Update app.py

Update app.py

Update ZeroGPU Limit To 180 Seconds

Update README.md

Update requirements.txt

Tune pyannote to segmentation 64 embedding 32

Set pyannote batch sizes directly on pipeline

Increase pyannote batch sizes to 64

Cache model locally and run diarization from memory

Move diarization post-processing outside GPU function

Return minimal diarization JSON schema

Simplify API-focused diarization app

Simplify API-only diarization inputs

Return diarization segments as JSON

Report measured ZeroGPU inference time

Normalize audio before diarization

Remove speaker count options from app

Update app for Gradio 6 API changes

Pin compatible audio runtime dependencies

Add pyannote diarization Gradio app

initial commit