metadata
library_name: transformers
license: mit
base_model: microsoft/speecht5_tts
tags:
- tts
- generated_from_trainer
datasets:
- microsoft/speecht5_tts
model-index:
- name: SyedNaqi_TechSpeechT5_TTS
results: []
SyedNaqi_TechSpeechT5_TTS
This model is a fine-tuned version of microsoft/speecht5_tts on the TTS_English_Technical_data dataset. It achieves the following results on the evaluation set:
- Loss: 0.4597
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: 14
- eval_batch_size: 10
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 28
- optimizer: Use adamw_torch 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: 1000
- training_steps: 4000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.3331 | 1.5649 | 500 | 0.5719 |
| 1.0675 | 3.1299 | 1000 | 0.4861 |
| 1.0427 | 4.6948 | 1500 | 0.4732 |
| 1.0088 | 6.2598 | 2000 | 0.4667 |
| 1.0052 | 7.8247 | 2500 | 0.4637 |
| 0.9862 | 9.3897 | 3000 | 0.4613 |
| 0.9788 | 10.9546 | 3500 | 0.4597 |
| 0.9792 | 12.5196 | 4000 | 0.4597 |
Framework versions
- Transformers 4.47.0.dev0
- Pytorch 2.5.0+cu121
- Datasets 3.0.2
- Tokenizers 0.20.1