README.md

metadata

title: TranStudio
colorFrom: indigo
colorTo: blue
sdk: docker
sdk_version: '3.10'
app_file: main.py
pinned: false
fullWidth: true
header: mini

TranStudio

TranStudio is a web-based application for generating and managing audio transcriptions. It provides an intuitive interface for uploading audio files, generating transcriptions, and managing transcribed content with advanced playback controls.

Features

• Audio Upload: Support for uploading audio files for transcription
• Real-time Transcription: Process audio files and generate text transcriptions
• Interactive Audio Player:
  • Waveform visualization
  • Play/pause controls
  • Volume control with mute option
  • Segment-based playback
  • Time formatting and seeking
• Transcription Management:
  • Save and load transcriptions
  • Copy text to clipboard
  • Delete segments
  • View transcription history
• Advanced Options:
  • Language selection
  • Response format customization
  • Temperature control
  • Chunk size and overlap settings

Technology Stack

• Backend: FastAPI (Python)
• Frontend: Vue.js with Buefy UI components
• Audio Processing:
  • Howler.js for audio playback
  • SoundFile for audio processing
  • Custom waveform visualization

Hugging Face Spaces Deployment

This application is ready to be deployed on Hugging Face Spaces:

1. Create a new Space on Hugging Face with Docker template
2. Link this repository to your Space
3. Configure the following environment variables in your Space settings:
   • SECRET_KEY: A secure random string for JWT encryption
   • DATABASE_URL: Database connection string (default: "sqlite:///./transtudio.db")
   • OPENAI_API_KEY: Your OpenAI API key
   • OPENAI_BASE_URL: OpenAI API base URL (default: "https://api.openai.com/v1")
   • CORS_ORIGINS: Comma-separated list of allowed origins

Local Setup

1. Clone the repository
2. Install dependencies:

pip install -r requirements.txt