Fine-tuned SpeechT5 on Amharic TTS
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: 0.4262
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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 2000
- training_steps: 16000
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss |
---|---|---|---|
0.5695 | 2.08 | 1000 | 0.5083 |
0.5247 | 4.16 | 2000 | 0.4753 |
0.495 | 6.24 | 3000 | 0.4570 |
0.4861 | 8.32 | 4000 | 0.4478 |
0.49 | 10.4 | 5000 | 0.4426 |
0.4868 | 12.47 | 6000 | 0.4395 |
0.4847 | 14.55 | 7000 | 0.4367 |
0.4733 | 16.63 | 8000 | 0.4336 |
0.4678 | 18.71 | 9000 | 0.4323 |
0.4608 | 20.79 | 10000 | 0.4301 |
0.4628 | 22.87 | 11000 | 0.4301 |
0.4532 | 24.95 | 12000 | 0.4284 |
0.4658 | 27.03 | 13000 | 0.4277 |
0.4733 | 29.11 | 14000 | 0.4278 |
0.4554 | 31.19 | 15000 | 0.4269 |
0.4542 | 33.26 | 16000 | 0.4262 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.0
- Downloads last month
- 3