README.md

metadata

language:
  - en
license: mit
base_model: microsoft/speecht5_tts
tags:
  - scottish
  - tts
  - glaswegian
  - generated_from_trainer
datasets:
  - divakaivan/glaswegian_audio
model-index:
  - name: GlaswegianTTS v0.1.0
    results: []

GlaswegianTTS v0.1.0

This model is a fine-tuned version of microsoft/speecht5_tts on the glaswegian_tts_v0.1.0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.7169

Training and evaluation data

Glaswegian audio clips 80% train/test split

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 1e-05
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• 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: 500
• training_steps: 4000
• mixed_precision_training: Native AMP

Framework versions

• Transformers 4.42.0.dev0
• Pytorch 2.2.1+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1