add accelerate to load models with smaller memory footprint #361
patrickvonplaten
merged
21 commits into
huggingface:main
from
main
add accelerate to load models with smaller memory footprint
0ea501b9
remove low_cpu_mem_usage as it is reduntant
7631dd68
Merge branch 'main' of github.com:huggingface/diffusers into main
973eb23c
patrickvonplaten
commented
on 2022-09-16
patrickvonplaten
commented
on 2022-09-16
patrickvonplaten
commented
on 2022-09-16
patrickvonplaten
commented
on 2022-09-16
move accelerate init weights context to modelling utils
8592e23c
add test to ensure results are the same when loading with accelerate
76b8e4a0
add tests to ensure ram usage gets lower when using accelerate
dd7f9b9c
move accelerate logic to single snippet under modelling utils and rem…
ec5f7aa7
Merge branch 'huggingface:main' into main
ae5f56db
format code using to pass quality check
8392e3ff
fix imports with isor
615054af
add accelerate to test extra deps
75c08a94
Merge branch 'main' into main
7e06f3d8
patrickvonplaten
commented
on 2022-09-21
only import accelerate if device_map is set to auto
6189b86f
Merge branch 'main' of github.com:piEsposito/diffusers into main
02818b5e
Merge branch 'main' of github.com:huggingface/diffusers into main
dc14ace0
patrickvonplaten
commented
on 2022-09-22
patrickvonplaten
commented
on 2022-09-22
move accelerate availability check to diffusers import utils
bc510615
Merge remote-tracking branch 'upstream/main' into main
ad1b55d0
format code
e020d733
Merge branch 'main' into main
c3778bb4
Merge branch 'main' into main
0e2319d3
patrickvonplaten
commented
on 2022-10-04
Merge branch 'main' into main
25e07d8d
patrickvonplaten
approved these changes
on 2022-10-04
patil-suraj
approved these changes
on 2022-10-04
Reviewers
Assignees
No one assigned
Labels
None yet
Milestone
No milestone
Login to write a write a comment.
Login via GitHub