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
12,662 Commits 274 Branches 239 Tags 397 MiB
e3a7bf12c1
Commit Graph

1163 Commits

Author SHA1 Message Date
Sergey Borisov
6af659b1da Handle t2i adapter in modular denoise 2024-07-24 02:55:33 +03:00
psychedelicious
76aa6bdf05 feat(nodes): split spandrel node 
`spandrel_image_to_image` now just runs the model with no changes.

`spandrel_image_to_image_autoscale` runs the model repeatedly until the desired scale is reached. previously, `spandrel_image_to_image` did this.
 2024-07-24 08:22:46 +10:00
Sergey Borisov
19c00241c6 Use non-inverted mask generally(except inpaint model handling) 2024-07-24 00:59:13 +03:00
Sergey Borisov
c323a760a5 Suggested changes 
Co-Authored-By: Ryan Dick <14897797+RyanJDick@users.noreply.github.com>
 2024-07-23 23:34:28 +03:00
Sergey Borisov
9d1fcba415 Fix create gradient mask node output 2024-07-23 23:29:28 +03:00
Sergey Borisov
ca21996a97 Remove old seamless class 2024-07-23 18:04:33 +03:00
Sergey Borisov
62aa064e56 Handle seamless in modular denoise 2024-07-23 18:03:59 +03:00
Ryan Dick
e2e47fd606 Merge branch 'main' into stalker-modular_controlnet 2024-07-23 10:19:12 -04:00
Ryan Dick
db52f5606f Merge branch 'main' into stalker-modular_freeu 2024-07-23 09:53:32 -04:00
Ryan Dick
d014dc94fd
Merge branch 'main' into stalker7779/modular_rescale_cfg 2024-07-23 09:34:22 -04:00
psychedelicious
29efab70b7 feat(nodes): spandrel_image_to_image.scale defaults to 4.0 2024-07-23 10:55:54 +10:00
psychedelicious
ac6adc392a feat(nodes): add scale and fit_to_multiple_of_8 to spandrel node 2024-07-23 10:55:54 +10:00
psychedelicious
a2ef5d56ee feat(nodes): split out spandrel node upscale logic into utils 2024-07-23 10:55:54 +10:00
Sergey Borisov
4e8dcb7a1a Suggested changes 
Co-Authored-By: Ryan Dick <14897797+RyanJDick@users.noreply.github.com>
 2024-07-23 01:46:29 +03:00
Sergey Borisov
87eb018380 Revert debug change 2024-07-22 23:49:20 +03:00
Sergey Borisov
5003e5d763 Same changes as in other PRs, add check for running inpainting on inpaint model without source image 
Co-Authored-By: Ryan Dick <14897797+RyanJDick@users.noreply.github.com>
 2024-07-22 23:47:39 +03:00
Sergey Borisov
58f3072b91 Handle inpainting on normal models 2024-07-21 22:17:29 +03:00
Sergey Borisov
9e7b470189 Handle inpaint models 2024-07-21 20:45:55 +03:00
Sergey Borisov
42356ec866 Add ControlNet support to denoise 2024-07-21 20:01:30 +03:00
Sergey Borisov
e046e60e1c Add FreeU support to denoise 2024-07-21 18:31:10 +03:00
Sergey Borisov
9a1420280e Add rescale cfg support to denoise 2024-07-21 17:33:43 +03:00
Ryan Dick
a8cc5caf96 Fix function call that we forgot to update in https://github.com/invoke-ai/InvokeAI/pull/6606 2024-07-19 17:07:52 -04:00
Ryan Dick
78d2b1b650 Merge branch 'main' into stalker-backend_base 2024-07-19 16:25:20 -04:00
Ryan Dick
e16faa6370 Add gradient blending to tile seams in MultiDiffusion. 2024-07-19 13:05:50 -07:00
Sergey Borisov
0c56d4a581 Ryan's suggested changes to extension manager/extensions 
Co-Authored-By: Ryan Dick <14897797+RyanJDick@users.noreply.github.com>
 2024-07-18 23:49:44 +03:00
StAlKeR7779
710dc6b487
Merge branch 'main' into stalker7779/backend_base 2024-07-18 01:08:04 +03:00
Sergey Borisov
2ef3b49a79 Add run cancelling logic to extension manager 2024-07-17 04:39:15 +03:00
Sergey Borisov
79e35bd0d3 Minor fixes 2024-07-17 03:48:37 +03:00
Ryan Dick
6b0ca88177
Merge branch 'main' into ryan/spandrel-upscale-tiling 2024-07-16 15:40:14 -04:00
Ryan Dick
81991e072b Merge branch 'main' into ryan/spandrel-upscale 2024-07-16 15:14:08 -04:00
Sergey Borisov
cec345cb5c Change attention processor apply logic 2024-07-16 20:03:29 +03:00
Sergey Borisov
608cbe3f5c Separate inputs in denoise context 2024-07-16 19:30:29 +03:00
