Commit
·
b2a47a2
1
Parent(s):
bc7a9dd
update model card README.md
Browse files
README.md
CHANGED
|
@@ -12,7 +12,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 12 |
|
| 13 |
# wav2vec2-large-asr-th
|
| 14 |
|
| 15 |
-
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on
|
|
|
|
|
|
|
|
|
|
| 16 |
|
| 17 |
## Model description
|
| 18 |
|
|
@@ -41,6 +44,17 @@ The following hyperparameters were used during training:
|
|
| 41 |
- training_steps: 1000
|
| 42 |
- mixed_precision_training: Native AMP
|
| 43 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 44 |
### Framework versions
|
| 45 |
|
| 46 |
- Transformers 4.26.1
|
|
|
|
| 12 |
|
| 13 |
# wav2vec2-large-asr-th
|
| 14 |
|
| 15 |
+
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the None dataset.
|
| 16 |
+
It achieves the following results on the evaluation set:
|
| 17 |
+
- Loss: 3.1431
|
| 18 |
+
- Cer: 0.9986
|
| 19 |
|
| 20 |
## Model description
|
| 21 |
|
|
|
|
| 44 |
- training_steps: 1000
|
| 45 |
- mixed_precision_training: Native AMP
|
| 46 |
|
| 47 |
+
### Training results
|
| 48 |
+
|
| 49 |
+
| Training Loss | Epoch | Step | Validation Loss | Cer |
|
| 50 |
+
|:-------------:|:-----:|:----:|:---------------:|:------:|
|
| 51 |
+
| 3.8093 | 1.13 | 200 | 3.7212 | 0.9999 |
|
| 52 |
+
| 3.5492 | 2.26 | 400 | 3.5525 | 0.9999 |
|
| 53 |
+
| 3.5977 | 3.39 | 600 | 3.5198 | 0.9999 |
|
| 54 |
+
| 3.2243 | 4.52 | 800 | 3.1431 | 0.9986 |
|
| 55 |
+
| 2.8584 | 5.65 | 1000 | 2.6861 | 0.8201 |
|
| 56 |
+
|
| 57 |
+
|
| 58 |
### Framework versions
|
| 59 |
|
| 60 |
- Transformers 4.26.1
|