a1ea8c01c3
k-diffusion-euler ( #1019 )
...
* k-diffusion-euler
* make style make quality
* make fix-copies
* fix tests for euler a
* Update src/diffusers/schedulers/scheduling_euler_ancestral_discrete.py
Co-authored-by: Anton Lozhkov <aglozhkov@gmail.com>
* Update src/diffusers/schedulers/scheduling_euler_ancestral_discrete.py
Co-authored-by: Anton Lozhkov <aglozhkov@gmail.com>
* Update src/diffusers/schedulers/scheduling_euler_discrete.py
Co-authored-by: Anton Lozhkov <aglozhkov@gmail.com>
* Update src/diffusers/schedulers/scheduling_euler_discrete.py
Co-authored-by: Anton Lozhkov <aglozhkov@gmail.com>
* remove unused arg and method
* update doc
* quality
* make flake happy
* use logger instead of warn
* raise error instead of deprication
* don't require scipy
* pass generator in step
* fix tests
* Apply suggestions from code review
Co-authored-by: Pedro Cuenca <pedro@huggingface.co>
* Update tests/test_scheduler.py
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
* remove unused generator
* pass generator as extra_step_kwargs
* update tests
* pass generator as kwarg
* pass generator as kwarg
* quality
* fix test for lms
* fix tests
Co-authored-by: patil-suraj <surajp815@gmail.com>
Co-authored-by: Anton Lozhkov <aglozhkov@gmail.com>
Co-authored-by: Pedro Cuenca <pedro@huggingface.co>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
2022-10-31 16:20:38 +01:00
bf7b0bc25b
Allow `safety_checker` to be `None` when using CPU offload ( #1078 )
...
Allow None safety_checker when using CPU offload.
2022-10-31 15:03:33 +01:00
e4d264e4eb
[GitBot] Automatically close issues after inactivitiy ( #1079 )
...
* [GitBot] Automatically close issues after inactivitiy
* improve
* Add unstale
* typo
Co-authored-by: anton-l <anton@huggingface.co>
2022-10-31 14:06:03 +01:00
1606eb994a
Fix pipelines user_agent, ignore CI requests ( #1058 )
...
* Fix pipelines user_agent, ignore CI requests
* fix circular import
* N/A versions
* N/A versions
2022-10-31 13:38:43 +01:00
82d56cf192
Merge branch 'main' of https://github.com/huggingface/diffusers into main
2022-10-31 09:13:40 +00:00
707b8684b3
fix slow test
2022-10-31 09:13:37 +00:00
8e4fd686e0
Move safety detection to model call in Flax safety checker ( #1023 )
...
* Move safety detection to model call in Flax safety checker
* Update src/diffusers/pipelines/stable_diffusion/safety_checker_flax.py
2022-10-30 20:07:55 +01:00
95414bd6bf
Experimental: allow fp16 in `mps` ( #961 )
...
* Docs: refer to pre-RC version of PyTorch 1.13.0.
* Remove temporary workaround for unavailable op.
* Update comment to make it less ambiguous.
* Remove use of contiguous in mps.
It appears to not longer be necessary.
* Special case: use einsum for much better performance in mps
* Update mps docs.
* MPS: make pipeline work in half precision.
2022-10-29 21:09:32 +02:00
a59f9990fc
Tests: upgrade PyTorch cuda to 11.7 to fix examples tests. ( #1048 )
...
Tests: upgrade PyTorch cuda to 11.7.
Otherwise the cuda versions of torch and torchvision mismatch, and
examples tests fail. We were requesting cuda 11.6 for PyTorch, and the
default torchvision (via setup.py).
Another option would be to include torchvision in the same pip install
line as torch.
2022-10-29 20:27:00 +02:00
1fc208825d
Add seed resizing to community pipelines ( #1011 )
...