Sergey Borisov
9f088d1bf5 Multiple small fixes 2024-07-16 00:51:25 +03:00
Sergey Borisov
499e4d4fde Add preview extension to check logic 2024-07-13 00:45:04 +03:00
Sergey Borisov
3a9dda9177 Renames 2024-07-12 22:44:00 +03:00
Sergey Borisov
87e96e1be2 Rename modifiers to callbacks, convert order to int, a bit unify injection points 2024-07-12 22:01:05 +03:00
Sergey Borisov
9cc852cf7f Base code from draft PR 2024-07-12 20:31:26 +03:00
psychedelicious
712cf00a82 fix(app): vae tile size field description 2024-07-12 06:30:27 -07:00
Ryan Dick
0428ce73a9 Add early cancellation to SpandrelImageToImageInvocation. 2024-07-11 15:42:33 -04:00
Ryan Dick
d0d2955992 Reduce peak VRAM utilization of SpandrelImageToImageInvocation. 2024-07-10 14:25:19 -04:00
Ryan Dick
d868d5d584 Make SpandrelImageToImage tiling much faster. 2024-07-10 14:25:19 -04:00
Ryan Dick
ab775726b7 Add tiling support to the SpoandrelImageToImage node. 2024-07-10 14:25:19 -04:00
Ryan Dick
7b5d4935b4 Merge branch 'main' into ryan/spandrel-upscale 2024-07-09 13:47:11 -04:00
Ryan Dick
ecbff2aa44 Whoops... forgot to commit this file. 2024-07-05 14:57:05 -04:00
Ryan Dick
d09999736c Rename spandrel models to 'Image-to-Image Model' throughout the UI. 2024-07-05 14:04:08 -04:00
Ryan Dick
35f8781ea2 Fix static type errors with SCHEDULER_NAME_VALUES. And, avoid bi-directional cross-directory imports, which contribute to circular import issues. 2024-07-05 07:38:35 -07:00
Ryan Dick
1d449097cc Apply ruff rule to disallow all relative imports. 2024-07-04 09:35:37 -04:00
Ryan Dick
9da5925287 Add ruff rule to disallow relative parent imports. 2024-07-04 09:35:37 -04:00
Ryan Dick
6161aa73af Move pil_to_tensor() and tensor_to_pil() utilities to the SpandrelImageToImage class. 2024-07-03 16:28:21 -04:00
Ryan Dick
504a42fe61 typo: fix UIType on Spandrel Upscaling node. 2024-07-03 16:28:21 -04:00
Ryan Dick
29c8ddfb88 WIP - A bunch of boilerplate to support Spandrel Image-to-Image models throughout the model manager and the frontend. 2024-07-03 16:28:21 -04:00
Ryan Dick
95079dc7d4 Use a ModelIdentifierField to identify the spandrel model in the UpscaleSpandrelInvocation. 2024-07-03 16:28:21 -04:00
Ryan Dick
c1afe35704 Add prototype invocation for running upscaling models with spandrel. 2024-07-03 16:28:21 -04:00
Ryan Dick
3752509066 Expose the VAE tile_size on the VAE encode and decode invocations. 2024-07-02 09:07:03 -04:00
Ryan Dick
79640ba14e Add context manager for overriding VAE tiling params. 2024-07-02 09:07:03 -04:00
Ryan Dick
bd74b84cc5 Revert "Remove the redundant init_timestep parameter that was being passed around. It is simply the first element of the timesteps array." 
This reverts commit fa40061eca.
 2024-06-26 12:51:51 -04:00
Ryan Dick
b9946e50f9 Use image-space tile dimensions on the TiledMultiDiffusionDenoiseLatents invocation. This is more natural for many users. 2024-06-25 11:31:52 -07:00
Ryan Dick
06f49a30f6 Mark TiledMultiDiffusionDenoiseLatents as a Beta node. 2024-06-25 11:31:52 -07:00
Ryan Dick
e1af78c702 Make the tile_overlap input to MultiDiffusion *strictly* control the amount of overlap rather than being a lower bound. 2024-06-25 11:31:52 -07:00
Ryan Dick
7c032ea604 (minor) Fix some documentation typos. 2024-06-25 11:31:52 -07:00
Ryan Dick
c5ee415607 Add progress image callbacks to TiledMultiDiffusionDenoiseLatentsInvocation. 2024-06-25 11:31:52 -07:00
Ryan Dick
fa40061eca Remove the redundant init_timestep parameter that was being passed around. It is simply the first element of the timesteps array. 2024-06-25 11:31:52 -07:00
Ryan Dick
7cafd78d6e Revert "Expose vae_decode(...) as a staticmethod on LatentsToImageInvocation." 
This reverts commit 753239b48d.
