File size: 2,451 Bytes
7a9c9d5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
6c68b02
7a9c9d5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
---
language:
- fi
license: apache-2.0
tags:
- whisper-event
- finnish
- speech-recognition
datasets:
- mozilla-foundation/common_voice_11_0
- google/fleurs
metrics:
- wer
- cer
model-index:
- name: Whisper Large V3 Finnish
  results:
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: Common Voice 11.0
      type: mozilla-foundation/common_voice_11_0
      config: fi
      split: test
      args: fi
    metrics:
    - name: Wer
      type: wer
      value: 8.23
    - name: Cer
      type: cer
      value: 1.43
  - task:
      name: Automatic Speech Recognition
      type: automatic-speech-recognition
    dataset:
      name: FLEURS
      type: google/fleurs
      config: fi_fi
      split: test
      args: fi_fi
    metrics:
    - name: Wer
      type: wer
      value: 8.21
    - name: Cer
      type: cer
      value: 3.23
library_name: transformers
pipeline_tag: automatic-speech-recognition
---
# This is a conversion of [Finnish-NLP/whisper-large-finnish-v3](https://huggingface.co./Finnish-NLP/whisper-large-finnish-v3) into faster-whisper format.

<h3>This is our improved Whisper v3 model that is now finetuned from OpenAI Whisper Large V3 </h3>
<p>We improve from our previously finetuned Whisper V2 model in the following manner<a>https://huggingface.co./Finnish-NLP/whisper-large-v2-finnish</a> </p>
<p>CV11 (Common Voice 11 test set) WER (Word error rate) 10.42 --> 8.23</p>
<p>Fleurs (A speech recognition test set by Google) WER (Word error rate) 10.20 --> 8.21</p>
<p>Model was trained on Nvidia RTX4080 for 32k steps with batch size 8, gradient accumulation 2</p>

<br>

<h3> Original OpenAI Whisper Large V3</h3>
- CV11
  - WER: 14.81
  - WER NORMALIZED: 10.82
  - CER: 2.7
  - CER NORMALIZED: 2.07

- Fleurs
  - WER: 12.04
  - WER NORMALIZED: 9.63
  - CER: 2.48
  - CER NORMALIZED: 3.64


<h3> After Finetuning with Finnish data our V3 got these scores on the test set:</h3>

  - @14000 finetuning steps
    - CV11
      - WER: 11.36
      - WER NORMALIZED: 8.31
      - CER: 1.93
      - CER NORMALIZED: 1.48

    - Fleurs
      - WER: 10.2
      - WER NORMALIZED: 8.56
      - CER: 2.26
      - CER NORMALIZED: 3.54

  - @32000 finetuning steps
    - CV11
      - WER: 11.47
      - WER NORMALIZED: 8.23
      - CER: 1.91
      - CER NORMALIZED: 1.43

    - Fleurs
      - WER: 10.1
      - WER NORMALIZED: 8.21
      - CER: 2.2
      - CER NORMALIZED: 3.23