Commit History
-
complete almost
04b615f -
pin vllm<=0.18.0 to fix TRL importance-sampling ratio collapse
8464e70 -
Add GRPO training scaffolding (Module 5 rollout_func pattern) + plot artifacts
caea08d -
shop manage eng phase 2 first
048f186
-
feat: jewelry shop RL environment with market, warehouse, and showroom phases
5c6ca01