fix: Use correct Gradio named endpoint for transcription

bump: Update semantic version to 1.2.6

fix: Revert to working Gradio queue system, remove FastAPI complexity

bump: Update semantic version to 1.2.5

fix: Remove duplicate tab_names parameter in TabbedInterface

bump: Update semantic version to 1.2.4

fix: Update client to use Gradio API endpoint format

fix: Move FastAPI imports to top of file for proper initialization

feat: Add direct API endpoints to bypass Gradio queue system

enhance: Add comprehensive base64 debugging for troubleshooting Invalid base64 data errors

feat: Add debugpy support for Cursor/VSCode Dev Mode debugging

revert: Downgrade STT service from Gradio 5.44.0 to 4.44.1

fix: Add base64 padding validation for Gradio 5.44.0 audio data

fix: Set OMP_NUM_THREADS=1 to resolve libgomp threading warning

fix: Return single string from gradio_transcribe_memory for voiceCal.ai compatibility

fix: Add missing output components to match 3-return function values

fix: Return 3 output values to match Gradio interface components

fix: Standardize Gradio outputs as arrays for consistent API behavior

fix: Improve STT service for pure Gradio HuggingFace Spaces compatibility

Resolve README merge conflict with HuggingFace metadata

initial commit

feat: STT GPU Service - WebRTC Speech-to-Text Microservice