pfluo commited on
Commit
706c9b7
1 Parent(s): e5d4fa6

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +7 -5
README.md CHANGED
@@ -3,13 +3,15 @@ license: apache-2.0
3
  ---
4
  ## Chinese-English ASR model using k2-zipformer-streaming
5
  ### AIShell-1 and Wenetspeech testset results with modified-beam-search streaming decode using epoch-14.pt
6
- | AIShell-1 | TEST_NET | TEST_MEETING |
7
- |-----------|----------|--------------|
8
- | 3.19 | 9.58 | 9.51 ||
 
9
 
10
- ### Training commond
11
  ```
12
  nohup ./pruned_transducer_stateless7_streaming/train.py --world-size 8 --num-epochs 30 --start-epoch 1 --feedforward-dims "1024,1024,1536,1536,1024" --exp-dir pruned_transducer_stateless7_streaming/exp --max-duration 360 > pruned_transducer_stateless7_streaming/exp/nohup.zipformer &
 
13
  ```
14
 
15
  ### Model unit is char+bpe as `data/lang_char_bpe/tokens.txt`
@@ -31,4 +33,4 @@ dims': '192,192,192,192,192', 'encoder_unmasked_dims': '256,256,256,256,256', 'z
31
  _dim': 512, 'joiner_dim': 512, 'short_chunk_size': 50, 'num_left_chunks': 4, 'decode_chunk_len': 32, 'manifest_dir': PosixPath('data/fbank'), 'max_duration': 360, 'bucketing
32
  _sampler': True, 'num_buckets': 300, 'concatenate_cuts': False, 'duration_factor': 1.0, 'gap': 1.0, 'on_the_fly_feats': False, 'shuffle': True, 'return_cuts': True, 'num_wor
33
  kers': 8, 'enable_spec_aug': True, 'spec_aug_time_warp_factor': 80, 'enable_musan': True, 'training_subset': '12k_hour', 'blank_id': 0, 'vocab_size': 6254}
34
- ```
 
3
  ---
4
  ## Chinese-English ASR model using k2-zipformer-streaming
5
  ### AIShell-1 and Wenetspeech testset results with modified-beam-search streaming decode using epoch-14.pt
6
+ | decode_chunk_len | AIShell-1 | TEST_NET | TEST_MEETING |
7
+ |------------------|-----------|----------|--------------|
8
+ | 32 | 3.19 | 9.58 | 9.51 ||
9
+ | 64 | 3.04 | 8.97 | 8.83 ||
10
 
11
+ ### Training and decoding commands
12
  ```
13
  nohup ./pruned_transducer_stateless7_streaming/train.py --world-size 8 --num-epochs 30 --start-epoch 1 --feedforward-dims "1024,1024,1536,1536,1024" --exp-dir pruned_transducer_stateless7_streaming/exp --max-duration 360 > pruned_transducer_stateless7_streaming/exp/nohup.zipformer &
14
+ nohup ./pruned_transducer_stateless7_streaming/decode.py --epoch 6 --avg 1 --exp-dir ./pruned_transducer_stateless7_streaming/exp --max-duration 600 --decode-chunk-len 32 --decoding-method modified_beam_search --beam-size 4 > nohup.zipformer.deocode &
15
  ```
16
 
17
  ### Model unit is char+bpe as `data/lang_char_bpe/tokens.txt`
 
33
  _dim': 512, 'joiner_dim': 512, 'short_chunk_size': 50, 'num_left_chunks': 4, 'decode_chunk_len': 32, 'manifest_dir': PosixPath('data/fbank'), 'max_duration': 360, 'bucketing
34
  _sampler': True, 'num_buckets': 300, 'concatenate_cuts': False, 'duration_factor': 1.0, 'gap': 1.0, 'on_the_fly_feats': False, 'shuffle': True, 'return_cuts': True, 'num_wor
35
  kers': 8, 'enable_spec_aug': True, 'spec_aug_time_warp_factor': 80, 'enable_musan': True, 'training_subset': '12k_hour', 'blank_id': 0, 'vocab_size': 6254}
36
+ ```