Hugging Face's logo

SungBeom
/
stt_kr_conformer_ctc_medium

Automatic Speech Recognition
NeMo
Korean
conformer-ctc
Model card Files Community
SungBeom commited on
Commit
0e52cef
β€’
1 Parent(s): dd0371d

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +7 -7
README.md CHANGED
@@ -20,6 +20,12 @@ V100 GPUμ—μ„œλŠ” RTFκ°€ 0.05, CPU(7 cores)μ—μ„œλŠ” 0.35 정도 λ‚˜μ˜€λŠ” 것
20
  Streaming μ½”λ“œμ™€ Denoise model이 ν¬ν•¨λœ μ½”λ“œλŠ” μ•„λž˜ κΉƒν—™μ—μ„œ 확인할 수 μžˆμŠ΅λ‹ˆλ‹€.
21
  [https://github.com/SUNGBEOMCHOI/Korean-Streaming-ASR](https://github.com/SUNGBEOMCHOI/Korean-Streaming-ASR)
22
 
 
 
 
 
 
 
23
 
24
  ### dataset
25
 
@@ -45,10 +51,4 @@ The following hyperparameters were used during training:
45
  - eval_batch_size: 16
46
  - num_train_epoch: 1
47
  - sample_rate: 16000
48
- - max_duration: 20.0
49
-
50
- ### Training results
51
-
52
- | Training Loss | Epoch | Wer |
53
- |:-------------:|:-----:|:-------:|
54
- | 9.09 | 1.0 | 11.51 |
 
20
  Streaming μ½”λ“œμ™€ Denoise model이 ν¬ν•¨λœ μ½”λ“œλŠ” μ•„λž˜ κΉƒν—™μ—μ„œ 확인할 수 μžˆμŠ΅λ‹ˆλ‹€.
21
  [https://github.com/SUNGBEOMCHOI/Korean-Streaming-ASR](https://github.com/SUNGBEOMCHOI/Korean-Streaming-ASR)
22
 
23
+ ### Training results
24
+
25
+ | Training Loss | Epoch | Wer |
26
+ |:-------------:|:-----:|:-------:|
27
+ | 9.09 | 1.0 | 11.51 |
28
+
29
 
30
  ### dataset
31
 
 
51
  - eval_batch_size: 16
52
  - num_train_epoch: 1
53
  - sample_rate: 16000
54
+ - max_duration: 20.0