Yoshiki
commited on
Commit
•
1f58bde
1
Parent(s):
6a61ad5
Update model
Browse files- README.md +71 -0
- exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/RESULTS.md +44 -0
- exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/config.yaml +0 -0
- exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/acc.png +0 -0
- exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/backward_time.png +0 -0
- exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/cer.png +0 -0
- exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/clip.png +0 -0
- exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/forward_time.png +0 -0
- exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/gpu_max_cached_mem_GB.png +0 -0
- exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/grad_norm.png +0 -0
- exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/iter_time.png +0 -0
- exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/loss.png +0 -0
- exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/loss_att.png +0 -0
- exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/loss_scale.png +0 -0
- exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/optim0_lr0.png +0 -0
- exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/optim_step_time.png +0 -0
- exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/train_time.png +0 -0
- exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/wer.png +0 -0
- exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/valid.acc.ave_3best.pth +3 -0
- meta.yaml +8 -0
README.md
CHANGED
@@ -1,3 +1,74 @@
|
|
1 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2 |
license: cc-by-4.0
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
+
tags:
|
3 |
+
- espnet
|
4 |
+
- audio
|
5 |
+
- automatic-speech-recognition
|
6 |
+
language: en
|
7 |
+
datasets:
|
8 |
+
- wsj
|
9 |
license: cc-by-4.0
|
10 |
---
|
11 |
+
|
12 |
+
## ESPnet2 ASR model
|
13 |
+
|
14 |
+
### `espnet/yoshiki_wsj_whisper_medium_finetuning`
|
15 |
+
|
16 |
+
This model was trained by Yoshiki using wsj recipe in [espnet](https://github.com/espnet/espnet/).
|
17 |
+
|
18 |
+
### Demo: How to use in ESPnet2
|
19 |
+
|
20 |
+
Follow the [ESPnet installation instructions](https://espnet.github.io/espnet/installation.html)
|
21 |
+
if you haven't done that already.
|
22 |
+
|
23 |
+
```bash
|
24 |
+
cd espnet
|
25 |
+
git checkout fe00740b80cd26fad7c550cd9e975609deb664db
|
26 |
+
pip install -e .
|
27 |
+
cd egs2/wsj/asr1
|
28 |
+
./run.sh --skip_data_prep false --skip_train true --download_model espnet/yoshiki_wsj_whisper_medium_finetuning
|
29 |
+
```
|
30 |
+
|
31 |
+
<!-- Generated by scripts/utils/show_asr_result.sh -->
|
32 |
+
# RESULTS
|
33 |
+
## Environments
|
34 |
+
- date: `Fri Jul 21 12:47:17 JST 2023`
|
35 |
+
- python version: `3.10.10 (main, Mar 21 2023, 18:45:11) [GCC 11.2.0]`
|
36 |
+
- espnet version: `espnet 202304`
|
37 |
+
- pytorch version: `pytorch 1.13.1`
|
38 |
+
- Git hash: `d7172fcb7181ffdcca9c0061400254b63e37bf21`
|
39 |
+
- Commit date: `Sat Jul 15 15:01:30 2023 +0900`
|
40 |
+
|
41 |
+
## /scratch/espnet-hackathon/egs2/wsj/asr1/exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual
|
42 |
+
### WER
|
43 |
+
|
44 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
45 |
+
|---|---|---|---|---|---|---|---|---|
|
46 |
+
|decode_asr_whisper_noctc_greedy_asr_model_valid.acc.ave/test_eval92|333|5564|98.5|1.4|0.1|0.2|1.7|19.5|
|
47 |
+
|
48 |
+
### CER
|
49 |
+
|
50 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
51 |
+
|---|---|---|---|---|---|---|---|---|
|
52 |
+
|decode_asr_whisper_noctc_greedy_asr_model_valid.acc.ave/test_eval92|333|32096|99.6|0.2|0.2|0.2|0.6|19.5|
|
53 |
+
|
54 |
+
### TER
|
55 |
+
|
56 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
57 |
+
|---|---|---|---|---|---|---|---|---|
|
58 |
+
## /scratch/espnet-hackathon/egs2/wsj/asr1/exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/decode_asr_whisper_noctc_greedy_asr_model_valid.acc.ave
|
59 |
+
### WER
|
60 |
+
|
61 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
62 |
+
|---|---|---|---|---|---|---|---|---|
|
63 |
+
|org/test_dev93|503|8132|97.7|2.0|0.4|0.3|2.6|24.9|
|
64 |
+
|
65 |
+
### CER
|
66 |
+
|
67 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
