github ggml-org/llama.cpp b8420

2 hours ago
Details

CANN: handle in-place ROPE on non-contiguous f32 tensors (#20274)

RotaryPositionEmbedding on CANN fails when src and dst share the same
non-contiguous buffer (inplace + view), because the operator overwrites
source data before it is fully read.

Add a branch that detects this case and uses contiguous temporary
buffers: copy src to temp, run ROPE into another temp, then copy back
to the non-contiguous dst. Fixes 20 failing ROPE tests (f32, v=1,
inplace=1).

Signed-off-by: noemotiovon 757486878@qq.com

macOS/iOS:

Linux:

Windows:

openEuler:

Don't miss a new llama.cpp release

NewReleases is sending notifications on new releases.