Brett Kuprel
|
5f4815775b
|
generate_image_stream
|
2 years ago |
|
Brett Kuprel
|
cf5b116284
|
update predict.py
|
2 years ago |
Brett Kuprel
|
d3beb3ab57
|
update colab to match predict.py
|
2 years ago |
|
Brett Kuprel
|
0ac2b07344
|
update cog.yaml
|
2 years ago |
|
Brett Kuprel
|
73f4a47fb5
|
update predict.py
|
2 years ago |
|
Brett Kuprel
|
970afc9631
|
Merge branch 'main' of https://github.com/kuprel/min-dalle
|
2 years ago |
|
Brett Kuprel
|
97d22f0fdd
|
handle log2_mid_count=0
|
2 years ago |
|
Brett Kuprel
|
41a9c9c69d
|
update colab to properly handle intermediate_image_count=1
|
2 years ago |
|
Brett Kuprel
|
25f5d8f800
|
update colab notebook to use log2_mid_count properly
|
2 years ago |
|
Brett Kuprel
|
1702d3c439
|
log2_mid_count
|
2 years ago |
|
Brett Kuprel
|
059d8f3448
|
update colab to use log2_mid_count
|
2 years ago |
|
Brett Kuprel
|
8c068f0dcd
|
display_size parameter in colab
|
2 years ago |
|
Brett Kuprel
|
0d9998926d
|
display intermediate images
|
2 years ago |
|
Brett Kuprel
|
b634375edf
|
display intermediate images
|
2 years ago |
|
Brett Kuprel
|
672f325b3b
|
14s now for 4x4 on A100
|
2 years ago |
|
Brett Kuprel
|
8f7f0ba756
|
update readme
|
2 years ago |
|
Brett Kuprel
|
ecc1534f87
|
5x5 grid works on colab P100 so should work on replicate
|
2 years ago |
|
Brett Kuprel
|
3dc92b19bd
|
update readme
|
2 years ago |
|
Brett Kuprel
|
e405577d99
|
4x4 works on colab T4 and P100
|
2 years ago |
|
Brett Kuprel
|
486c73c741
|
update cog.yaml
|
2 years ago |
|
Brett Kuprel
|
deefd24919
|
decode_row
|
2 years ago |
|
Brett Kuprel
|
884202239f
|
update notebook
|
2 years ago |
|
Brett Kuprel
|
e0fd4d4bff
|
update cog.yaml
|
2 years ago |
|
Brett Kuprel
|
6f617fe98f
|
inplace attention state, faster and less memory
|
2 years ago |
|
Brett Kuprel
|
aca617dc64
|
new decoder self attention reduced memory usage and 3x3 grids now work on T4 in colab
|
2 years ago |
|
Brett Kuprel
|
461890486e
|
update python version to 0.2.15
|
2 years ago |
|
Brett Kuprel
|
377d15cb16
|
faster decoder self attention
|
2 years ago |
|
Brett Kuprel
|
a79d30f718
|
png -> jpg
|
2 years ago |
|
Brett Kuprel
|
a21ea6ea14
|
update replicate, clear cuda cache if cuda available
|
2 years ago |
|
Brett Kuprel
|
256fd1aab0
|
empty cache if cuda available
|
2 years ago |
|
Brett Kuprel
|
9602055eaa
|
Merge pull request #57 from w4ffl35/bugfix-for-mega-model
Fixes #16 - mega model running out of memory
|
2 years ago |
w4ffl35
|
7d549505da
|
Fixes #16 - mega model running out of memory
Prior to this fix the mega model will (more often than not) fail when running in succession.
Clearing the cache seems to fix the issue.
|
2 years ago |
|
Brett Kuprel
|
6b433d5ba6
|
update colab
|
2 years ago |
|
Brett Kuprel
|
53741154c7
|
update readme
|
2 years ago |
|
Brett Kuprel
|
aa08f7f93c
|
update readme
|
2 years ago |
|
Brett Kuprel
|
ec14eeba62
|
convert pngs to jpgs
|
2 years ago |
|
Brett Kuprel
|
2661074589
|
update readme
|
2 years ago |
|
Brett Kuprel
|
6ce60179ff
|
update cog.yaml
|
2 years ago |
|
Brett Kuprel
|
948e05c0dd
|
update readme
|
2 years ago |
|
Brett Kuprel
|
e20c4a8c9f
|
updated colab
|
2 years ago |
|
Brett Kuprel
|
a5004efad0
|
update readme
|
2 years ago |
|
Brett Kuprel
|
b939a1e5ac
|
update cog.yaml
|
2 years ago |
|
Brett Kuprel
|
f0f9c5e83e
|
update replicate predict.py
|
2 years ago |
|
Brett Kuprel
|
8d808599d3
|
h100 stat tbd
|
2 years ago |
|
Brett Kuprel
|
229af7233d
|
h100 stat tbd
|
2 years ago |
|
Brett Kuprel
|
107213b28b
|
3x3 only works on colab if P100 is allocated
|
2 years ago |
|
Brett Kuprel
|
f1bb562ac1
|
update readme
|
2 years ago |
|
Brett Kuprel
|
134fdaf76b
|
add nvidia-smi to notebook
|
2 years ago |
|
Brett Kuprel
|
9bdba57195
|
update replicate
|
2 years ago |
|
Brett Kuprel
|
9e146e6675
|
update readme
|
2 years ago |
