Model save

README.md

@@ -4,7 +4,7 @@ base_model: microsoft/speecht5_tts
 tags:
 - generated_from_trainer
 datasets:
-- voxpopuli
+- common_voice_1_0
 model-index:
 - name: speecht5_tts
   results: []
@@ -15,9 +15,9 @@ should probably proofread and complete it, then remove this comment. -->
 
 # speecht5_tts
 
-This model is a fine-tuned version of [microsoft/speecht5_tts](https://huggingface.co/microsoft/speecht5_tts) on the voxpopuli dataset.
+This model is a fine-tuned version of [microsoft/speecht5_tts](https://huggingface.co/microsoft/speecht5_tts) on the common_voice_1_0 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.7094
+- Loss: 0.7290
 
 ## Model description
 
@@ -37,25 +37,21 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 1e-05
-- train_batch_size: 32
+- train_batch_size: 8
 - eval_batch_size: 8
 - seed: 42
 - gradient_accumulation_steps: 2
-- total_train_batch_size: 64
+- total_train_batch_size: 16
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- lr_scheduler_warmup_steps: 500
-- training_steps: 4000
+- training_steps: 100
 - mixed_precision_training: Native AMP
 
 ### Training results
 
-| Training Loss | Epoch  | Step | Validation Loss |
-|:-------------:|:------:|:----:|:---------------:|
-| 0.1945        | 1000.0 | 1000 | 0.6777          |
-| 0.1781        | 2000.0 | 2000 | 0.6907          |
-| 0.1697        | 3000.0 | 3000 | 0.7122          |
-| 0.1699        | 4000.0 | 4000 | 0.7094          |
+| Training Loss | Epoch | Step | Validation Loss |
+|:-------------:|:-----:|:----:|:---------------:|
+| 0.7528        | 0.8   | 100  | 0.7290          |
 
 
 ### Framework versions