Update README.md
Browse files
README.md
CHANGED
|
@@ -64,7 +64,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 64 |
|
| 65 |
# deepdml/whisper-medium-mix-fr
|
| 66 |
|
| 67 |
-
This model is a fine-tuned version of [deepdml/whisper-medium-mix-fr](https://huggingface.co/deepdml/whisper-medium-mix-fr) on the common_voice_11_0
|
| 68 |
It achieves the following results on the evaluation set:
|
| 69 |
- Loss: 0.2599
|
| 70 |
- Wer: 11.2278
|
|
@@ -79,7 +79,13 @@ More information needed
|
|
| 79 |
|
| 80 |
## Training and evaluation data
|
| 81 |
|
| 82 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 83 |
|
| 84 |
## Training procedure
|
| 85 |
|
|
|
|
| 64 |
|
| 65 |
# deepdml/whisper-medium-mix-fr
|
| 66 |
|
| 67 |
+
This model is a fine-tuned version of [deepdml/whisper-medium-mix-fr](https://huggingface.co/deepdml/whisper-medium-mix-fr) on the mozilla-foundation/common_voice_11_0, google/fleurs, facebook/multilingual_librispeech and facebook/voxpopuli datasets.
|
| 68 |
It achieves the following results on the evaluation set:
|
| 69 |
- Loss: 0.2599
|
| 70 |
- Wer: 11.2278
|
|
|
|
| 79 |
|
| 80 |
## Training and evaluation data
|
| 81 |
|
| 82 |
+
Training data used:
|
| 83 |
+
- **mozilla-foundation/common_voice_11_0:** fr, train+validation
|
| 84 |
+
- **google/fleurs:** fr_fr, train
|
| 85 |
+
- **facebook/multilingual_librispeech:** french, train
|
| 86 |
+
- **facebook/voxpopuli:** fr, train
|
| 87 |
+
-
|
| 88 |
+
Evaluating over test split from mozilla-foundation/common_voice_11_0 dataset.
|
| 89 |
|
| 90 |
## Training procedure
|
| 91 |
|