File size: 896 Bytes
a3c0c78
 
 
84403d1
a3c0c78
2df817d
a3c0c78
 
ea349b4
6d16672
f025181
 
b7b2571
 
 
 
2519041
6d16672
a218480
bbee575
84403d1
f73f733
84403d1
 
 
 
 
 
9e7aa43
 
 
 
bbd9364
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
---
language: 
  - uk
license: "apache-2.0"
datasets:
- mozilla-foundation/common_voice_10_0
---

🇺🇦 Join Ukrainian Speech Recognition Community - https://t.me/speech_recognition_uk

⭐ See other Ukrainian models - https://github.com/egorsmkv/speech-recognition-uk

**ATTENTION!**

> USE UPDATED MODEL: https://huggingface.co/Yehor/w2v-bert-2.0-uk

This model has apostrophes and hyphens.

The language model is trained on the texts of the Common Voice dataset, which is used during training.

Metrics:

| Dataset | CER | WER |
|-|-|-|
| CV7 (no LM) |  0.0432 | 0.2288 |
| CV7 (with LM) | 0.0169 | 0.0706 |
| CV10 (no LM) | 0.0412 | 0.2206 |
| CV10 (with LM) | 0.0118 | 0.0463 |

More:

- The same model, but trained on noisy data: https://huggingface.co/Yehor/wav2vec2-xls-r-300m-uk-with-small-lm-noisy
- Traced JIT version: https://huggingface.co/Yehor/wav2vec2-xls-r-300m-uk-traced-jit