llama.cpp: version bump b10729 (#18160)

mirror of https://github.com/ollama/ollama.git synced 2026-09-21 13:38:14 -05:00

* llama.cpp: version bump b10729

Regenerate the compat hooks patch for b10729: upstream removed the
whole-tensor load_data_for read (last consumer was llama-quantize,
which now reads slabs via load_data_range). Keep the existing hook
surface (constructor, skip loops, load_all_data, mtmd/clip) unchanged
and add maybe_load_text_tensor_range, which materializes a text load
op's output once per tensor and serves the new (offset, size) slab
reads from that cache.

* address comments

This commit is contained in:

Daniel Hiltgen

2026-09-01 16:24:15 -07:00

committed by

GitHub

parent 3ba380d0be

commit 5ec5804360

6 changed files with 208 additions and 19 deletions

LLAMA_CPP_VERSION

+1 -1

View File

@@ -1 +1 @@
 b10630
 b10729