Mirrored_Repos/InvokeAI
1
0
Fork 0
mirror of https://github.com/invoke-ai/InvokeAI synced 2024-08-30 20:32:17 +00:00
Code Issues Packages Projects Releases Wiki Activity
2,252 Commits 274 Branches 239 Tags 397 MiB
6e4dad60c8
Commit Graph

110 Commits

Author SHA1 Message Date
Damian Stewart
23eb80b404 attention maps callback stuff for diffusers 2022-12-14 21:04:55 +01:00
Kevin Turner
1f86e527aa
Merge branch 'main' into dev/diffusers 2022-12-14 08:25:03 -08:00
Ivan Efimov
ccf8a46320 Fix: define path as None before usage 2022-12-13 19:46:03 -05:00
rmagur1203
f4e7383490
Load model in inpaint when using free_gpu_mem option (#1938) 
* Load model in inpaint when using free_gpu_mem option

* Passing free_gpu_mem option to inpaint generator
 2022-12-12 09:14:30 -05:00
rmagur1203
96a12099ed
Fix the mistake of not importing the gc (#1939) 2022-12-12 09:14:09 -05:00
rmagur1203
bd0c0d77d2
Reduce more memories on free_gpu_mem option (#1915) 
* Enhance free_gpu_mem option
Unload cond_stage_model on free_gpu_mem option is setted

* Enhance free_gpu_mem option
Unload cond_stage_model on free_gpu_mem option is setted
 2022-12-11 13:49:55 -05:00
Kevin Turner
520c17ab86 diffusers(AddsMaskedGuidance): partial fix for k-schedulers 
Prevents them from crashing, but results are still hot garbage.
 2022-12-10 21:19:32 -08:00
Kevin Turner
cd358c40ad diffusers: stopgap fix for attention_maps_callback crash after recent merge 2022-12-10 19:48:15 -08:00
Kevin Turner
061c5369a2 Merge remote-tracking branch 'origin/main' into dev/diffusers 2022-12-10 19:18:00 -08:00
Damian Stewart
8cbb50c204 avoid further crash under low-memory conditions 2022-12-10 15:32:11 -05:00
Kevin Turner
63532226a5 Merge remote-tracking branch 'origin/main' into dev/diffusers 
# Conflicts:
#	backend/invoke_ai_web_server.py
#	ldm/generate.py
#	ldm/invoke/CLI.py
#	ldm/invoke/generator/base.py
#	ldm/invoke/generator/txt2img.py
#	ldm/models/diffusion/cross_attention_control.py
#	ldm/modules/attention.py
 2022-12-10 08:43:37 -08:00
Kevin Turner
50c48cffc7 diffusers: make masked img2img behave better with multi-step schedulers 
re-randomizing the noise each step was confusing them.
 2022-12-10 08:27:46 -08:00
Kevin Turner
1a67836012 diffusers: use xformers when available 
diffusers no longer auto-enables this as of 0.10.2.
 2022-12-10 08:26:24 -08:00
Damian Stewart
786b8878d6
Save and display per-token attention maps (#1866) 
* attention maps saving to /tmp

* tidy up diffusers branch backporting of cross attention refactoring

* base64-encoding the attention maps image for generationResult

* cleanup/refactor conditioning.py

* attention maps and tokens being sent to web UI

* attention maps: restrict count to actual token count and improve robustness

* add argument type hint to image_to_dataURL function

Co-authored-by: psychedelicious <4822129+psychedelicious@users.noreply.github.com>

