SpeechT5 Finetuned VoxPopuli Spanish
This model is a fine-tuned version of microsoft/speecht5_tts on the VoxPopuli Spanish dataset. It achieves the following results on the evaluation set:
- Loss: 0.4376
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: 1e-05
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 16
- 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: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 4000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 8.5272 | 1.3521 | 500 | 0.4793 |
| 7.9654 | 2.7042 | 1000 | 0.4562 |
| 7.6000 | 4.0542 | 1500 | 0.4480 |
| 7.6937 | 5.4063 | 2000 | 0.4424 |
| 7.6689 | 6.7584 | 2500 | 0.4405 |
| 7.5710 | 8.1083 | 3000 | 0.4385 |
| 7.5896 | 9.4605 | 3500 | 0.4379 |
| 7.5567 | 10.8126 | 4000 | 0.4376 |
Framework versions
- Transformers 5.0.1.dev0
- Pytorch 2.9.0+cu126
- Datasets 2.18.0
- Tokenizers 0.22.2