Update model
Browse files- README.md +69 -1
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/10epoch.pth +3 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/RESULTS.md +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/config.yaml +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/acc.png +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/backward_time.png +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/cer.png +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/cer_ctc.png +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/cer_interctc_layer3.png +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/clip.png +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/forward_time.png +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/gpu_max_cached_mem_GB.png +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/grad_norm.png +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/iter_time.png +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/loss.png +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/loss_att.png +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/loss_ctc.png +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/loss_interctc_layer3.png +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/loss_scale.png +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/optim0_lr0.png +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/optim_step_time.png +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/train_time.png +0 -0
- exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/wer.png +0 -0
- meta.yaml +8 -0
README.md
CHANGED
@@ -1,3 +1,71 @@
|
|
1 |
---
|
2 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
+
tags:
|
3 |
+
- espnet
|
4 |
+
- audio
|
5 |
+
- automatic-speech-recognition
|
6 |
+
language: multilingual
|
7 |
+
datasets:
|
8 |
+
- fleurs
|
9 |
+
license: cc-by-4.0
|
10 |
---
|
11 |
+
|
12 |
+
## ESPnet2 ASR model
|
13 |
+
|
14 |
+
### `akreal/bloomzmms-ctc`
|
15 |
+
|
16 |
+
This model was trained by Pavel Denisov using fleurs 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 99cc51a342acfd13ca2b7475bd32c636f467ca0f
|
26 |
+
pip install -e .
|
27 |
+
cd egs2/fleurs/asr1
|
28 |
+
./run.sh --skip_data_prep false --skip_train true --download_model akreal/bloomzmms-ctc
|
29 |
+
```
|
30 |
+
|
31 |
+
|
32 |
+
|
33 |
+
## ASR config
|
34 |
+
|
35 |
+
<details><summary>expand</summary>
|
36 |
+
|
37 |
+
|
38 |
+
</details>
|
39 |
+
|
40 |
+
|
41 |
+
|
42 |
+
### Citing ESPnet
|
43 |
+
|
44 |
+
```BibTex
|
45 |
+
@inproceedings{watanabe2018espnet,
|
46 |
+
author={Shinji Watanabe and Takaaki Hori and Shigeki Karita and Tomoki Hayashi and Jiro Nishitoba and Yuya Unno and Nelson Yalta and Jahn Heymann and Matthew Wiesner and Nanxin Chen and Adithya Renduchintala and Tsubasa Ochiai},
|
47 |
+
title={{ESPnet}: End-to-End Speech Processing Toolkit},
|
48 |
+
year={2018},
|
49 |
+
booktitle={Proceedings of Interspeech},
|
50 |
+
pages={2207--2211},
|
51 |
+
doi={10.21437/Interspeech.2018-1456},
|
52 |
+
url={http://dx.doi.org/10.21437/Interspeech.2018-1456}
|
53 |
+
}
|
54 |
+
|
55 |
+
|
56 |
+
|
57 |
+
|
58 |
+
```
|
59 |
+
|
60 |
+
or arXiv:
|
61 |
+
|
62 |
+
```bibtex
|
63 |
+
@misc{watanabe2018espnet,
|
64 |
+
title={ESPnet: End-to-End Speech Processing Toolkit},
|
65 |
+
author={Shinji Watanabe and Takaaki Hori and Shigeki Karita and Tomoki Hayashi and Jiro Nishitoba and Yuya Unno and Nelson Yalta and Jahn Heymann and Matthew Wiesner and Nanxin Chen and Adithya Renduchintala and Tsubasa Ochiai},
|
66 |
+
year={2018},
|
67 |
+
eprint={1804.00015},
|
68 |
+
archivePrefix={arXiv},
|
69 |
+
primaryClass={cs.CL}
|
70 |
+
}
|
71 |
+
```
|
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/10epoch.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f63c3b933022129d914c8914e9998cc6fc170b283507aec0cd58c5d0e73308d4
|
3 |
+
size 10114401449
|
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/RESULTS.md
ADDED
File without changes
|
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/config.yaml
ADDED
The diff for this file is too large to render.
See raw diff
|
|
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/acc.png
ADDED
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/backward_time.png
ADDED
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/cer.png
ADDED
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/cer_ctc.png
ADDED
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/cer_interctc_layer3.png
ADDED
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/clip.png
ADDED
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/forward_time.png
ADDED
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/gpu_max_cached_mem_GB.png
ADDED
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/grad_norm.png
ADDED
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/iter_time.png
ADDED
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/loss.png
ADDED
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/loss_att.png
ADDED
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/loss_ctc.png
ADDED
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/loss_interctc_layer3.png
ADDED
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/loss_scale.png
ADDED
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/optim0_lr0.png
ADDED
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/optim_step_time.png
ADDED
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/train_time.png
ADDED
exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/images/wer.png
ADDED
meta.yaml
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
espnet: '202304'
|
2 |
+
files:
|
3 |
+
asr_model_file: exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/10epoch.pth
|
4 |
+
python: "3.9.18 (main, Feb 28 2024, 00:00:00) \n[GCC 13.2.1 20231205 (Red Hat 13.2.1-6)]"
|
5 |
+
timestamp: 1711918248.319074
|
6 |
+
torch: 2.1.2+cu121
|
7 |
+
yaml_files:
|
8 |
+
asr_train_config: exp/asr_train_asr_e_branchformer_mms1b-asr_bloomz7b_ctc_raw_hugging_face_bigscience-bloomz-7b1_sp/config.yaml
|