r18-a2-q2-chrono4_2012-r2a-s400
anacoluthe89/chrono4-2012 + quality RFT round 1: SFT on its own answers that beat king-2025 (both judge orders, gemini-2.5-flash-lite) w-sft 2.0, lr 1e-05 last:8, 2011 known protection (LLM-known + 48k clozes), <=2011 text KD 1.0 / CE 0.5, no forget, step 400