All issues

Search results

Allow Clean and Compacting project from context menu and inspector

Project ManagerAsset manager / library / provenance workstreamAsset manager / library / provenance workstream

Status: Open.

#2512 In MrNeRF/LichtFeld-Studio;

· kmeirlaen opened on Sep 26, 2026

CUDA Tensor::multinomial with replacement scans every weight per draw

Status: Open.

#2497 In MrNeRF/LichtFeld-Studio;

· MrNeRF opened on Sep 26, 2026

CUDA has no dedicated argmax/argmin (indexed extrema) kernel

Status: Open.

#2493 In MrNeRF/LichtFeld-Studio;

· MrNeRF opened on Sep 26, 2026

copy_slice, all_close, has_nan/has_inf (non-Float32) and functional::* compute GPU tensors on the host

Status: Open.

#2480 In MrNeRF/LichtFeld-Studio;

· MrNeRF opened on Sep 26, 2026

GPU backends reject scatter_ modes, append_gather dtypes and non-contiguous Bool reductions that CPU accepts

Status: Open.

#2479 In MrNeRF/LichtFeld-Studio;

· MrNeRF opened on Sep 26, 2026

pad supports only Float32 on every device, and keeps a dead warn-and-return-zeros branch

Status: Open.

#2478 In MrNeRF/LichtFeld-Studio;

· MrNeRF opened on Sep 26, 2026

flip is CPU-only

Status: Open.

#2477 In MrNeRF/LichtFeld-Studio;

· MrNeRF opened on Sep 26, 2026

Index and weight validation downloads whole GPU tensors

Status: Open.

#2476 In MrNeRF/LichtFeld-Studio;

· MrNeRF opened on Sep 26, 2026

nonzero() on rank >= 2 GPU tensors runs on the host

Status: Open.

#2475 In MrNeRF/LichtFeld-Studio;

· MrNeRF opened on Sep 26, 2026

Tensor::to(Bool) round-trips through the host for Int32, Int64 and Float16

Status: Open.

#2474 In MrNeRF/LichtFeld-Studio;

· MrNeRF opened on Sep 26, 2026

index_put_ scatters GPU tensors on the host

Status: Open.

#2473 In MrNeRF/LichtFeld-Studio;

· MrNeRF opened on Sep 26, 2026

Vulkan attention returns NaN for queries whose keys are all masked

Status: Open.

#2458 In MrNeRF/LichtFeld-Studio;

· MrNeRF opened on Sep 26, 2026