tech-daskalos/CyberSecChatbot

Update Dockerfile with main.py copy

Copy main.py to Docker image to pick up syntax fixes

Fix remaining f-string syntax errors

Fix f-string syntax error in streaming endpoint

Switch to GPU-enabled Docker image

Add thread-safety for concurrent users

Configure uvicorn for concurrent request handling

Add concurrent request handling with model pool

Add GPU support

Revert to simpler configuration - optimizations caused slowdown

Optimize for faster inference

Increase threads to 8 for faster inference

Disable RAG for faster inference

Optimize for 8vCPU/32GB instance

Optimize model parameters for faster CPU inference

Fix API endpoint to use relative URL instead of localhost

Add test interface HTML to Docker image

Add knowledge_db directory and remove deprecated env var

Fix Hugging Face cache permissions

Fix /app/models directory permissions

Fix /data directory permissions for HF Spaces

Use pre-built Docker image from Docker Hub

Use pre-built Docker image from Docker Hub

Fix: Install llama-cpp-python at startup to /tmp to avoid build timeout

Fix: Install cmake and build-essential for llama-cpp-python build

Fix: Force use of pre-built llama-cpp-python wheels

Fix: Install llama-cpp-python at runtime (HF Spaces workaround)

Optimize: Use pre-built llama-cpp-python wheels for faster builds

Initial deployment to Spaces

initial commit