Search results

Distributed SHT fails when channel count is smaller than polar group size

Status: Open.

#207 In NVIDIA/torch-harmonics;

· bkravitz opened on May 29, 2026

add FP64 support for attention kernels

Status: Open.

#204 In NVIDIA/torch-harmonics;

· azrael417 opened on May 26, 2026

Implement AT_DISPATCH for CUDA attention kernel dispatch

Status: Open.

#203 In NVIDIA/torch-harmonics;

· azrael417 opened on May 22, 2026

Ensuring autocast eligibility for all custom ops

Status: Open.

#202 In NVIDIA/torch-harmonics;

· azrael417 opened on May 22, 2026

need single source of truth for default theta_cutoff

Status: Open.

#184 In NVIDIA/torch-harmonics;

· azrael417 opened on May 8, 2026

Question about Spherical Harmonic Function

Status: Open.

#143 In NVIDIA/torch-harmonics;

· Shikamaru-Nara1729 opened on Dec 10, 2025