File size: 1,913 Bytes
61c55fa
ad60f70
 
 
 
 
a53d524
 
 
61c55fa
a53d524
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
61c55fa
a53d524
 
 
 
 
 
5fbce16
a53d524
 
 
 
 
 
 
 
49409b9
45f78a2
a53d524
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
---
tags:
  - espnet
  - audio
  - automatic-speech-recognition

language:
- et

license: apache-2.0

metrics:
- wer

model-index:
- name: e-branchformer et

  results:
  - task:
  
      name: Automatic Speech Recognition
  
      type: automatic-speech-recognition
  
    dataset:
  
      name: ERR2020
  
      type: audio
  
    metrics:
  
    - name: Wer
    
      type: wer
    
      value: 9.9
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# e-branchformer et

Espnet2 e-branchformer based recipe (https://github.com/espnet/espnet/tree/master/egs2/librispeech_100/asr1) trained Estonian ASR model using ERR2020 dataset
- WER on ERR2020: 9.9
- WER on mozilla commonvoice_11: 20.8

 
For usage:
- clone this repo (`git clone https://huggingface.co/rristo/espnet_ebranchformer_et`)
- go to repo (`cd espnet_ebranchformer_et`)
- build docker image for needed libraries (`build.sh` or `build.bat`)
- run docker container (`run.sh` or `run.sh`). This mounts current directory
- run notebook `example_usage.ipynb` for example usage
  - currently expects audio to be in .wav format

## Model description

ASR model for Estonian, uses Estonian Public Broadcasting data ERR2020 data (around 340 hours of audio)

## Intended uses & limitations

Pretty much a toy model, trained on limited amount of data. Might not work well on data out of domain 
(especially spontaneous/noisy data).

## Training and evaluation data

Trained on ERR2020 data, evaluated on ERR2020 and mozilla commonvoice test data.

## Training procedure

Used espnet e-branchformer based recipe (https://github.com/espnet/espnet/tree/master/egs2/librispeech_100/asr1)

### Training results


Look into folder exp/images.

Validation results are in exp/RESULTS.md

### Framework versions

- espnet2