这一路破空,苦难有千重谁人懂
| Repository | Role | Purpose |
|---|---|---|
mit-han-lab/once-for-all |
SPECIALIZE | [ICLR 2020] Once for All trains one network and specializes it for efficient deployment. |
mit-han-lab/apq |
CO-DESIGN | [CVPR 2020] Joint search for network architecture, pruning, and quantization policy. |
binghong-ml/MolEvol |
EVOLVE | [ICLR 2021] Molecule optimization by explainable evolution. |
ACM-Class-2016/probability-theory-lecture-notes |
FOUNDATION | Lecture notes of Probability Theory. |
Efficient AI · 2 modules
| Module | Purpose |
|---|---|
mit-han-lab/once-for-all |
Train once, then specialize for efficient deployment. |
mit-han-lab/apq |
Jointly search architecture, pruning, and quantization policy. |
Research foundations · 2 modules
| Module | Purpose |
|---|---|
binghong-ml/MolEvol |
Explainable evolutionary molecule optimization. |
ACM-Class-2016/probability-theory-lecture-notes |
Lecture notes of Probability Theory. |

