End of training

README.md

@@ -15,7 +15,7 @@ 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 an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.0584
+- Loss: 0.7314
 
 ## Model description
 
@@ -36,19 +36,24 @@ More information needed
 The following hyperparameters were used during training:
 - learning_rate: 1e-05
 - train_batch_size: 4
-- eval_batch_size: 3
+- eval_batch_size: 2
 - seed: 42
+- gradient_accumulation_steps: 8
+- total_train_batch_size: 32
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
-- num_epochs: 3.0
+- lr_scheduler_warmup_steps: 500
+- training_steps: 4000
+- mixed_precision_training: Native AMP
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
-| 1.8403        | 0.83  | 500  | 1.2271          |
-| 1.2393        | 1.65  | 1000 | 1.1105          |
-| 1.1409        | 2.48  | 1500 | 1.0584          |
+| 0.9147        | 13.22 | 1000 | 0.8691          |
+| 0.812         | 26.45 | 2000 | 0.7710          |
+| 0.7837        | 39.67 | 3000 | 0.7394          |
+| 0.7648        | 52.89 | 4000 | 0.7314          |
 
 
 ### Framework versions