asahi417 commited on
Commit
f033ea6
1 Parent(s): a18dc33

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +13 -27
README.md CHANGED
@@ -8,31 +8,17 @@ pinned: false
8
  ---
9
  # Japanese ASR
10
 
11
- Hello! This repository contains Japanese ASR models and Japanese ASR dataset for model training and evaluation.
 
 
12
 
13
-
14
- | model | data | wer |
15
- |:--------------------------------------------------------------------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------------------|-------:|
16
- | [japanese-asr/distil-whisper-large-v3-ja-reazonspeech-large](https://huggingface.co/japanese-asr/distil-whisper-large-v3-ja-reazonspeech-large) | [japanese-asr/common_voice_8_0](https://huggingface.co/datasets/japanese-asr/common_voice_8_0) | 59.27 |
17
- | [japanese-asr/distil-whisper-large-v3-ja-reazonspeech-medium](https://huggingface.co/japanese-asr/distil-whisper-large-v3-ja-reazonspeech-medium) | [japanese-asr/common_voice_8_0](https://huggingface.co/datasets/japanese-asr/common_voice_8_0) | 64.38 |
18
- | [japanese-asr/distil-whisper-large-v3-ja-reazonspeech-small](https://huggingface.co/japanese-asr/distil-whisper-large-v3-ja-reazonspeech-small) | [japanese-asr/common_voice_8_0](https://huggingface.co/datasets/japanese-asr/common_voice_8_0) | 85.1 |
19
- | [japanese-asr/distil-whisper-large-v3-ja-reazonspeech-tiny](https://huggingface.co/japanese-asr/distil-whisper-large-v3-ja-reazonspeech-tiny) | [japanese-asr/common_voice_8_0](https://huggingface.co/datasets/japanese-asr/common_voice_8_0) | 99.96 |
20
- | [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) | [japanese-asr/common_voice_8_0](https://huggingface.co/datasets/japanese-asr/common_voice_8_0) | 55.41 |
21
- | [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) | [japanese-asr/common_voice_8_0](https://huggingface.co/datasets/japanese-asr/common_voice_8_0) | 63.64 |
22
- | [openai/whisper-small](https://huggingface.co/openai/whisper-small) | [japanese-asr/common_voice_8_0](https://huggingface.co/datasets/japanese-asr/common_voice_8_0) | 74.21 |
23
- | [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) | [japanese-asr/common_voice_8_0](https://huggingface.co/datasets/japanese-asr/common_voice_8_0) | 93.78 |
24
- | [japanese-asr/distil-whisper-large-v3-ja-reazonspeech-large](https://huggingface.co/japanese-asr/distil-whisper-large-v3-ja-reazonspeech-large) | [japanese-asr/jsut-basic5000](https://huggingface.co/datasets/japanese-asr/jsut-basic5000) | 64.36 |
25
- | [japanese-asr/distil-whisper-large-v3-ja-reazonspeech-medium](https://huggingface.co/japanese-asr/distil-whisper-large-v3-ja-reazonspeech-medium) | [japanese-asr/jsut-basic5000](https://huggingface.co/datasets/japanese-asr/jsut-basic5000) | 72.02 |
26
- | [japanese-asr/distil-whisper-large-v3-ja-reazonspeech-small](https://huggingface.co/japanese-asr/distil-whisper-large-v3-ja-reazonspeech-small) | [japanese-asr/jsut-basic5000](https://huggingface.co/datasets/japanese-asr/jsut-basic5000) | 94.18 |
27
- | [japanese-asr/distil-whisper-large-v3-ja-reazonspeech-tiny](https://huggingface.co/japanese-asr/distil-whisper-large-v3-ja-reazonspeech-tiny) | [japanese-asr/jsut-basic5000](https://huggingface.co/datasets/japanese-asr/jsut-basic5000) | 100 |
28
- | [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) | [japanese-asr/jsut-basic5000](https://huggingface.co/datasets/japanese-asr/jsut-basic5000) | 59.34 |
29
- | [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) | [japanese-asr/jsut-basic5000](https://huggingface.co/datasets/japanese-asr/jsut-basic5000) | 69.52 |
30
- | [openai/whisper-small](https://huggingface.co/openai/whisper-small) | [japanese-asr/jsut-basic5000](https://huggingface.co/datasets/japanese-asr/jsut-basic5000) | 82.02 |
31
- | [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) | [japanese-asr/jsut-basic5000](https://huggingface.co/datasets/japanese-asr/jsut-basic5000) | 97.72 |
32
- | [japanese-asr/distil-whisper-large-v3-ja-reazonspeech-large](https://huggingface.co/japanese-asr/distil-whisper-large-v3-ja-reazonspeech-large) | [japanese-asr/reazonspeech_test](https://huggingface.co/datasets/japanese-asr/reazonspeech_test) | 56.62 |
33
- | [japanese-asr/distil-whisper-large-v3-ja-reazonspeech-medium](https://huggingface.co/japanese-asr/distil-whisper-large-v3-ja-reazonspeech-medium) | [japanese-asr/reazonspeech_test](https://huggingface.co/datasets/japanese-asr/reazonspeech_test) | 62.99 |
34
- | [japanese-asr/distil-whisper-large-v3-ja-reazonspeech-small](https://huggingface.co/japanese-asr/distil-whisper-large-v3-ja-reazonspeech-small) | [japanese-asr/reazonspeech_test](https://huggingface.co/datasets/japanese-asr/reazonspeech_test) | 82.18 |
35
- | [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) | [japanese-asr/reazonspeech_test](https://huggingface.co/datasets/japanese-asr/reazonspeech_test) | 60.23 |
36
- | [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) | [japanese-asr/reazonspeech_test](https://huggingface.co/datasets/japanese-asr/reazonspeech_test) | 76.04 |
37
- | [openai/whisper-small](https://huggingface.co/openai/whisper-small) | [japanese-asr/reazonspeech_test](https://huggingface.co/datasets/japanese-asr/reazonspeech_test) | 82.99 |
38
- | [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) | [japanese-asr/reazonspeech_test](https://huggingface.co/datasets/japanese-asr/reazonspeech_test) | 94.85 |
 
