Commit
·
47c5523
1
Parent(s):
9612251
Update README.md
Browse files
README.md
CHANGED
|
@@ -25,29 +25,21 @@ model-index:
|
|
| 25 |
value: 35.613112100364226
|
| 26 |
---
|
| 27 |
|
| 28 |
-
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 29 |
-
should probably proofread and complete it, then remove this comment. -->
|
| 30 |
-
|
| 31 |
# Whisper Small Lithuanian and Serbian sequentially trained
|
| 32 |
|
| 33 |
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the Common Voice 11.0 dataset.
|
| 34 |
It achieves the following results on the evaluation set:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 35 |
- Loss: 0.6311
|
| 36 |
- Wer: 35.6131
|
| 37 |
|
| 38 |
-
## Model description
|
| 39 |
-
|
| 40 |
-
More information needed
|
| 41 |
-
|
| 42 |
-
## Intended uses & limitations
|
| 43 |
-
|
| 44 |
-
More information needed
|
| 45 |
-
|
| 46 |
-
## Training and evaluation data
|
| 47 |
-
|
| 48 |
-
More information needed
|
| 49 |
-
|
| 50 |
## Training procedure
|
|
|
|
| 51 |
|
| 52 |
### Training hyperparameters
|
| 53 |
|
|
@@ -62,7 +54,7 @@ The following hyperparameters were used during training:
|
|
| 62 |
- training_steps: 2000
|
| 63 |
- mixed_precision_training: Native AMP
|
| 64 |
|
| 65 |
-
### Training results
|
| 66 |
|
| 67 |
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 68 |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
|
|
|
|
| 25 |
value: 35.613112100364226
|
| 26 |
---
|
| 27 |
|
|
|
|
|
|
|
|
|
|
| 28 |
# Whisper Small Lithuanian and Serbian sequentially trained
|
| 29 |
|
| 30 |
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the Common Voice 11.0 dataset.
|
| 31 |
It achieves the following results on the evaluation set:
|
| 32 |
+
|
| 33 |
+
### Lithuanian
|
| 34 |
+
- Loss:
|
| 35 |
+
- Wer:
|
| 36 |
+
|
| 37 |
+
### Serbian
|
| 38 |
- Loss: 0.6311
|
| 39 |
- Wer: 35.6131
|
| 40 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 41 |
## Training procedure
|
| 42 |
+
It was first trained 2000 steps on Lithuanian and then 2000 steps on Serbian, continuing from the last checkpoint for Lithuanian.
|
| 43 |
|
| 44 |
### Training hyperparameters
|
| 45 |
|
|
|
|
| 54 |
- training_steps: 2000
|
| 55 |
- mixed_precision_training: Native AMP
|
| 56 |
|
| 57 |
+
### Training results for Serbian
|
| 58 |
|
| 59 |
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 60 |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
|