Automatic Speech Recognition
Transformers
Safetensors
Japanese
whisper
audio
hf-asr-leaderboard
Eval Results
Inference Endpoints
asahi417 commited on
Commit
bc781c0
·
verified ·
1 Parent(s): 4d61b55

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +8 -8
README.md CHANGED
@@ -26,27 +26,27 @@ model-index:
26
  - task:
27
  type: automatic-speech-recognition
28
  dataset:
29
- name: CommonVoice_8.0 (Japanese)
30
  type: japanese-asr/ja_asr.common_voice_8_0
31
  metrics:
32
  - name: WER
33
  type: WER
34
- value: 59.27
35
  - name: CER
36
  type: CER
37
- value: 9.44
38
  - task:
39
  type: automatic-speech-recognition
40
  dataset:
41
- name: ReazonSpeech (Test)
42
  type: japanese-asr/ja_asr.reazonspeech_test
43
  metrics:
44
  - name: WER
45
  type: WER
46
- value: 56.62
47
  - name: CER
48
  type: CER
49
- value: 12.6
50
  - task:
51
  type: automatic-speech-recognition
52
  dataset:
@@ -55,10 +55,10 @@ model-index:
55
  metrics:
56
  - name: WER
57
  type: WER
58
- value: 64.36
59
  - name: CER
60
  type: CER
61
- value: 8.48
62
  datasets:
63
  - japanese-asr/whisper_transcriptions.reazonspeech.large
64
  - japanese-asr/whisper_transcriptions.reazonspeech.large.wer_10.0
 
26
  - task:
27
  type: automatic-speech-recognition
28
  dataset:
29
+ name: CommonVoice 8 (Japanese test set)
30
  type: japanese-asr/ja_asr.common_voice_8_0
31
  metrics:
32
  - name: WER
33
  type: WER
34
+ value: 59.2
35
  - name: CER
36
  type: CER
37
+ value: 9.4
38
  - task:
39
  type: automatic-speech-recognition
40
  dataset:
41
+ name: ReazonSpeech (held out test set)
42
  type: japanese-asr/ja_asr.reazonspeech_test
43
  metrics:
44
  - name: WER
45
  type: WER
46
+ value: 56.6
47
  - name: CER
48
  type: CER
49
+ value: 12.2
50
  - task:
51
  type: automatic-speech-recognition
52
  dataset:
 
55
  metrics:
56
  - name: WER
57
  type: WER
58
+ value: 64.3
59
  - name: CER
60
  type: CER
61
+ value: 8.5
62
  datasets:
63
  - japanese-asr/whisper_transcriptions.reazonspeech.large
64
  - japanese-asr/whisper_transcriptions.reazonspeech.large.wer_10.0