Hugging Face's logo

geninhu
/
xls-asr-vi-40h-1B

Automatic Speech Recognition
Transformers
PyTorch
TensorBoard
Vietnamese
wav2vec2
common-voice
hf-asr-leaderboard
robust-speech-event
Eval Results
Inference Endpoints
Model card Files Metrics Community
1
geninhu commited on
Commit
60dafac
·
1 Parent(s): 45d54c6

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +7 -7
README.md CHANGED
@@ -18,12 +18,12 @@ model-index:
18
 
19
  args: vi
20
  metrics:
21
- - name: Test WER
22
  type: wer
23
- value: 34.210
24
- - name: Test CER
25
  type: cer
26
- value: 19.938
27
  ---
28
 
29
 
@@ -35,13 +35,13 @@ This model is a fine-tuned version of [facebook/wav2vec2-xls-r-1b](https://huggi
35
  | | [VIVOS](https://huggingface.co/datasets/vivos) | [COMMON VOICE 7.0](https://huggingface.co/datasets/mozilla-foundation/common_voice_7_0) |
36
  |---|---|---|
37
  |without LM| 25.93 | 34.21 |
38
- |with 4-grams LM| | |
39
 
40
  ### Benchmark CER result:
41
  | | [VIVOS](https://huggingface.co/datasets/vivos) | [COMMON VOICE 7.0](https://huggingface.co/datasets/mozilla-foundation/common_voice_7_0) |
42
  |---|---|---|
43
- |without LM| 9.243 | 19.938 |
44
- |with 4-grams LM| | |
45
 
46
  ## Evaluation
47
 
 
18
 
19
  args: vi
20
  metrics:
21
+ - name: Test WER (with LM)
22
  type: wer
23
+ value: 28.269
24
+ - name: Test CER (with LM)
25
  type: cer
26
+ value: 14.421
27
  ---
28
 
29
 
 
35
  | | [VIVOS](https://huggingface.co/datasets/vivos) | [COMMON VOICE 7.0](https://huggingface.co/datasets/mozilla-foundation/common_voice_7_0) |
36
  |---|---|---|
37
  |without LM| 25.93 | 34.21 |
38
+ |with 4-grams LM| 0.24 | 28.27 |
39
 
40
  ### Benchmark CER result:
41
  | | [VIVOS](https://huggingface.co/datasets/vivos) | [COMMON VOICE 7.0](https://huggingface.co/datasets/mozilla-foundation/common_voice_7_0) |
42
  |---|---|---|
43
+ |without LM| 9.24 | 19.94 |
44
+ |with 4-grams LM| 0.104 | 14.42 |
45
 
46
  ## Evaluation
47