Upload post_training_eval.json

Upload training_log_200.json

Replace old reward curve with 200-step version

Upload 200-step graph: training_dashboard.png

Upload 200-step graph: grpo_dual_reward_curve.png

Upload 200-step graph: base_vs_trained.png

Upload 200-step graph: grpo_reward_curve_200.png

feat: full project files — server, training, evaluation, models, outputs