|
--- |
|
license: mit |
|
base_model: microsoft/speecht5_tts |
|
tags: |
|
- generated_from_trainer |
|
model-index: |
|
- name: speecht5_finetuned_kazakh_tts2 |
|
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_finetuned_kazakh_tts2 |
|
|
|
This model is a fine-tuned version of [microsoft/speecht5_tts](https://huggingface.co/microsoft/speecht5_tts) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.5067 |
|
|
|
## 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: 2 |
|
- eval_batch_size: 2 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 16 |
|
- 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: 200 |
|
- training_steps: 1000 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 0.7136 | 0.03 | 100 | 0.6539 | |
|
| 0.6471 | 0.06 | 200 | 0.5934 | |
|
| 0.5851 | 0.08 | 300 | 0.5392 | |
|
| 0.5764 | 0.11 | 400 | 0.5275 | |
|
| 0.5666 | 0.14 | 500 | 0.5213 | |
|
| 0.5577 | 0.17 | 600 | 0.5138 | |
|
| 0.5605 | 0.2 | 700 | 0.5115 | |
|
| 0.5622 | 0.22 | 800 | 0.5088 | |
|
| 0.5603 | 0.25 | 900 | 0.5082 | |
|
| 0.558 | 0.28 | 1000 | 0.5067 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.38.1 |
|
- Pytorch 2.2.1+cu118 |
|
- Datasets 2.18.0 |
|
- Tokenizers 0.15.2 |
|
|