68 |
+
|---|---|---|---|---|---|---|---|---|
|
69 |
+
|org/test_dev93|503|46683|99.4|0.3|0.3|0.2|0.8|24.9|
|
70 |
+
|
71 |
+
### TER
|
72 |
+
|
73 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
74 |
+
|---|---|---|---|---|---|---|---|---|
|
exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/RESULTS.md
ADDED
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
<!-- Generated by scripts/utils/show_asr_result.sh -->
|
2 |
+
# RESULTS
|
3 |
+
## Environments
|
4 |
+
- date: `Fri Jul 21 12:47:17 JST 2023`
|
5 |
+
- python version: `3.10.10 (main, Mar 21 2023, 18:45:11) [GCC 11.2.0]`
|
6 |
+
- espnet version: `espnet 202304`
|
7 |
+
- pytorch version: `pytorch 1.13.1`
|
8 |
+
- Git hash: `d7172fcb7181ffdcca9c0061400254b63e37bf21`
|
9 |
+
- Commit date: `Sat Jul 15 15:01:30 2023 +0900`
|
10 |
+
|
11 |
+
## /scratch/espnet-hackathon/egs2/wsj/asr1/exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual
|
12 |
+
### WER
|
13 |
+
|
14 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
15 |
+
|---|---|---|---|---|---|---|---|---|
|
16 |
+
|decode_asr_whisper_noctc_greedy_asr_model_valid.acc.ave/test_eval92|333|5564|98.5|1.4|0.1|0.2|1.7|19.5|
|
17 |
+
|
18 |
+
### CER
|
19 |
+
|
20 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
21 |
+
|---|---|---|---|---|---|---|---|---|
|
22 |
+
|decode_asr_whisper_noctc_greedy_asr_model_valid.acc.ave/test_eval92|333|32096|99.6|0.2|0.2|0.2|0.6|19.5|
|
23 |
+
|
24 |
+
### TER
|
25 |
+
|
26 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
27 |
+
|---|---|---|---|---|---|---|---|---|
|
28 |
+
## /scratch/espnet-hackathon/egs2/wsj/asr1/exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/decode_asr_whisper_noctc_greedy_asr_model_valid.acc.ave
|
29 |
+
### WER
|
30 |
+
|
31 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
32 |
+
|---|---|---|---|---|---|---|---|---|
|
33 |
+
|org/test_dev93|503|8132|97.7|2.0|0.4|0.3|2.6|24.9|
|
34 |
+
|
35 |
+
### CER
|
36 |
+
|
37 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
38 |
+
|---|---|---|---|---|---|---|---|---|
|
39 |
+
|org/test_dev93|503|46683|99.4|0.3|0.3|0.2|0.8|24.9|
|
40 |
+
|
41 |
+
### TER
|
42 |
+
|
43 |
+
|dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
|
44 |
+
|---|---|---|---|---|---|---|---|---|
|
exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/config.yaml
ADDED
The diff for this file is too large to render.
See raw diff
|
|
exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/acc.png
ADDED
exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/backward_time.png
ADDED
exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/cer.png
ADDED
exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/clip.png
ADDED
exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/forward_time.png
ADDED
exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/gpu_max_cached_mem_GB.png
ADDED
exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/grad_norm.png
ADDED
exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/iter_time.png
ADDED
exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/loss.png
ADDED
exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/loss_att.png
ADDED
exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/loss_scale.png
ADDED
exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/optim0_lr0.png
ADDED
exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/optim_step_time.png
ADDED
exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/train_time.png
ADDED
exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/images/wer.png
ADDED
exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/valid.acc.ave_3best.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8124a87993b59790a0cadb4c106d4d05f9b98f575f9a0c2c6885d30ee284cec7
|
3 |
+
size 3055787223
|
meta.yaml
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
espnet: '202304'
|
2 |
+
files:
|
3 |
+
asr_model_file: exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/valid.acc.ave_3best.pth
|
4 |
+
python: 3.10.10 (main, Mar 21 2023, 18:45:11) [GCC 11.2.0]
|
5 |
+
timestamp: 1690041531.683268
|
6 |
+
torch: 1.13.1
|
7 |
+
yaml_files:
|
8 |
+
asr_train_config: exp/asr_train_asr_whisper_full_warmup3000_raw_en_whisper_multilingual/config.yaml
|