metadata
license: mit
base_model: microsoft/speecht5_tts
tags:
- generated_from_trainer
model-index:
- name: speecht5_allvoices_semcommon
results: []
speecht5_allvoices_semcommon
This model is a fine-tuned version of microsoft/speecht5_tts on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.3819
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-06
- train_batch_size: 4
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 8
- 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 |
|---|---|---|---|
| 0.5047 | 1.48 | 1000 | 0.4460 |
| 0.4601 | 2.96 | 2000 | 0.4096 |
| 0.4471 | 4.44 | 3000 | 0.3987 |
| 0.4384 | 5.92 | 4000 | 0.3911 |
| 0.4314 | 7.4 | 5000 | 0.3909 |
| 0.4285 | 8.88 | 6000 | 0.3861 |
| 0.427 | 10.36 | 7000 | 0.3842 |
| 0.4307 | 11.84 | 8000 | 0.3839 |
| 0.4244 | 13.32 | 9000 | 0.3829 |
| 0.4228 | 14.8 | 10000 | 0.3819 |
Framework versions
- Transformers 4.39.0.dev0
- Pytorch 2.2.1
- Datasets 2.18.0
- Tokenizers 0.15.2