llama.cpp
Vulkan Repack PoC
#21024
Open
Overview
Commits
10
Changes
View On GitHub

Vulkan Repack PoC #21024

0cc4m wants to merge 10 commits into master from 0cc4m/vulkan-repack
0cc4m
github-actions github-actions added Vulkan
github-actions github-actions added ggml
savvadesogle
0cc4m
inforithmics
0cc4m vulkan: repack q4_0 into aligned arrays
c285bb98
0cc4m add coopmat2 support
5c1e95c9
0cc4m fix double semicolon
b1243aa9
0cc4m add mxfp4 repacking
b4e2621d
0cc4m add missing repacking functions
b64f294c
0cc4m replace malloc/free with thread_local memory
6906f781
0cc4m add q4_1, q8_0, iq4_nl repacking
57fb74fb
0cc4m deduplicate repacking code
13a55c8e
0cc4m 0cc4m force pushed from df488da9 to 13a55c8e 34 days ago
inforithmics
0cc4m wider loads
ff6ad609
0cc4m fix partial writes
938872e9
0cc4m
0cc4m
inforithmics
0cc4m 0cc4m changed the title Vulkan Q4_0 Repack PoC Vulkan Repack PoC 30 days ago
AndreNijman

Login to write a write a comment.

Login via GitHub

Reviewers
No reviews
Assignees
No one assigned
Labels
Vulkan ggml
Milestone
No milestone
Terms Of Use Privacy Policy Refund Policy Impressum