deploy: fresh upload via HF API — force rebuild with latest code

feat: curriculum training, 4 reward funcs, session API, preflight tests, compare_inference

Create colab_notebook_guide.py

Create preflight_test.py

updated readme

added inference comparisor

fix: defensive .get() in grader to handle GRPO parallel rollout state

Fix .gitignore and add env module

feat(rl): Add Unsloth + TRL GRPO training pipeline scaffolding

update

update inference

readme update

Prepare OpenEnv benchmark for submission