Brett Kuprel
|
798e6ac5a3
|
optionally tile images in token space
|
2 years ago |
Brett Kuprel
|
3c28b1059b
|
0.3.13, simplified code, specify device when initializing MinDalle
|
2 years ago |
Brett Kuprel
|
d64e957731
|
add temperature parameter
|
2 years ago |
Brett Kuprel
|
2ab4365059
|
add performance graph to readme
|
2 years ago |
Brett Kuprel
|
dba3f11b3f
|
faster inference with cuda/cudnn backends flags
|
2 years ago |
Brett Kuprel
|
703bfb231d
|
detach unnecessary
|
2 years ago |
Brett Kuprel
|
985c5fad4c
|
removing unused import
|
2 years ago |
Brett Kuprel
|
24162ee717
|
change grid_size argument of generate_images to image_count
|
2 years ago |
Brett Kuprel
|
736904ef2f
|
fix typing
|
2 years ago |
Brett Kuprel
|
2cac9220b5
|
generate_images_stream and generate_images
|
2 years ago |
Brett Kuprel
|
c199507a7a
|
vqgan needs to be float32
|
2 years ago |
Brett Kuprel
|
da62298f06
|
support bfloat16
|
2 years ago |
Brett Kuprel
|
f071b31bdd
|
properly limit input to 64 tokens
|
2 years ago |
Brett Kuprel
|
89a125b4b9
|
control top_k value
|
2 years ago |
Brett Kuprel
|
ba6080668f
|
update readme
|
2 years ago |
Brett Kuprel
|
9a4d9dc118
|
remove unused imports
|
2 years ago |
Brett Kuprel
|
6d81d514a0
|
control super condition factor
|
2 years ago |
Brett Kuprel
|
5f4815775b
|
generate_image_stream
|
2 years ago |
Brett Kuprel
|
97d22f0fdd
|
handle log2_mid_count=0
|
2 years ago |
Brett Kuprel
|
1702d3c439
|
log2_mid_count
|
2 years ago |
Brett Kuprel
|
0d9998926d
|
display intermediate images
|
2 years ago |
Brett Kuprel
|
deefd24919
|
decode_row
|
2 years ago |
Brett Kuprel
|
377d15cb16
|
faster decoder self attention
|
2 years ago |
Brett Kuprel
|
256fd1aab0
|
empty cache if cuda available
|
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
|
313635e914
|
fixed typing error for older python versions
|
2 years ago |
Brett Kuprel
|
1eb56737d8
|
added grid_size parameter to generate a grid of images
|
2 years ago |
Brett Kuprel
|
cf9656baa2
|
added is_verbose flag
|
2 years ago |
Brett Kuprel
|
35e97768a5
|
v0.2.0, MinDalleTorch -> MinDalle, breaking change
|
2 years ago |
Brett Kuprel
|
fff44d683e
|
update readme, random seed if none specified
|
2 years ago |
Brett Kuprel
|
be2beca7c0
|
added to pypi
|
2 years ago |
Brett Kuprel
|
18c72ed34d
|
simplified MinDalleTorch
|
2 years ago |
Brett Kuprel
|
07ce93d5f8
|
moved flax model and conversion code to separate repository
|
2 years ago |
Brett Kuprel
|
85f5866eff
|
remove config.json dependency, default to torch in image_from_text.py
|
2 years ago |
Brett Kuprel
|
09a0f85b8e
|
separate setup processes for flax and torch
|
2 years ago |
Brett Kuprel
|
7bf76deafb
|
fixed wrong file path
|
2 years ago |
Brett Kuprel
|
e4c2be54cb
|
save converted detokenizer params
|
2 years ago |
Brett Kuprel
|
08b158d580
|
updated readme
|
2 years ago |
Brett Kuprel
|
b913b58353
|
pre converting params to torch allows mega to run in standard colab runtime
|
2 years ago |
Brett Kuprel
|
c2a3858c96
|
delete params sooner
|
2 years ago |
Brett Kuprel
|
f951424e38
|
is_reusable
|
2 years ago |
Brett Kuprel
|
b55bcba4c0
|
removed deepcopy, delete expendable parameters after use
|
2 years ago |
Brett Kuprel
|
41a44068d0
|
keep params in expendable mode
|
2 years ago |
Brett Kuprel
|
fb97ba5e20
|
update readme, cleanup
|
2 years ago |
Brett Kuprel
|
1e18ba0ffa
|
is_expendable argument reduces memory usage for command line script
|
2 years ago |
Brett Kuprel
|
ed91ab4a30
|
refactored to load models once and run multiple times
|
2 years ago |
Adam Novak
|
28c812c832
|
Use all logical cores in Torch mode
|
2 years ago |
Brett Kuprel
|
1fbb209623
|
fixed bug with cuda in detokenizer
|
2 years ago |
Brett Kuprel
|
764b0bc685
|
cuda in detokenizer from previous commit broke colab flax model, fixed
|
2 years ago |
Brett Kuprel
|
17c96fe110
|
works with cuda
|
2 years ago |