Whisper Small Cantonese Yue English

🚀 Transformers Library

This project utilizes the transformers library to achieve specific speech - related tasks. It leverages multiple datasets and is based on a pre - trained model, with clear training parameters and hardware requirements.

🚀 Quick Start

Installation

If you encounter tokenizer issues during inference, please update your transformers version to >= 4.46.3:

pip install --upgrade transformers==4.46.3

✨ Features

• Datasets: Utilizes multiple high - quality speech datasets, including AlienKevin/mixed_cantonese_and_english_speech, mozilla - foundation/common_voice_17_0, and mozilla - foundation/common_voice_11_0.
• Metrics: Uses the CER (Character Error Rate) metric to evaluate model performance.
• Base Model: Built upon the openai/whisper - small pre - trained model.

📦 Installation

The project relies on the transformers library. You can install the required version as shown in the Quick Start section.

📚 Documentation

Model Information

Training Arguments

• per_device_train_batch_size = 32
• gradient_accumulation_steps = 1
• learning_rate = 1e - 5
• gradient_checkpointing = True
• per_device_eval_batch_size = 64
• generation_max_length = 225

Hardware Requirements

• NVIDIA Tesla V100 16GB * 4

💡 Usage Tip

💡 Usage Tip

If having tokenizer issue during inference, please update your transformers version to >= 4.46.3

📄 License

This project is licensed under the MIT license.