8
  ---
9
  # Japanese ASR
10
 
11
+ This repository contains all the models and datasets for train/evaluate the Japanese ASR dataset generated through the process of achieving https://huggingface.co/kotoba-tech/kotoba-whisper-v1.0.
12
+ Following table shows CER comparison with different data size of ReazonSpeech used to distill the whisper-large-v3. The model names follows
13
+ `japanese-asr/distil-whisper-large-v3-ja-reazonspeech-{size of reazonspeech}`.
14
 
15
+ | model | common_voice_8_0 | jsut_basic5000 | reazonspeech_test |
16
+ |:--------------------------------------------------------------------------------------------------------------------------------------------------|-------------------:|-----------------:|--------------------:|
17
+ | [japanese-asr/distil-whisper-large-v3-ja-reazonspeech-large](https://huggingface.co/japanese-asr/distil-whisper-large-v3-ja-reazonspeech-large) | 9.44 | 8.48 | 12.6 |
18
+ | [japanese-asr/distil-whisper-large-v3-ja-reazonspeech-medium](https://huggingface.co/japanese-asr/distil-whisper-large-v3-ja-reazonspeech-medium) | 10.89 | 11.25 | 16.37 |
19
+ | [japanese-asr/distil-whisper-large-v3-ja-reazonspeech-small](https://huggingface.co/japanese-asr/distil-whisper-large-v3-ja-reazonspeech-small) | 30.48 | 38.96 | 42.29 |
20
+ | [japanese-asr/distil-whisper-large-v3-ja-reazonspeech-tiny](https://huggingface.co/japanese-asr/distil-whisper-large-v3-ja-reazonspeech-tiny) | 94.69 | 95.32 | 95.82 |
21
+ | [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) | 8.52 | 7.18 | 15.18 |
22
+ | [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) | 11.34 | 9.87 | 29.56 |
23
+ | [openai/whisper-small](https://huggingface.co/openai/whisper-small) | 15.26 | 14.22 | 34.29 |
24
+ | [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) | 46.86 | 35.69 | 96.69 |