mschoi/ai-rfm
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Improve RMSD with weighted flow objective.

Browse Source 
Increase rotational and torsional emphasis in normalized velocity loss with gradient clipping, yielding a better 100-run final-time RMSD while preserving random-time flow-matching training.

Made-with: Cursor
This commit is contained in:
demian3b
2026-04-16 17:29:43 +09:00
parent e7274cb680
commit 3ddae9d815
4 changed files with 23 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
README.md
View File

@@ -49,3 +49,4 @@ This repository is intentionally pinned to CUDA 12.6 PyTorch wheels and matching
- 2026-04-16: Added pre-commit artifact refresh: on best update it now stages `BEST_PRACTICE.json`, `artifacts/best_model.pt`, and regenerates 6 trajectory visualizations in `reports/trajectories/`.
- 2026-04-16: Enforced random-time flow-matching rule (no fixed training time), saved best checkpoint to git-tracked artifact path, and improved metric to `mean_rmsd_100=2.519821` with `gcn hidden=512 layers=8 batch=96`.
- 2026-04-16: Added a general multi-layer diagnosis principle to `GUIDELINES.md` so experiments are judged with quantitative + qualitative + structural evidence, not metric-only optimization.
- 2026-04-16: Tried weighted objective to counter weak rotation/torsion motion (`w_center=0.8, w_omega=2.0, w_torsion=3.0, grad_clip=0.8`) and improved to `mean_rmsd_100=2.505556`.