8e844f2eae
version: 1.6.2
2 years ago
cc014685bd
fix: leave sample in float32 for all final processing
2 years ago
09f00f9652
version: 1.6.1
2 years ago
9614a82f12
fix: use float32 at image render time for wider cpu compatibility
...
If the x_sample was a bfloat on the gpu but the cpu doesn't support bfloat, that can cause a TypeError
```
File "/home/stdiff/.local/lib/python3.10/site-packages/imaginairy/api.py", line 292, in imagine
x_sample.cpu().numpy(), "c h w -> h w c"
TypeError: Got unsupported ScalarType BFloat16`
```
2 years ago
08fca72033
Merge pull request #18 from brycedrennan/bugfixes
...
Bugfixes + per-prompt tile mode
2 years ago
a275d9c9cd
version: 1.6.0 + docs
2 years ago
cdfeaa4c6f
tests: add docker image for testing environment. minor test improvements
2 years ago
09bc1c70e6
fix: try to address #13 BFloat16 issue
...
Seems to be caused by incompatible types in group_norm when we use autocast.
Patch group_norm to cast the weights to the same type as the inputs
From what I can understand all the other repos just switch to full precision instead
of addressing this. I think this would make things slower but I'm not sure. So maybe
the patching solution I'm doing is better?
https://github.com/pytorch/pytorch/pull/81852
2 years ago
e23e363bf5
lint
2 years ago
ff7fc0dab7
feature: tile mode can now be specified per-prompt
2 years ago
41b0ae99b3
version 1.5.4
...
minor changes to help with google colab compat
2 years ago
6f77c23c1d
version 1.5.4
...
minor changes to help with google colab compat
2 years ago
83c809e2b7
version 1.5.3
...
bugfix missing config file
2 years ago
3d64867e1f
version 1.5.2
...
bugfix missing config file
2 years ago
37db8db1a2
version 1.5.1
2 years ago
a46424c673
feature: img2img now supported with PLMS (instead of just DDIM)
...
Kinda hacky copy/pasting from ddim. Need to cleanup
2 years ago
3a10a2eb80
docs: osx => macOS
2 years ago
ced6ec8151
Merge pull request #11 from brycedrennan/describe-images
...
feature: generate captions for images
2 years ago
4705d182d5
feature: generate captions for images
...
- add wip functionality for negative masks
- ci: add code linter that removes unused imports
- add instructions to install rust on osx
2 years ago
60a04a5d68
version: 1.4.0
...
we can get the stem of the pear
2 years ago
f040e3bffe
feature: allow multiple additive targets for text masking targets
2 years ago
a0087c1304
Merge branch 'clip-masking'
2 years ago
be85133070
version: 1.3.0
2 years ago
a5e010d01f
Merge pull request #9 from brycedrennan/clip-masking
...
Clip masking
2 years ago
930295d840
feature: txt2mask - automated text replacement
...
from https://github.com/timojl/clipseg
2 years ago
7087c4a680
feature: automatic mask generation
...
mask generation working but not integrated
2 years ago
8238e59067
tests: add more tests
2 years ago
d7cbf6e416
refactor: cleanup ddim
...
- delete more unused code
- fix some lints
2 years ago
6307a0daf5
refactor: cleanup ddim
2 years ago
89908603cf
fix: don't repeatedly download the same url
2 years ago
59648dbe61
test: add some tests/logging
...
add some experiments
2 years ago
19d0b563ac
style: lint fixes. remove unused code
2 years ago
793df6e9dc
docs update
2 years ago
33f1e2aeec
version 1.2.0
...
- add url init images
- don't install protobuf versions that are causing problems
https://github.com/brycedrennan/imaginAIry/issues/7
https://github.com/protocolbuffers/protobuf/issues/10571
2 years ago
af3dc6bacd
fix: restrict usage of difficult protobuf versions
2 years ago
6b4c348132
Merge pull request #6 from brycedrennan/bd/init-img-url
...
feature: urls as init images
2 years ago
c7a822d701
feature: urls as init images
...
- --init-image accepts urls
- cleanup command line code
2 years ago
47db34647b
version 1.1.4
2 years ago
23f0e2be12
fix: select ddim for img2img
2 years ago
da0ba308b2
version: 1.1.3
...
fix compat with python 3.8,3.9
2 years ago
8c064587d9
Merge pull request #1 from wseagar/ws/nsfw_fix
...
fix: is_nsfw_img is None rather than false
2 years ago
56fc414724
fix: is_nsfw_img is None rather than false
2 years ago
a1675fa599
docs: add dockerfile example
2 years ago
fa3673ef56
fix: use context manager syntax compatible with python 3.8
...
parentheses supported in python 3.10
2 years ago
51a8b1b8d8
fix: filter logic was wrong
2 years ago
0834db63c8
version 1.1.1
2 years ago
967eb76365
refactor/test: logging suppression + hashed image test
...
- simpler logging suppression for `transformers` library
- suppress logging noise for running tests
- get test running for all samplers on mps and cuda platforms
- refactor safety model env variable to allow classification
2 years ago
8c88f495d2
version
2 years ago
bb665b9eb6
fix: make k-diffusion samplers deterministic
...
- add test for hashes on mps. images look same on CUDA but are slightly different.
2 years ago
b4a3b8c2b3
feature: k-diffusion samplers
...
- improved image logging functionality. can just stick log_latent wherever you want
- improved some variable naming
- moved all the samplers together
- vendored k-diffusion library
2 years ago
Bryce
wseagar