llama.cpp
metal : implement q5_0 and q5_1 kernels
#3648
Merged

metal : implement q5_0 and q5_1 kernels #3648

ggerganov merged 7 commits into ggml-org:master from jhen0409:metal-q5
jhen0409
jhen0409 metal : implement dequantize_q5_0
9c3e05d5
jhen0409 metal : block_q_n_dot_y for block_q5_0 (broken)
7ebd4acb
jhen0409
jhen0409 commented on 2023-10-17
jhen0409 jhen0409 force pushed from e8bc2a3a to e924f6c3 1 year ago
jhen0409 jhen0409 force pushed from e924f6c3 to 4f87b243 1 year ago
jhen0409 metal : revert unnecessary change
a7a4887b
jhen0409 jhen0409 force pushed from 4f87b243 to a7a4887b 1 year ago
jhen0409 metal : implement dequantize_q5_1
fce44a76
jhen0409 metal : block_q_n_dot_y for q5_1 (broken)
79d4732c
ggerganov
ggerganov commented on 2023-10-17
jhen0409 metal : fix block_q_n_dot_y
9db276f0
jhen0409
jhen0409 jhen0409 marked this pull request as ready for review 1 year ago
jhen0409 jhen0409 changed the title metal : implement q5_0 / q5_1 kernels metal : implement q5_0 and q5_1 kernels 1 year ago
ggerganov
jhen0409
ggerganov minor : spaces / formatting
7a885229
ggerganov
ggerganov approved these changes on 2023-10-18
ggerganov ggerganov merged c67fe68e into master 1 year ago
jhen0409 jhen0409 deleted the metal-q5 branch 1 year ago

Login to write a write a comment.

Login via GitHub

Reviewers
Assignees
No one assigned
Labels
Milestone