metadata
library_name: transformers
license: mit
base_model: microsoft/speecht5_tts
tags:
- generated_from_trainer
model-index:
- name: speecht5_finetuned_librispeech_polish_epo10_batch2_gas5
results: []
speecht5_finetuned_librispeech_polish_epo10_batch2_gas5
This model is a fine-tuned version of microsoft/speecht5_tts on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.3722
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: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 5
- total_train_batch_size: 10
- 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: 200
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 2.2629 | 1.2749 | 200 | 0.4141 |
| 2.1473 | 2.5499 | 400 | 0.3946 |
| 2.06 | 3.8248 | 600 | 0.3917 |
| 1.9595 | 5.0959 | 800 | 0.3812 |
| 1.9753 | 6.3708 | 1000 | 0.3773 |
| 1.9364 | 7.6458 | 1200 | 0.3752 |
| 1.9184 | 8.9207 | 1400 | 0.3722 |
Framework versions
- Transformers 4.47.1
- Pytorch 2.5.1+cu121
- Datasets 3.2.0
- Tokenizers 0.21.0