senga-nt-fused-tts
This model is a fine-tuned version of microsoft/speecht5_tts on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0868
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 3407
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 2000
- training_steps: 20000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.1376 | 9.0499 | 2000 | 0.1013 |
| 0.1242 | 18.0999 | 4000 | 0.0932 |
| 0.1111 | 27.1498 | 6000 | 0.0903 |
| 0.1081 | 36.1998 | 8000 | 0.0895 |
| 0.1038 | 45.2497 | 10000 | 0.0889 |
| 0.0995 | 54.2997 | 12000 | 0.0870 |
| 0.1009 | 63.3496 | 14000 | 0.0875 |
| 0.0969 | 72.3995 | 16000 | 0.0868 |
| 0.0997 | 81.4495 | 18000 | 0.0871 |
| 0.0962 | 90.4994 | 20000 | 0.0868 |
Framework versions
- Transformers 4.57.1
- Pytorch 2.8.0+cu128
- Datasets 4.2.0
- Tokenizers 0.22.2