llama.cpp
[SYCL] Optimize mul_mat for Q4_0 on Intel GPU
#12035
Merged

[SYCL] Optimize mul_mat for Q4_0 on Intel GPU #12035

NeoZhangJianyu
arthw opt performance by reorder for Intel GPU
78e232a0
arthw detect hw type and save opt feature, and print opt feature
7a6b48d7
arthw correct name
63e5285d
NeoZhangJianyu support optimize graph once when compute graph, record the opt status…
5cfde909
NeoZhangJianyu add env variable GGML_SYCL_DISABLE_OPT for debug
4eaab12e
github-actions github-actions added examples
github-actions github-actions added ggml
github-actions github-actions added SYCL
NeoZhangJianyu NeoZhangJianyu requested a review from airMeng airMeng 298 days ago
airMeng
airMeng commented on 2025-02-24
NeoZhangJianyu use syclex::architecture replace the custom hw define, update the gui…
0e91e0eb
NeoZhangJianyu add performance data
b3570b92
NeoZhangJianyu mv getrows functions to separeted files
f1117218
github-actions github-actions added documentation
airMeng
airMeng approved these changes on 2025-02-24
NeoZhangJianyu Merge branch 'master' into opt_reorder
c541d6ae
NeoZhangJianyu fix global variables
30ddc909
NeoZhangJianyu NeoZhangJianyu merged 08d59862 into master 297 days ago
Alcpz
Alcpz commented on 2025-02-24
qnixsynapse
Alcpz
qnixsynapse
Alcpz
Alcpz
qnixsynapse
ky438
NeoZhangJianyu
ky438
NeoZhangJianyu

Login to write a write a comment.

Login via GitHub

Reviewers
Assignees
No one assigned
Labels
Milestone