metadata
license: mit
base_model: microsoft/speecht5_tts
tags:
- generated_from_trainer
datasets:
- lj_speech
model-index:
- name: speecht5_finetuned_sysdesign
results: []
speecht5_finetuned_sysdesign
This model is a fine-tuned version of microsoft/speecht5_tts on the lj_speech dataset. It achieves the following results on the evaluation set:
- Loss: 0.3717
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: 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: 100
- training_steps: 1500
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 0.4684 | 0.31 | 100 | 0.4141 |
| 0.4496 | 0.61 | 200 | 0.4108 |
| 0.4334 | 0.92 | 300 | 0.3955 |
| 0.4245 | 1.22 | 400 | 0.3921 |
| 0.4225 | 1.53 | 500 | 0.3892 |
| 0.4207 | 1.83 | 600 | 0.3858 |
| 0.4151 | 2.14 | 700 | 0.3820 |
| 0.4136 | 2.44 | 800 | 0.3803 |
| 0.4105 | 2.75 | 900 | 0.3782 |
| 0.4083 | 3.05 | 1000 | 0.3763 |
| 0.4046 | 3.36 | 1100 | 0.3764 |
| 0.4012 | 3.66 | 1200 | 0.3748 |
| 0.4004 | 3.97 | 1300 | 0.3733 |
| 0.3998 | 4.27 | 1400 | 0.3726 |
| 0.4013 | 4.58 | 1500 | 0.3717 |
Framework versions
- Transformers 4.38.0
- Pytorch 2.6.0+cu124
- Datasets 3.5.1
- Tokenizers 0.15.2