-model.txt
2024-10-01 17:29:46,548 INFO [ctc_decode.py:783] 
For test-clean, WER of different settings are:
prefix-beam-search	2.02	best for test-clean

2024-10-01 17:29:52,962 INFO [ctc_decode.py:720] batch 0/?, cuts processed until now is 142
2024-10-01 17:31:35,541 INFO [ctc_decode.py:739] The transcripts are stored in zipformer/exp_cr_large/prefix-beam-search/recogs-test-other-epoch-50_avg-26_beam-4_use-averaged-model.txt
2024-10-01 17:31:35,635 INFO [utils.py:668] [test-other_prefix-beam-search] %WER 4.35% [2276 / 52343, 224 ins, 180 del, 1872 sub ]
2024-10-01 17:31:35,842 INFO [ctc_decode.py:767] Wrote detailed error stats to zipformer/exp_cr_large/prefix-beam-search/errs-test-other-epoch-50_avg-26_beam-4_use-averaged-model.txt
2024-10-01 17:31:35,848 INFO [ctc_decode.py:783] 
For test-other, WER of different settings are:
prefix-beam-search	4.35	best for test-other

2024-10-01 17:31:35,848 INFO [ctc_decode.py:1087] Done!