Perf: skip regex on every token unless think tag present

Perf: short system prompt 530->50 tokens, n_threads=2, ctx=512, max_tokens=80

Fix Gradio 6: remove type param from Chatbot (removed in v6, always messages format)

Fix Gradio 6: remove show_copy_button from Chatbot

Fix Gradio 6: move theme/css to launch(), remove invalid Chatbot kwargs

Upload README.md with huggingface_hub

Upload app.py with huggingface_hub

Upload Dockerfile with huggingface_hub

Upload requirements-hf.txt with huggingface_hub

Upload config.py with huggingface_hub

Upload app.py with huggingface_hub

Enable true real-time streaming by default and disable chat DB logging latency

Ultra-fast preset: shorter context/output, reduced history, lower thinking overhead

Upload config.py with huggingface_hub

Speed tuning: smooth chunked output mode + faster CPU defaults

Upload app.py with huggingface_hub

Improve streaming smoothness and reduce latency on low hardware

Upload app.py with huggingface_hub

Fix Gradio chat history format compatibility (tuple vs messages)

Fix chat streaming duplication, robust message normalization, and cleaner response style

Upload app.py with huggingface_hub

Complete backend upload

Fix: Add error handling for vision model loading

Add local models: faster-whisper (STT) + BLIP (vision) - all features now 100% local except image generation

Upload app.py with huggingface_hub

Fix all features: Update API URLs to router.huggingface.co, fix chat_fn signature

Fix: Use tuple format for Gradio 4.19.2 compatibility

Fix: Downgrade to Gradio 4.19.2 to avoid API schema bug

Fix Gradio API schema bug - manual chat interface

Compile from source for Debian (glibc compatibility)

Add build-essential + pin llama-cpp v0.2.90

Fix llama-cpp install order

Switch to Docker SDK

Remove gradio (auto-added by HF Spaces)

Fix version compatibility: pin huggingface_hub==0.20.3 gradio==4.19.2

Remove llama-cpp from requirements

Add Dockerfile for prebuilt wheel

Use Python 3.11 + prebuilt wheel

Trigger rebuild

Upload requirements.txt with huggingface_hub

Upload database.py with huggingface_hub

Upload app.py with huggingface_hub

Upload README.md with huggingface_hub

Upload config.py with huggingface_hub

Upload app.py with huggingface_hub

Upload inference.py with huggingface_hub

Upload conversation.py with huggingface_hub

Upload config.py with huggingface_hub

Upload app.py with huggingface_hub

Upload requirements.txt with huggingface_hub