Upload README.md
Browse files
README.md
ADDED
@@ -0,0 +1,67 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
language:
|
3 |
+
- fr
|
4 |
+
license: apache-2.0
|
5 |
+
tags:
|
6 |
+
- whisper
|
7 |
+
- generated_from_trainer
|
8 |
+
datasets:
|
9 |
+
- mozilla-foundation/common_voice_15_0
|
10 |
+
- BrunoHays/multilingual-tedx-fr
|
11 |
+
- PolyAI/minds14
|
12 |
+
- facebook/multilingual_librispeech
|
13 |
+
- facebook/voxpopuli
|
14 |
+
- google/fleurs
|
15 |
+
metrics:
|
16 |
+
- wer
|
17 |
+
model-index:
|
18 |
+
- name: Whisper tiny French
|
19 |
+
results:
|
20 |
+
- task:
|
21 |
+
name: Automatic Speech Recognition
|
22 |
+
type: automatic-speech-recognition
|
23 |
+
dataset1:
|
24 |
+
name: mozilla-foundation/common_voice_15_0 fr
|
25 |
+
type: mozilla-foundation/common_voice_15_0
|
26 |
+
config: fr
|
27 |
+
split: test
|
28 |
+
args: fr
|
29 |
+
metrics:
|
30 |
+
- name: Wer
|
31 |
+
type: wer
|
32 |
+
value: 40.0
|
33 |
+
dataset2:
|
34 |
+
name: facebook/multilingual_librispeech fr
|
35 |
+
type: facebook/multilingual_librispeech
|
36 |
+
config: fr
|
37 |
+
split: test
|
38 |
+
args: fr
|
39 |
+
wer : 26.1
|
40 |
+
dataset3:
|
41 |
+
name: facebook/voxpopuli fr
|
42 |
+
type: facebook/voxpopuli
|
43 |
+
config: fr
|
44 |
+
split: test
|
45 |
+
args: fr
|
46 |
+
wer : 29.4
|
47 |
+
dataset4:
|
48 |
+
name: google/fleurs fr
|
49 |
+
type: google/fleurs
|
50 |
+
config: fr
|
51 |
+
split: test
|
52 |
+
args: fr
|
53 |
+
wer : 33.7
|
54 |
+
---
|
55 |
+
|
56 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
57 |
+
should probably proofread and complete it, then remove this comment. -->
|
58 |
+
|
59 |
+
# Whisper tiny fr - JaepaX
|
60 |
+
This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the fr datasets.
|
61 |
+
|
62 |
+
## WER Result
|
63 |
+
It achieves the following results on the evaluation sets
|
64 |
+
- Mulit-Libri : "26.1",
|
65 |
+
- common : "40.0"
|
66 |
+
- voxpopuli : "29.4"
|
67 |
+
- fleurs : "33.7"
|