Two-decimal scores, single tasks/graders.py, min three inference episodes 97032c2 Maheswar01 commited on Apr 12
feat: six canonical tasks, task_registry, DaddyCoder-style graders layout 34a3f52 Maheswar01 commited on Apr 10
feat: canonical tasks/graders.py layout for Phase 2; wire inference and manifests ea40fd4 Maheswar01 commited on Apr 10
fix: grader scores strictly in (0,1); declare tasks_with_graders; README Phase 2 notes 0e74385 Maheswar01 commited on Apr 8
fix: default to LLM when API_KEY/HF_TOKEN set; proxy API_KEY first d5c1dd5 Maheswar01 commited on Apr 8
fix: use str.format for prompts (Py3.10/3.11 Phase2 SyntaxError) cd7211a Maheswar01 commited on Apr 8