Skip to content
#

pgsr

Here is 1 public repository matching this topic...

Drop-in, idempotent speed patches for PGSR surface-reconstruction training — removes per-iteration PCIe transfers, redundant ones-kernel convolutions, and GPU-CPU logging syncs (plus a real EMA-logging bug) with CPU-proven numerical parity: same geometry, lower GPU bill.

  • Updated Jul 26, 2026
  • Python

Add this topic to your repo

To associate your repository with the pgsr topic, visit your repo's landing page and select "manage topics."

Learn more