Speecht5 Finetuned Voxpopuli It

🚀 SpeechT5-it

This model, SpeechT5-it, is a fine - tuned variant of microsoft/speecht5_tts on the VOXPOPULI dataset. It offers high - quality text - to - speech conversion and achieves a loss of 0.46 on the evaluation set.

📚 Documentation

Model Information

Evaluation Results

This model achieves the following results on the evaluation set:

• Loss: 0.4600

Training and Evaluation Data

This model was trained on the VOXPOPULI dataset with the Italian configuration (it). The evaluation was performed on the validation split of the same dataset.

Training Procedure

Training Hyperparameters

The following hyperparameters were used during training:

• learning_rate: 1e - 05
• train_batch_size: 4
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• 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: 100
• num_epochs: 40

Training Results

Framework Versions

• Transformers 4.30.0.dev0
• Pytorch 2.0.1+cu117
• Datasets 2.13.1
• Tokenizers 0.13.3

📄 License

This model is released under the MIT license.