Hugging Face's logo

Arbi-Houssem
/
speecht5_ar_tn_1.6

Text-to-Audio
Transformers
TensorBoard
Safetensors
Arabic
speecht5
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
speecht5_ar_tn_1.6 / README.md
Arbi-Houssem's picture
Arbi-Houssem
End of training
85b6753 verified
preview code
|
raw
history blame contribute delete
No virus
1.88 kB
metadata
language:
  - ara
license: mit
base_model: MBZUAI/speecht5_tts_clartts_ar
tags:
  - generated_from_trainer
datasets:
  - Arbi-Houssem/Tunisian_dataset_STT-TTS15s_filtred1.0_Mixed
model-index:
  - name: SpeechT5 TTS Tunisien
    results: []

SpeechT5 TTS Tunisien

This model is a fine-tuned version of MBZUAI/speecht5_tts_clartts_ar on the Tunisian_dataset_STT-TTS15s_filtred1.0_Mixed dataset. It achieves the following results on the evaluation set:

  • Loss: 0.5743

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: 8
  • 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
  • training_steps: 4000
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss
0.57 6.67 500 0.6392
0.5333 13.33 1000 0.6088
0.5289 20.0 1500 0.5913
0.5002 26.67 2000 0.5848
0.5045 33.33 2500 0.5783
0.5325 40.0 3000 0.5803
0.5215 46.67 3500 0.5742
0.4773 53.33 4000 0.5743

Framework versions

  • Transformers 4.35.1
  • Pytorch 2.3.1+cu121
  • Datasets 2.14.7
  • Tokenizers 0.14.1