chore: untrack server/__pycache__ files

fix: use correct HF router URL /hf-inference/v1/ (not /models/{id}/v1/)

fix: COGS profit model - charge unit cost on sold units not on orders, preventing end-of-period bias against high-inventory strategies

fix: update HF proxy URL to router.huggingface.co (api-inference deprecated 410)

fix: serve index.html with no-cache headers to prevent CDN/browser stale cache

fix: proxy HF Inference API through FastAPI to bypass HF Spaces CSP

metrics: profit first, then service level, then fill rate (React UI)

fix: use minimal requirements-server.txt in Docker to avoid downloading torch/ML deps

fix: add package-lock.json for npm ci in Docker build

fix: switch HF Space SDK from gradio to docker

feat: replace Gradio with React UI — GRPO tab, 730-day sim, 200-entry memory bank

refactor: remove Unsloth, use standard transformers + PEFT

fix: use nvidia/cuda devel base image so vllm can build with CUDA toolkit

fix: pin torch 2.6.0 + xformers 0.0.29, use Python 3.12 for ML compat

fix: install torch, xformers, vllm before requirements to avoid build failures

fix: install torch before vllm/xformers to satisfy build deps

feat: add vllm for Unsloth fast_inference

fix: force-reinstall unsloth to fix PreTrainedConfig NameError

fix: align Unsloth config with recommended GRPO settings

Remove ideas

metrics: profit first, then service level, then fill rate

Merge branch 'main' of https://github.com/ademcodesproducts/OpenEnv-Inventory-Simulations

fix: pipeline-aware ordering, YoY demand signal, reward rebalancing

feat: integrate Unsloth into GRPO training pipeline

feat: full-horizon lookahead reward (365 days, <0.5ms)

feat: crash-resilient training with dataset caching and iteration resume

feat: tee training logs to persistent volume

Remove PPO MLP agent, update README

feat: add Northflank training Dockerfile and start.sh

feat: improve GRPO training logging and fix torch_dtype deprecation

fix: let Gradio auto-select port locally, fix via env vars only

fixed all environment files

fixed all environment files

fix: use 127.0.0.1 locally, 0.0.0.0 in Docker via env var

fix: Dockerfile runs Gradio app, not FastAPI server

Merge branch 'main' of https://huggingface.co/spaces/ademarteau/RL-Inventory-Simulations

fix: fall back to HF_TOKEN env var for Inference API auth

Upload ppo_inventory.zip with huggingface_hub

Merge: accept remote ppo model and gitattributes

Upload ppo_inventory.zip with huggingface_hub

Track zip files with LFS

Merge teammate changes, unify reward via reward.py, add PPO model

Added PPO model and reward.py

feat: improve training logging with tqdm, timings, GPU memory, ETA

Merge branch 'main' of https://github.com/ademcodesproducts/OpenEnv-Inventory-Simulations

Added trained PPO model + app.py UI changes for HF Spaces

fix: add missing ML and simulation packages to requirements.txt

fix: remove pywin32 Windows-only packages, use Python 3.13

fix: bump to Python 3.13 to match requirements.txt (audioop-lts)

fix: remove scripts/ COPY, scripts dir not in repo