Files
evil_MoE/src/projected_grpo
wassname afcd4a1e88 refactor(c): extract data.py (load_problems) + vhack.py (load/postprocess v_hack)
data.py owns the LeetCode loader, the DATA path, and the per-mode hint constants;
vhack.py owns v_hack load + the k-slice/noise-floor postprocess. Both are pure
(no train globals beyond what moved with them). train.py imports them back.
Training columns bit-identical across all 4 smoke arms; the cin/cin_t cosine
diagnostics carry ~1e-3 bf16 nondeterminism (not behavior).

Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
2026-06-01 09:29:03 +00:00
..
2026-05-23 10:40:02 +08:00
wip
2026-05-29 06:29:46 +00:00