* add seed resizing to community examples
* actually add the file responsible for seed resizing
2022-10-29 09:31:42 +02:00
12fd0736dc
clean incomplete pages ( #1008 )
2022-10-29 09:28:26 +02:00
fc0ca47456
Fix speedup ratio in fp16.mdx ( #837 )
2022-10-29 09:26:23 +02:00
6b185b6acd
Update training and fine-tuning docs ( #1020 )
...
* Update training and fine-tuning docs.
* Update examples README.
* Update README.
* Add Flax fine-tuning section.
* Accept suggestion
Co-authored-by: Anton Lozhkov <anton@huggingface.co>
* Accept suggestion
Co-authored-by: Anton Lozhkov <anton@huggingface.co>
Co-authored-by: Anton Lozhkov <anton@huggingface.co>
2022-10-28 21:02:08 +02:00
81b6fbf19d
higher precision for vae
2022-10-28 18:19:06 +00:00
a7ae808ee2
increase tolerance
2022-10-28 17:50:22 +00:00
ea01a4c7f9
fix
2022-10-28 16:55:43 +00:00
cbbb29398a
hot fix
2022-10-28 16:55:21 +00:00
d37f08da72
[Tests] no random latents anymore ( #1045 )
2022-10-28 18:52:25 +02:00
c4ef1efe46
[Tests] Better prints ( #1043 )
2022-10-28 17:38:31 +02:00
8d6487f3cb
Fix some failing tests ( #1041 )
...
* up
* up
* up
* Update src/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion.py
* Apply suggestions from code review
2022-10-28 17:05:00 +02:00
d2d9764f35
[Tests] Speed up slow tests ( #1040 )
...
* [Tests] Speed up slow tests
* Up
* up
2022-10-28 14:46:39 +02:00
a80480f0f2
[Tests] Improve unet / vae tests ( #1018 )
...
* improve tests
* up
* finish
* upload
* add init
* up
* finish vae
* finish
* reduce loading time with device_map
* remove device_map from CPU
* uP
2022-10-28 13:43:26 +02:00
ab079f27cf
fix `F.interpolate()` for large batch sizes ( #1006 )
...
* fix `upsample_nearest_nhwc` for large bsz
* fix `upsample_nearest_nhwc` for large bsz
2022-10-28 11:25:21 +02:00
1e07b6b334
[Flax SD finetune] Fix dtype ( #1038 )
...
fix jnp dtype
2022-10-28 11:21:34 +02:00
fb38bb1621
Support grayscale images in `numpy_to_pil` ( #1025 )
2022-10-27 22:44:35 +02:00
de00c63217
Document sequential CPU offload method on Stable Diffusion pipeline ( #1024 )
...
* document cpu offloading method
* address review comments
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
2022-10-27 16:52:21 +02:00
a6314a8d4e
Add `--dataloader_num_workers` to the DDPM training example ( #1027 )
2022-10-27 15:55:36 +02:00
939ec17e91
Probably nicer to specify dependency on tensorboard in the training example ( #998 )
...
tensorboard import in readme, otherwise accelerator.trackers[0] out of range
Co-authored-by: lukovnikov <lukovnikov@users.noreply.github.com>
2022-10-27 15:55:18 +02:00
eceeebdf91
Update train_dreambooth.py
2022-10-27 15:51:11 +02:00
52f2128dc6
update readme for flax examples ( #1026 )
2022-10-27 15:25:25 +02:00
fbcc383340
Deprecate `init_git_repo`, refactor `train_unconditional.py` ( #1022 )
...
Deprecate `init_git_repo` and `push_to_hub`, refactor `train_unconditional.py`
2022-10-27 15:16:59 +02:00
90f91adb0e
[Flax] Add DreamBooth ( #1001 )
...
* [Flax] Add DreamBooth
* fix sample rng
* style
* not reuse rng
* add dtype for mixed precision training
* Add Flax example
2022-10-27 14:25:04 +02:00
4623f095f3
[DreamBooth] Set train mode for text encoder ( #1012 )
...
Set train mode for text encoder
2022-10-27 14:19:13 +02:00
abe058221c
[Flax] Add finetune Stable Diffusion ( #999 )
...
