 w-e-w
|
1b16c62608
use shared.hf_endpoint (#16611)
|
9 maanden geleden |
 Jabasukuriputo Wang
|
895456c4a2
change state dict comparison to ref compare
|
1 jaar geleden |
 AUTOMATIC1111
|
eaba3d7349
send weights to target device instead of CPU memory
|
2 jaren geleden |
|
AUTOMATIC1111
|
86221269f9
RAM optimization round 2
|
2 jaren geleden |
|
AUTOMATIC1111
|
0a89cd1a58
Use less RAM when creating models
|
2 jaren geleden |
|
AUTOMATIC
|
762265eab5
autofixes from ruff
|
2 jaren geleden |
|
AUTOMATIC
|
668d7e9b9a
make it possible to load SD1 checkpoints without CLIP
|
2 jaren geleden |
|
AUTOMATIC
|
0792fae078
fix missing field for aesthetic embedding extension
|
2 jaren geleden |
|
AUTOMATIC
|
4bd490727e
fix for an error caused by skipping initialization, for realsies this time: TypeError: expected str, bytes or os.PathLike object, not NoneType
|
2 jaren geleden |
|
AUTOMATIC
|
0f8603a559
add support for transformers==4.25.1
|
2 jaren geleden |
|
AUTOMATIC
|
ce3f639ec8
add more stuff to ignore when creating model from config
|
2 jaren geleden |
|
AUTOMATIC
|
0c3feb202c
disable torch weight initialization and CLIP downloading/reading checkpoint to speedup creating sd model from config
|
2 jaren geleden |
