metadata
license: mit
base_model: microsoft/speecht5_tts
tags:
- generated_from_trainer
datasets:
- common_voice_1_0
model-index:
- name: speecht5_tts
results: []
speecht5_tts
This model is a fine-tuned version of microsoft/speecht5_tts on the common_voice_1_0 dataset. It achieves the following results on the evaluation set:
- Loss: 0.5189
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: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 10000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.5395 | 16.0 | 1000 | 0.4726 |
| 0.4727 | 32.0 | 2000 | 0.4819 |
| 0.4513 | 48.0 | 3000 | 0.4871 |
| 0.4526 | 64.0 | 4000 | 0.5006 |
| 0.4474 | 80.0 | 5000 | 0.5022 |
| 0.4147 | 96.0 | 6000 | 0.5039 |
| 0.423 | 112.0 | 7000 | 0.5154 |
| 0.4271 | 128.0 | 8000 | 0.5217 |
| 0.4232 | 144.0 | 9000 | 0.5198 |
| 0.4044 | 160.0 | 10000 | 0.5189 |
Framework versions
- Transformers 4.36.0.dev0
- Pytorch 2.1.0+cu121
- Datasets 2.14.6
- Tokenizers 0.14.1