yuyunfrancis/kin_to_en_api

A robust FastAPI-based API service for transcribing, translating, and generating captions for audio files, with a focus on supporting Kinyarwanda and English languages, featuring comprehensive fallback mechanisms for ensuring reliable translation services. # Audio Transcription & Translation API A robust FastAPI-based API service for transcribing, translating, and generating captions for audio files, with a focus on supporting Kinyarwanda and English languages, featuring comprehensive fallback mechanisms for ensuring reliable translation services. ## Features - **Audio Transcription**: Convert speech to text in Kinyarwanda or English - **Translation**: Translate text from Kinyarwanda to English with multi-model fallback - **Caption Generation**: Create SRT or VTT caption files with bilingual support - **Combined Pipeline**: Process audio through a full transcription, translation, and captioning pipeline - **Fault Tolerance**: Robust fallback mechanisms when primary models are unavailable ## API Endpoints ### Transcription Endpoints - `POST /api/transcribe/`: Transcribe an audio file - `GET /api/transcribe/languages`: List supported transcription languages ### Translation Endpoints - `POST /api/translate/text`: Translate text from source to target language - `POST /api/translate/chunks`: Translate transcription chunks with timing information - `POST /api/translate/audio`: Transcribe and translate audio in one step - `GET /api/translate/languages`: List supported translation language pairs ### Caption Endpoints - `POST /api/captions/generate`: Generate captions from transcription data - `POST /api/captions/from-audio`: Generate captions directly from audio file - `GET /api/captions/formats`: List supported caption formats ### Pipeline Endpoint - `POST /api/pipeline/process`: Complete audio processing pipeline (transcription, translation, captioning) ## Installation and Setup ### Prerequisites - Python 3.9 or higher - FFmpeg installed on your system - (Optional) CUDA-compatible GPU for faster processing ### Local Development Setup 1. Clone the repository: ```bash git clone github.com cd audio-transcription-api ``` 2. Create a virtual environment: ```bas …