|
--- |
|
license: mit |
|
base_model: microsoft/speecht5_tts |
|
tags: |
|
- generated_from_trainer |
|
datasets: |
|
- common_voice_16_1 |
|
model-index: |
|
- name: speecht5_tts_common_voice_uk |
|
results: [] |
|
--- |
|
|
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You |
|
should probably proofread and complete it, then remove this comment. --> |
|
|
|
# speecht5_tts_common_voice_uk |
|
|
|
This model is a fine-tuned version of [microsoft/speecht5_tts](https://huggingface.co/microsoft/speecht5_tts) on the common_voice_16_1 dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.4191 |
|
|
|
## 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: 32 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 500 |
|
- num_epochs: 10 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 0.902 | 1.0 | 146 | 0.7113 | |
|
| 0.7054 | 2.0 | 292 | 0.4888 | |
|
| 0.5368 | 3.0 | 438 | 0.4549 | |
|
| 0.5188 | 4.0 | 584 | 0.4441 | |
|
| 0.5007 | 5.0 | 730 | 0.4526 | |
|
| 0.4836 | 6.0 | 876 | 0.4299 | |
|
| 0.4722 | 7.0 | 1022 | 0.4249 | |
|
| 0.4666 | 8.0 | 1168 | 0.4234 | |
|
| 0.4652 | 9.0 | 1314 | 0.4218 | |
|
| 0.4639 | 10.0 | 1460 | 0.4191 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.37.2 |
|
- Pytorch 1.12.1+cu116 |
|
- Datasets 2.4.0 |
|
- Tokenizers 0.15.2 |
|
|