* [Flax] Add finetune Stable Diffusion
* temporary fix
* drop_last and seed
* add dtype for mixed precision training
* style
* Add Flax example
2022-10-27 14:08:21 +02:00
3be9fa97d6
[Accelerate model loading] Fix meta device and super low memory usage ( #1016 )
...
* [Accelerate model loading] Fix meta device and super low memory usage
* better naming
2022-10-27 12:11:42 +02:00
e92a603cab
fix dreambooth script. ( #1017 )
...
make input_args optional
2022-10-27 11:44:06 +02:00
1d04e1b4de
Continuation of #942 : additional float64 failure ( #996 )
...
* Add failing test for #940 .
* Do not use torch.float64 in mps.
* style
* Temporarily skip add_noise for IPNDMScheduler.
Until #990 is addressed.
* Fix additional float64 error in mps.
* Improve add_noise test
* Slight edit – I think it's clearer this way.
2022-10-27 10:21:40 +02:00
a23ad87d7a
[Flax] Add Textual Inversion ( #880 )
...
* add textual inversion flax
* make style
* make style
* replicate vae and unet params
* make style
* minor
* save after end of training
* style
* Temporary fix
Co-authored-by: Suraj Patil <surajp815@gmail.com>
* Add Flax instruction
Co-authored-by: Suraj Patil <surajp815@gmail.com>
2022-10-26 22:28:55 +02:00
d3d22ce5a8
Small modification to enable usage by external scripts ( #956 )
...
* Make training code usable by external scripts
Add parameter inputs to training and argument parsing function to allow this script to be used by an external call.
* Apply suggestions from code review
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
2022-10-26 18:46:56 +02:00
8332c1a6d9
Enable multi-process DataLoader for dreambooth ( #950 )
2022-10-26 17:24:48 +02:00
bd06dd023f
[inpaint pipeline] fix bug for multiple prompts inputs ( #959 )
2022-10-26 16:41:57 +02:00
b2e2d1411c
minimal stable diffusion GPU memory usage with accelerate hooks ( #850 )
...
* add method to enable cuda with minimal gpu usage to stable diffusion
* add test to minimal cuda memory usage
* ensure all models but unet are onn torch.float32
* move to cpu_offload along with minor internal changes to make it work
* make it test against accelerate master branch
* coming back, its official: I don't know how to make it test againt the master branch from accelerate
* make it install accelerate from master on tests
* go back to accelerate>=0.11
* undo prettier formatting on yml files
* undo prettier formatting on yml files againn
2022-10-26 15:52:57 +02:00
2f0fcf4fa8
Add missing import ( #979 )
2022-10-26 15:45:39 +02:00
cc436087d3
Fix typos ( #978 )
2022-10-26 15:32:47 +02:00
d7d6841406
fix a bug in the new version ( #957 )
...
remove tensor_format in the new version
2022-10-26 14:26:17 +02:00
d9cfe325a5
CompVis -> diffusers script - allow converting from merged checkpoint to either EMA or non-EMA ( #991 )
...
* improve script
* up
2022-10-26 12:32:07 +02:00
0343d8f531
Do not use torch.float64 on the mps device ( #942 )
...
* Add failing test for #940 .
* Do not use torch.float64 in mps.
* style
* Temporarily skip add_noise for IPNDMScheduler.
Until #990 is addressed.
2022-10-26 11:56:43 +02:00
4b9f58952a
Add --pretrained_model_name_revision option to train_dreambooth.py ( #933 )
...
* Add --pretrained_model_name_revision option to train_dreambooth.py
* Renamed --pretrained_model_name_revision to --revision
2022-10-25 21:38:23 +02:00
e2243de5f2
Fix typo in documentation title ( #975 )
2022-10-25 20:20:16 +02:00
59f0ce82eb
[Dance Diffusion] Better naming ( #981 )
...
uP
2022-10-25 19:52:41 +02:00
hlky