stillerman/stammer

README.md

@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3328
-- Wer: 39.2157
+- Loss: 0.4497
+- Wer: 35.7430
 
 ## Model description
 
@@ -44,15 +44,17 @@ The following hyperparameters were used during training:
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_steps: 10
-- training_steps: 50
+- training_steps: 100
 - mixed_precision_training: Native AMP
 
 ### Training results
 
 | Training Loss | Epoch  | Step | Validation Loss | Wer     |
 |:-------------:|:------:|:----:|:---------------:|:-------:|
-| 0.4879        | 3.5714 | 25   | 0.4535          | 37.2549 |
-| 0.0704        | 7.1429 | 50   | 0.3328          | 39.2157 |
+| 0.9039        | 0.3968 | 25   | 0.7568          | 42.1687 |
+| 0.4742        | 0.7937 | 50   | 0.4766          | 35.9438 |
+| 0.3073        | 1.1905 | 75   | 0.4589          | 35.1406 |
+| 0.3061        | 1.5873 | 100  | 0.4497          | 35.7430 |
 
 
 ### Framework versions