kasimisik01/xtts-turkish-tts

Fix: Generate silent speaker reference locally

Fix: Download and use default speaker reference file

Debug: Add error logging and handle missing speaker_wav

Fix: Remove speakers attribute, use optional speaker_wav

Fix: Use speaker parameter instead of speaker_wav

Fix: Download and use default speaker reference

Fix: Support both POST (JSON body) and GET (query params)

Fix: Use Query params instead of JSON body

Fix: Pin PyTorch to 2.5.1 for XTTS compatibility

Fix: Pin transformers to 4.39.3 for XTTS compatibility

Fix: Remove start.sh reference, use python3 app.py

Fix: Patch input at very top of app.py

Fix: Use start.sh to patch input() before TTS import

Fix: Patch input() to auto-agree license

Fix: Add COQUI_TOS_AGREED env var for license

Add XTTS v2 API server

initial commit