metadata
library_name: transformers
language:
- en
license: mit
base_model: microsoft/speecht5_tts
tags:
- English
- generated_from_trainer
datasets:
- Yassmen/TTS_English_Technical_data
model-index:
- name: SpeechT5-fine-tune-en
results: []
SpeechT5-fine-tune-en
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: 0.4469
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: 0.0001
- train_batch_size: 4
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 32
- 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: 100
- num_epochs: 3.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.578 | 0.3870 | 100 | 0.5009 |
| 0.5417 | 0.7741 | 200 | 0.4890 |
| 0.5072 | 1.1611 | 300 | 0.4739 |
| 0.4956 | 1.5481 | 400 | 0.4615 |
| 0.4915 | 1.9352 | 500 | 0.4560 |
| 0.4837 | 2.3222 | 600 | 0.4498 |
| 0.4856 | 2.7092 | 700 | 0.4469 |
Framework versions
- Transformers 4.46.0.dev0
- Pytorch 2.4.1+cu121
- Datasets 3.0.1
- Tokenizers 0.20.1