diff options
| author | Yuren Hao <yurenh2@illinois.edu> | 2026-08-05 19:14:33 -0500 |
|---|---|---|
| committer | Yuren Hao <yurenh2@illinois.edu> | 2026-08-05 19:14:33 -0500 |
| commit | 2de7789ecf05eba74cc52dbd5f4d8298e4084e50 (patch) | |
| tree | 5546f171017966ea7294c4c09a67e4a09778a891 /ep_run/watch_contraction.py | |
| parent | 0df61c1d36faf3e985394c943853b5bab5adaa8c (diff) | |
PSGD-QUAD入列: vendor psgd.py + Optimizer子类适配器(合成损失closure桥接EP预填梯度), 13路全量回归冒烟PASS
修复过程记录: 首版适配器非Optimizer子类被LambdaLR拒; 重写时splice错序造成muon.py重复段
(旧类后定义胜出), 按结构图手术去重并断言全符号在位。三轴定位: GPU配方候选(追踪慢变统计=
R4幸存哲学+噪声鲁棒卖点), 硬件不合适(d²记忆), 偏置公理最差档(逐方向白化放大低方差偏置)。
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014FAPDWQ49M5Ye3NpTndTpn
Diffstat (limited to 'ep_run/watch_contraction.py')
0 files changed, 0 insertions, 0 deletions