Spaces:
Paused
Paused
Commit History
-
fix: Improve fallback handling and prevent TTS errors v0.5.4
Peter Michael Gits +1: Claude- Shortened WebSocket timeout to 5s for faster fallback - Fixed fallback message to be TTS-friendly instead of technical details - Ensured HTTP fallback is properly attempted before final fallback - Better error messages to distinguish WebSocket vs HTTP failures - Prevents "STT fallback: Large audio file" from being sent to TTS 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com>08815de -
feat: Force WebSocket as primary method for STT/TTS integration v0.5.3
Peter Michael Gits +1: Claude- Reverted to WebSocket-FIRST strategy as requested - WebSocket connections are now the PRIMARY method for both STT and TTS - HTTP API serves only as fallback when WebSocket fails - Added detailed debug logging to identify WebSocket connection issues - Better error messages to distinguish timeout vs endpoint availability - This prioritizes the WebRTC/WebSocket architecture we built 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com>5be66b9 -
fix: Prioritize HTTP API over WebSocket for STT/TTS services v0.5.2
Peter Michael Gits +1: Claude- Changed strategy to use HTTP API as primary method instead of WebSocket - Added comprehensive HTTP fallback for both STT and TTS services - WebSocket remains as fallback when HTTP fails - Better error handling and timeout management for TTS (60s timeout) - Should resolve STT fallback and TTS generation failed errors - Uses existing Gradio API endpoints that are confirmed working 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com>f48b31d -
fix: Add HTTP API fallback for STT when WebSocket unavailable v0.5.1
Peter Michael Gits +1: Claude- Added timeout handling for WebSocket connections to STT/TTS services - Implemented HTTP API fallback when WebSocket endpoints not available - Better error handling and connection timeout management - Graceful degradation to Gradio API format for existing services - Prevents "STT fallback" messages when real API is available Fixes WebSocket connection issues when external services lack WS endpoints 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com>2e4a0b8 -
feat: Complete TTS playback integration for voice responses v0.4.9
Peter Michael Gits +1: Claude- Added WebSocket TTS handler to TTS service with ZeroGPU synthesis - Integrated TTS WebSocket client in ChatCal WebRTC handler - Real-time text-to-speech with base64 audio transmission - Auto-generate demo TTS responses after STT transcription - Client-side audio playback with proper error handling - Complete voice interaction loop: Speech → Text → Response → Audio 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com>228bc17 -
feat: Add real STT WebSocket integration to replace demo responses v0.4.8
Peter Michael Gits +1: Claude- Added WebSocket STT handler to STT service with ZeroGPU transcription - Integrated real STT WebSocket client in ChatCal WebRTC handler - Support for dual mode: Gradio interface + WebSocket STT endpoints - Real-time audio transmission via base64 encoding over WebSocket - Automatic cleanup of STT connections on client disconnect - Fallback to demo responses if STT service unavailable 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com>ccbd055 -
fix: Audio buffer size error - handle MediaRecorder format v0.4.6
Peter Michael Gits +1: Claude- Fix buffer size error by properly handling MediaRecorder audio format - Replace numpy.frombuffer with file-based audio processing - MediaRecorder produces WebM/OGG format, not raw PCM data - Add process_audio_file_webrtc() method for proper file handling - Implement file size analysis for WebRTC demo functionality - Clean up duplicate transcription sending code - Better error handling and temporary file cleanup Audio Processing: - Save received audio data as temporary WebM file - Analyze file size instead of attempting raw buffer conversion - Proper cleanup of temporary files after processing 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com>aab69ad -
feat: WebRTC-first implementation with demo audio processing v0.4.4
Peter Michael Gits +1: Claude- Comment out failing HTTP STT calls (preserve for future) - Enable WebRTC FastAPI app as primary deployment - Implement WebRTC audio processing with characteristics detection - Remove WSGI middleware complexity - direct FastAPI deployment - Create functional WebSocket endpoints for real-time audio - Add audio duration and sample rate analysis in WebRTC demo - Fallback to Gradio-only if WebRTC fails WebRTC Endpoints Available: - /ws/webrtc/{client_id} - WebSocket for audio streaming - /webrtc/demo - Interactive demo with microphone - /webrtc/test - API status check Note: HTTP STT services preserved in comments for later integration 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com>65f90da -
fix: Add missing WebRTC module files for HF deployment v0.4.1
Peter Michael Gits +1: Claude- Create complete WebRTC directory structure with __init__.py files - Add websocket_handler.py for real-time audio streaming - Add audio_processor.py for STT service integration - Ensure all imports resolve correctly for HF Space deployment - Maintain no-demo-mode policy across all WebRTC components 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com>edd83ad