ESPnet Text to Byte Array

This project leverages ESPnet2 for converting text to a byte array and subsequently reconstructing the audio from the byte data.

Setup

Follow the steps below to set up and run the project:

Step 1: Clone the Repository

git clone https://github.com/samiabat/ESPnet-text-2-byte-audio-buffer.git

Step 2: Install Dependencies

pip install -r requirements.txt

Step 3: Organize Model Files

Create a folder named 'model' in the project root directory. Place the 'config.yaml' and 'train.total_count.ave_10best.pth' files inside the 'model' folder.

Step 4: Prepare Text File

Create a file named 'text.txt' in the project root directory and add the text you want to synthesize.

Step 5: Run the Code

python3 ESPnetT2S.py

Additional Notes

• The ESPnetTextToByte class in ESPnetT2S.py handles the text-to-speech conversion and byte array creation.
• The get_byte_data method in the class writes the byte data to a file named 'audio_byte_file.raw'.
• ⚠️ The byte data is in float32 format so when load the buffur file it should be in float32.
• Adjust the file paths as needed, and feel free to customize the code to suit your requirements.