Upload README.md with huggingface_hub

Upload app.py with huggingface_hub

Upload requirements.txt with huggingface_hub

Upload requirements.txt with huggingface_hub

Upload requirements.txt with huggingface_hub

Upload requirements.txt with huggingface_hub

Upload requirements.txt with huggingface_hub

Upload README.md with huggingface_hub

Upload app.py with huggingface_hub

Upload requirements.txt with huggingface_hub

Remove gradio_client - let Gradio 4.0.0 auto-install it

Fix import: use gradio_client instead of gradio.networking

Add gradio_client==0.8.0 for Gradio 4.0.0 compatibility

Update for Gradio 4.0.0 compatibility (Free CPU)

Update sdk_version to 4.0.0

Downgrade to Gradio 4.0.0 for Free CPU compatibility

v12-V2-SPACES: Use Bandila-v2 and Amigo-v2 (hash: 2039db4e)

v11: HTTP workaround for old Gradio (hash: 20cc3792)

v10-FINAL: Gradio 4.44.1 compatible (hash: aa3fb3fb)

FIX: Update sdk_version to 4.44.1 in README

v9: Mark as GRADIO-UPDATE

v9: Upgrade Gradio to 4.44.1

v8-SERIALIZE-FIX: Skip API info parsing (hash: b4181bb6)

v7-RETRY: Fallback without token + detailed debug (hash: 28713834)

v6-WAKEUP: Use env HF_TOKEN (hash: 83ddcb93)

v5-DETAILED: Pinpoint exact failure point (hash: 88cea635)

v4-DEBUG: Enhanced logging (hash: 36bc7e29)

Force rebuild - update requirements

Add debug logging to see actual specialist responses

Force rebuild v2 - add version marker

Fix response extraction - handle tuple format from Gradio chatbot

Fix API endpoint: use /respond instead of /predict

REAL FIX: Use eager attention to prevent DynamicCache creation

Pin transformers==4.44.2 to fix DynamicCache

Fix DynamicCache error - disable cache + pin transformers

Deploy GPU-optimized Amihan 1.2 with trained model

Switch to lightweight mode - no model loading, instant startup, keyword routing

Optimize for CPU: use float32, force CPU device, add loading progress

Fix chatbox: correct input/output order and add error handling

Fix API endpoint calls to Bandila and Amigo Spaces

Add standalone fallback mode when specialists are sleeping

Replace mythology reference with technical description

Add link to detailed installation guide

Add detailed local installation guide for beginners

Fix JSON parsing error with better error handling

Add thumbnail.png for social media previews

Upload app.py with huggingface_hub

Upload README.md with huggingface_hub

Upload app.py with huggingface_hub

Upload README.md with huggingface_hub