rishithayanidhi/datacenter-cooling-optimization

fix: simplify __main__ block to call main() directly for validator compatibility

fix: add server/ and uv.lock to root, update pyproject.toml for multi-mode deployment

fix: add pyproject.toml to repo root for multi-mode deployment validator

fix: add openenv.yaml to repo root for validator

perf: initialize LLM client once in main() instead of per episode

fix: set default API_BASE_URL to OpenAI, MODEL_NAME to gpt-4o

fix: update default API_BASE_URL to router.huggingface.co, MODEL_NAME to Qwen2.5-72B

fix: enforce action format via system prompt, sanitize newlines in [STEP] output

fix: add LOCAL_IMAGE_NAME, top-level OpenAI import, score in [END] docstring

fix: set API_BASE_URL/MODEL_NAME defaults to OpenAI, score in [END], HF_TOKEN as api_key

fix: inject HF token into WebSocket handshake for private Space auth

fix: replace BaseHTTPMiddleware with ASGI middleware to unblock WebSocket /ws route

fix: add root GET / endpoint to resolve 404 from HF Spaces log polling

refactor: make all config dynamic via env vars, add structured file logging and try/catch error handling

Fix: Use openenv-core from GitHub instead of PyPI openenv

Add missing openenv and torch dependencies to Docker requirements

Increase startup timeout to 1 hour

Fix colorTo to valid value

Add UTF-8 README

Remove UTF-16 file

Fix HF Space config with app_port

Remove UTF-16 README

Merge HF remote

initial commit

project completed

Update verified deployment configuration

Fix Dockerfile paths and remove redundant server Dockerfile

Add complete project files: inference, models, server, validation scripts

first commit