End of training

README.md

@@ -18,8 +18,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [openai/whisper-base](https://huggingface.co/openai/whisper-base) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.9605
-- Wer: 40.5810
+- Loss: 0.9478
+- Wer: 40.9025
 
 ## Model description
 
@@ -52,10 +52,10 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss | Wer     |
 |:-------------:|:-----:|:----:|:---------------:|:-------:|
-| 0.151         | 10.0  | 500  | 0.7583          | 45.9840 |
-| 0.0226        | 20.0  | 1000 | 0.9046          | 43.0211 |
-| 0.0033        | 30.0  | 1500 | 0.9496          | 41.0893 |
-| 0.0006        | 40.0  | 2000 | 0.9605          | 40.5810 |
+| 0.5893        | 10.0  | 500  | 0.7149          | 51.7322 |
+| 0.0835        | 20.0  | 1000 | 0.8352          | 42.5328 |
+| 0.0199        | 30.0  | 1500 | 0.9229          | 41.2373 |
+| 0.0066        | 40.0  | 2000 | 0.9478          | 40.9025 |
 
 
 ### Framework versions

generation_config.json

@@ -39,7 +39,7 @@
   ],
   "bos_token_id": 50257,
   "decoder_start_token_id": 50258,
-  "dropout": 0.01,
+  "dropout": 0.1,
   "eos_token_id": 50257,
   "is_multilingual": true,
   "lang_to_id": {