Kakeya OProver — Lean Proof Training Journey Lean proof LoRA journey: SFT, DPO regression, compiler-feedback repair, strict evals, and Jensen SEARCH_EXHAUSTED (no RH proof). FluffyAIcode/Kakeya-OProver-Stage1-SFT90 Text Generation • Updated Aug 8 • 8 FluffyAIcode/Kakeya-OProver-Stage2-SFT1000 Text Generation • Updated Aug 8 • 7 FluffyAIcode/Kakeya-OProver-Stage3-DPO Text Generation • Updated Aug 8 • 5 FluffyAIcode/Kakeya-OProver-Stage4-Repair-SFT Text Generation • Updated Aug 8 • 6
Kakeya OProver — Lean Proof Training Journey Lean proof LoRA journey: SFT, DPO regression, compiler-feedback repair, strict evals, and Jensen SEARCH_EXHAUSTED (no RH proof). FluffyAIcode/Kakeya-OProver-Stage1-SFT90 Text Generation • Updated Aug 8 • 8 FluffyAIcode/Kakeya-OProver-Stage2-SFT1000 Text Generation • Updated Aug 8 • 7 FluffyAIcode/Kakeya-OProver-Stage3-DPO Text Generation • Updated Aug 8 • 5 FluffyAIcode/Kakeya-OProver-Stage4-Repair-SFT Text Generation • Updated Aug 8 • 6
