Search results

Delegate FP16 <-> BF16 to a kernel or the compiler

Status: Open.

#179 In ypapadop-amd/ggml;

· ypapadop-amd opened on Mar 12, 2026

Support sequences of views

Status: Open.

#155 In ypapadop-amd/ggml;

· ypapadop-amd opened on Dec 15, 2025

Support operations with in-place tensors

Status: Open.

#154 In ypapadop-amd/ggml;

· ypapadop-amd opened on Dec 15, 2025

Error result in test-vector-hsa

Status: Open.

#143 In ypapadop-amd/ggml;

· MaoxinYee opened on Oct 28, 2025

Add Support for different GGML_TYPE inputs

Status: Open.

#140 In ypapadop-amd/ggml;

· MaoxinYee opened on Oct 20, 2025

MNIST inference on NPU

Status: Open.

#119 In ypapadop-amd/ggml;

· ypapadop-amd opened on Oct 6, 2025

ggml-hsa in llama.cpp will get errors about pybind11 and numpy

Status: Open.

#117 In ypapadop-amd/ggml;

· MaoxinYee opened on Oct 1, 2025

How can I use this ggml-hsa to enable llama.cpp running on AMD NPU？

Status: Open.

#114 In ypapadop-amd/ggml;

· MaoxinYee opened on Sep 25, 2025

Does ggml_hsa support NPU on PHX/HPT platforms, or only STX or newer?

Status: Open.

#87 In ypapadop-amd/ggml;

· netqer opened on Jul 2, 2025

Create tool to pregenerate kernels

Status: Open.

#76 In ypapadop-amd/ggml;

· ypapadop-amd opened on Jun 6, 2025

Add design document

documentationImprovements or additions to documentationImprovements or additions to documentation

Status: Open.

#36 In ypapadop-amd/ggml;

· ypapadop-amd opened on Mar 31, 2025

Allocator for kernargs

enhancementNew feature or requestNew feature or request

Status: Open.

#27 In ypapadop-amd/ggml;

· ypapadop-amd opened on Mar 5, 2025