PR #5132 New Features to run MobileVLM on orin

New Features to run MobileVLM on orin #5132

ggerganov merged 17 commits into ggml-org:master from JidongZhang-THU:MobileVLM

New Feature:

d89e0e09

fix tabs instead of spaces

b08c6b1a

slaren commented on 2024-01-26

code clean

c29a8554

ggerganov commented on 2024-01-26

CUDA POOL2D

ba5592c6

ADD POOL2D test case in test-backend-ops.cpp

1a827880

slaren commented on 2024-01-29

code clean

41a34cb3

fix pool2d_kernel

1556d4ca

JidongZhang-THU force pushed from 00925f83 to 1556d4ca 1 year ago

fix bug in pool2d kernel

379f89fb

fix avg pooling, count_include_pad

49f09aa7

JidongZhang-THU force pushed from 212da076 to 49f09aa7 1 year ago

test-backend-ops : add more pool_2d tests

04f10a22

cuda : fix warnings and formatting

caf2fc82

slaren commented on 2024-01-30

ggml : check types in release builds too in pool_2d

bdf3b8ad

test-backend-ops : remove f16 pool_2d tests

8824e427

cuda : more style fixes

0d94da7c

JidongZhang-THU closed this 1 year ago

JidongZhang-THU reopened this 1 year ago

Add assert in ggml_cuda_op_pool2d

ca4ec6d8

pool2d float padding fallback

66dd123b

test-backend-ops : add dst_type to im2col

18fd0b0c

slaren approved these changes on 2024-01-31

ggerganov approved these changes on 2024-01-31

ggerganov merged 15606309 into master 1 year ago

Reviewers

ggerganov

slaren

Assignees

No one assigned

Labels

None yet

Milestone

No milestone

llama.cpp New Features to run MobileVLM on orin #5132 Merged

New Features to run MobileVLM on orin #5132

llama.cpp
New Features to run MobileVLM on orin
#5132

Merged