train(grpo): unified hint prompt, no-history chat, positive-advantage filter

fix: align notebook with 15-day horizon, drop unused replies field

update

Set TASK_HORIZON to 15 days and align graders, UI, and training prompts.

update

fix(env): tolerate malformed predict_engagement scheduled_actions

train: shrink to weekly horizon + bounded steps

train: per-step credit + drop replies + larger batches

update

la la la --123

firstiteration

reduced steps to fit out free tier

Viraltest OpenEnv: deploy to HF Space