Automatic Speech Recognition
TensorBoard
Safetensors
Welsh
wav2vec2
Generated from Trainer
File size: 887 Bytes
7db016f
 
 
 
1c61e48
7db016f
 
 
 
 
 
b594d32
 
 
 
 
 
7db016f
 
b594d32
7db016f
b594d32
 
 
7db016f
b594d32
7db016f
b594d32
 
 
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
---
license: apache-2.0
base_model: facebook/wav2vec2-large-xlsr-53
tags:
- automatic-speech-recognition
- generated_from_trainer
metrics:
- wer
model-index:
- name: wav2vec2-xlsr-53-ft-btb-cv-cy-cand
  results: []
datasets:
- techiaith/commonvoice_18_0_cy
- techiaith/banc-trawsgrifiadau-bangor
language:
- cy
pipeline_tag: automatic-speech-recognition
---

# wav2vec2-xlsr-53-ft-btb-cv-cy

This model is a version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) 
fine-tuned on Welsh language read-speech from [commonvoice_18_cy](https://huggingface.co/datasets/techiaith/commonvoice_18_0_cy) 
and spontaneous speech from [Bangor Transcriptions 24.10](https://huggingface.co/datasets/techiaith/banc-trawsgrifiadau-bangor/tree/24.10)

It achieves the following WER results on the Bangor Transcriptions test set:

- WER: 36.23
- CER: 12.55