Co-authored-by: damian <git@damianstewart.com>
Co-authored-by: psychedelicious <4822129+psychedelicious@users.noreply.github.com>
 2022-12-10 15:57:41 +01:00
Kevin Turner
a3e41d63ec Merge remote-tracking branch 'origin/main' into dev/diffusers 
# Conflicts:
#	.github/workflows/test-invoke-conda.yml
#	.github/workflows/test-invoke-pip.yml
#	environments-and-requirements/requirements-base.txt
 2022-12-09 10:34:29 -08:00
Jonathan
4e61069821 Update embiggen.py 2022-12-08 22:34:49 -05:00
Kevin Turner
9199d698f8 diffusers: upgrade to diffusers 0.10, add Heun scheduler 2022-12-08 13:02:47 -08:00
Kevin Turner
9bcb3b1bf7 fixup! refactor(diffusers): reduce some code duplication amongst the different tasks 2022-12-07 18:26:08 -08:00
Kevin Turner
821c7df240 refactor(diffusers): reduce some code duplication amongst the different tasks 2022-12-07 18:20:56 -08:00
Daya Adianto
d7ba041007 Enable force free GPU memory in img2img 2022-12-07 19:25:21 -05:00
Kevin Turner
04a5bc938e diffusers: txt2img2img (hires_fix) 
with so much slicing and dicing of pipeline methods to stitch them together
 2022-12-06 22:28:09 -08:00
Kevin Turner
bf6376417a diffusers: reset num_vectors_per_token 
sync with 44a0055571
 2022-12-06 15:52:49 -08:00
Kevin Turner
5c7e6751e0 inpainting for the normal model. I think it works this time. 2022-12-05 12:37:20 -08:00
Damian Stewart
69d42762de refactor common CrossAttention stuff into a mixin so that the old ldm code can still work if necessary 2022-12-05 20:00:18 +01:00
Kevin Turner
f3570d8344 inpainting for the normal model [WIP] 
This seems to be performing well until the LAST STEP, at which point it dissolves to confetti.
 2022-12-04 23:36:12 -08:00
Kevin Turner
875312080d diffusers support for the inpainting model 2022-12-04 20:12:04 -08:00
Damian Stewart
ff42027a00 add cross-attention control support to diffusers (fails on MPS) 
For unknown reasons MPS produces garbage output with .swap(). Use
--always_use_cpu arg to invoke.py for now to test this code on MPS.
 2022-12-04 16:07:54 +01:00
Damian Stewart
f48706efee add always_use_cpu arg to bypass MPS 2022-12-04 15:15:39 +01:00
Damian Stewart
9ea10ece95 monkey-patch diffusers.attention and use Invoke lowvram code 2022-12-03 10:10:37 +01:00
Kevin Turner
9d7fe45702 Merge remote-tracking branch 'origin/main' into dev/diffusers 
# Conflicts:
#	scripts/configure_invokeai.py
 2022-12-01 21:38:46 -08:00
Lincoln Stein
bde456f9fa fix startup messages and a startup crash 
- make the warnings about patchmatch less redundant
- only warn about being unable to load concepts from Hugging Face
  library once
- do not crash when unable to load concepts from Hugging Face
  due to network connectivity issues
 2022-12-01 07:42:31 -05:00
Kevin Turner
ea1cf83c20 preliminary support for outpainting (no masking yet) 2022-11-30 22:05:58 -08:00
Kevin Turner
494936a8d2 🚧 post-rebase repair 2022-11-30 15:39:40 -08:00
Kevin Turner
7c558d5056 fix "config attributes…not expected" diffusers warnings. 2022-11-30 15:34:19 -08:00
Kevin Turner
8157bff4bc internal: avoid importing diffusers DummyObject 
see https://github.com/huggingface/diffusers/issues/1479
 2022-11-30 15:34:19 -08:00
Kevin Turner
ca1f76b7ba create an embedding_manager for diffusers 2022-11-30 15:34:19 -08:00
Kevin Turner
56153c2ebf lint: correct annotations for Python 3.9. 2022-11-30 15:33:02 -08:00
Kevin Turner
b7864aa1a7 refactor: remove backported img2img.get_timesteps 
now that we can use it directly from diffusers 0.8.1
 2022-11-30 15:33:02 -08:00
Kevin Turner
ceb53ccdfb dev: upgrade to diffusers 0.8 (from 0.7.1) 
We get to remove some code by using methods that were factored out in the base class.
 2022-11-30 15:33:02 -08:00
Kevin Turner
efbb807905 diffusers integration: support img2img 2022-11-30 15:33:02 -08:00
Kevin Turner
95db6e80ee cross_attention_control: stub (no-op) implementations for diffusers 2022-11-30 15:01:27 -08:00
Kevin Turner
b6b1a8d97c diffusers: use InvokeAIDiffuserComponent for conditioning 2022-11-30 15:01:18 -08:00
Kevin Turner
e99faeb8d7 diffusers: restore prompt weighting feature 2022-11-30 15:01:15 -08:00
Kevin Turner
05a1d68ef4 web server: update image_progress callback for diffusers data 2022-11-30 15:01:06 -08:00
Kevin Turner
1e98f4bafc diffusers: let the scheduler do its scaling of the initial latents 
Remove IPNDM scheduler; it is not behaving.
 2022-11-30 15:01:06 -08:00
Kevin Turner
d55e22981a txt2img: support switching diffusers schedulers 2022-11-30 15:01:06 -08:00
Kevin Turner
ae9b482acf model_cache: fix model default image dimensions 2022-11-30 15:01:06 -08:00
Kevin Turner
4c3858e079 model_cache: add ability to load a diffusers model pipeline 
and update associated things in Generate & Generator to not instantly fail when that happens
 2022-11-30 15:00:59 -08:00
Kevin Turner
dcfdb83513 doc: type hints for Generator 2022-11-30 14:56:30 -08:00
Kevin Turner
58ea3bf4c8 spike: proof of concept using diffusers for txt2img 2022-11-30 14:56:26 -08:00