Model save

README.md

@@ -18,7 +18,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [microsoft/speecht5_tts](https://huggingface.co/microsoft/speecht5_tts) on the common_voice_17_0 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.6216
+- Loss: 0.5678
 
 ## Model description
 
@@ -52,15 +52,15 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| 0.7014        | 500.0  | 1000 | 0.5671          |
-| 0.6384        | 1000.0 | 2000 | 0.6343          |
-| 0.5742        | 1500.0 | 3000 | 0.6314          |
-| 0.5597        | 2000.0 | 4000 | 0.6216          |
+| 0.3551        | 500.0  | 1000 | 0.4988          |
+| 0.3218        | 1000.0 | 2000 | 0.5528          |
+| 0.2891        | 1500.0 | 3000 | 0.5593          |
+| 0.282         | 2000.0 | 4000 | 0.5678          |
 
 
 ### Framework versions
 
-- Transformers 4.46.0.dev0
+- Transformers 4.47.0.dev0
 - Pytorch 2.5.0+cu121
-- Datasets 3.0.2
-- Tokenizers 0.20.1
+- Datasets 3.1.0
+- Tokenizers 0.20.3

generation_config.json

@@ -5,5 +5,5 @@
   "eos_token_id": 2,
   "max_length": 1876,
   "pad_token_id": 1,
-  "transformers_version": "4.46.0.dev0"
+  "transformers_version": "4.47.0.dev0"
 }