Commit
•
caa1790
1
Parent(s):
eff1062
Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,57 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
tags:
|
| 4 |
+
- automatic-speech-recognition
|
| 5 |
+
- asr
|
| 6 |
+
- pytorch
|
| 7 |
+
- wav2vec2
|
| 8 |
+
- wolof
|
| 9 |
+
|
| 10 |
+
---
|
| 11 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 12 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 13 |
+
# wav2vec2-xls-r-300m-wolof
|
| 14 |
+
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the Common Voice 7.0 dataset.
|
| 15 |
+
It achieves the following results on the evaluation set:
|
| 16 |
+
- Loss:0.367826
|
| 17 |
+
- Wer: 0.212565
|
| 18 |
+
|
| 19 |
+
## Model description
|
| 20 |
+
More information needed
|
| 21 |
+
## Intended uses & limitations
|
| 22 |
+
More information needed
|
| 23 |
+
## Training and evaluation data
|
| 24 |
+
More information needed
|
| 25 |
+
## Training procedure
|
| 26 |
+
### Training hyperparameters
|
| 27 |
+
More information needed
|
| 28 |
+
### Training results
|
| 29 |
+
|
| 30 |
+
|
| 31 |
+
|
| 32 |
+
| Step | Training Loss | Validation Loss | Wer |
|
| 33 |
+
|:-------:|:-------------:|:---------------:|:------:|
|
| 34 |
+
| 1500 | 2.854200 |0.642243 |0.543964 |
|
| 35 |
+
| 3000 | 0.599200 | 0.468138 | 0.429549|
|
| 36 |
+
| 4500 | 0.468300 | 0.433436 | 0.405644|
|
| 37 |
+
| 6000 | 0.427000 | 0.384873 | 0.344150|
|
| 38 |
+
| 7500 | 0.377000 | 0.374003 | 0.323892|
|
| 39 |
+
| 9000 | 0.337000 | 0.363674 | 0.306189|
|
| 40 |
+
| 10500 | 0.302400 | 0.349884 |0 .283908 |
|
| 41 |
+
| 12000 | 0.264100 | 0.344104 |0.277120|
|
| 42 |
+
| 13500 |0 .254000 |0.341820 |0.271316|
|
| 43 |
+
| 15000 | 0.208400| 0.326502 | 0.260695|
|
| 44 |
+
| 16500 | 0.203500| 0.326209 | 0.250313|
|
| 45 |
+
| 18000 |0.159800 |0.323539 | 0.239851|
|
| 46 |
+
| 19500 | 0.158200 | 0.310694 | 0.230028|
|
| 47 |
+
| 21000 | 0.132800 | 0.338318 | 0.229283|
|
| 48 |
+
| 22500 | 0.112800 | 0.336765 | 0.224145|
|
| 49 |
+
| 24000 | 0.103600 | 0.350208 | 0.227073 |
|
| 50 |
+
| 25500 | 0.091400 | 0.353609 | 0.221589 |
|
| 51 |
+
| 27000 | 0.084400 | 0.367826 | 0.212565 |
|
| 52 |
+
|
| 53 |
+
|
| 54 |
+
### Framework versions
|
| 55 |
+
- Transformers 4.11
|
| 56 |
+
- Pytorch 1.10.0
|
| 57 |
+
- Datasets 1.13
|