 2024-06-25 11:31:52 -07:00
Ryan Dick
8a43656cf9 (minor) Address a few small TODOs. 2024-06-25 11:31:52 -07:00
Ryan Dick
bd3b6ca11b Remove TiledStableDiffusionRefineInvocation. It was a proof-of-concept that has been superseded by TiledMultiDiffusionDenoiseLatents. 2024-06-25 11:31:52 -07:00
Ryan Dick
ceae5fe1db (minor) typo 2024-06-25 11:31:52 -07:00
Ryan Dick
c22526b9d0 Fix handling of stateful schedulers in MultiDiffusionPipeline. 2024-06-25 11:31:52 -07:00
Ryan Dick
c881882f73 Connect TiledMultiDiffusionDenoiseLatents to the MultiDiffusionPipeline backend. 2024-06-25 11:31:52 -07:00
Ryan Dick
80a67572f1 Fix invocation name of tiled_multi_diffusion_denoise_latents. 2024-06-25 11:31:52 -07:00
Ryan Dick
60ac937698 Improve clarity of comments regarded when 'noise' and 'latents' are expected to be set. 2024-06-25 11:31:52 -07:00
Ryan Dick
196f3b721d Stricter typing for the is_gradient_mask: bool. 2024-06-25 11:31:52 -07:00
Ryan Dick
40ae174c41 Fix typing of timesteps and init_timestep. 2024-06-25 11:31:52 -07:00
Ryan Dick
ffc28176fe Remove unused num_inference_steps. 2024-06-25 11:31:52 -07:00
Ryan Dick
230e205541 WIP TiledMultiDiffusionDenoiseLatents. Updated parameter list and first half of the logic. 2024-06-25 11:31:52 -07:00
Ryan Dick
7e94350351 Tidy DenoiseLatentsInvocation.prep_control_data(...) and fix some type errors. 2024-06-25 11:31:52 -07:00
Ryan Dick
c4e8549c73 Make DenoiseLatentsInvocation.prep_control_data(...) a staticmethod so that it can be called externally. 2024-06-25 11:31:52 -07:00
Ryan Dick
350a210835 Copy TiledStableDiffusionRefineInvocation as a starting point for TiledMultiDiffusionDenoiseLatents.py 2024-06-25 11:31:52 -07:00
Ryan Dick
ed781dbb0c Change tiling strategy to make TiledStableDiffusionRefineInvocation work with more tile shapes and overlaps. 2024-06-25 11:31:52 -07:00
Ryan Dick
b41ea963e7 Expose a few more params from TiledStableDiffusionRefineInvocation. 2024-06-25 11:31:52 -07:00
Ryan Dick
da5d105049 Add support for LoRA models in TiledStableDiffusionRefineInvocation. 2024-06-25 11:31:52 -07:00
Ryan Dick
5301770525 Add naive ControlNet support to TiledStableDiffusionRefineInvocation 2024-06-25 11:31:52 -07:00
Ryan Dick
534640ccde Rough prototype of TiledStableDiffusionRefineInvocation is working. 2024-06-25 11:31:52 -07:00
Ryan Dick
d5ab8cab5c WIP - TiledStableDiffusionRefine 2024-06-25 11:31:52 -07:00
Ryan Dick
4767301ad3 Minor improvements to LatentsToImageInvocation type hints. 2024-06-25 11:31:52 -07:00
Ryan Dick
21d7ca45e6 Expose vae_decode(...) as a staticmethod on LatentsToImageInvocation. 2024-06-25 11:31:52 -07:00
Ryan Dick
020e8eb413 Fix return type of prepare_noise_and_latents(...). 2024-06-25 11:31:52 -07:00
Ryan Dick
3d49541c09 Make init_scheduler() a staticmethod on DenoiseLatentsInvocation so that it can be called externally. 2024-06-25 11:31:52 -07:00
Ryan Dick
1ef266845a Only allow a single positive/negative prompt conditioning input for tiled refine. 2024-06-25 11:31:52 -07:00
Ryan Dick
a37589ca5f WIP on TiledStableDiffusionRefine 2024-06-25 11:31:52 -07:00
Ryan Dick
171a505f5e Convert several methods in DenoiseLatentsInvocation to staticmethods so that they can be called externally. 2024-06-25 11:31:52 -07:00
Ryan Dick
8004a0d5f5 Simplify the logic in prepare_noise_and_latents(...). 2024-06-25 11:31:52 -07:00
Ryan Dick
610a1fd611 Split out the prepare_noise_and_latents(...) logic in DenoiseLatentsInvocation so that it can be called from other invocations. 2024-06-25 11:31:52 -07:00
Ryan Dick
43108eec13 (minor) Add a TODO note to get_scheduler(...). 2024-06-25 11:31:52 -07:00
Ryan Dick
79ceac2f82 (minor) Use SilenceWarnings as a decorator rather than a context manager to save an indentation level. 2024-06-18 15:06:22 -04:00
Ryan Dick
d13aafb514 Tidy denoise_latents.py imports to all use absolute import paths. 2024-06-18 15:06:22 -04:00
Ryan Dick
785bb1d9e4 Fix all comparisons against the DEFAULT_PRECISION constant. DEFAULT_PRECISION is a torch.dtype. Previously, it was compared to a str in a number of places where it would always resolve to False. This is a bugfix that results in a change to the default behavior. In practice, this will not change the behavior for many users, because it only causes a change in behavior if a users has configured float32 as their default precision. 2024-06-14 11:26:10 -07:00
Lincoln Stein
7d19af2caa
Merge branch 'main' into lstein/feat/simple-mm2-api 2024-06-08 18:55:06 -04:00
Ryan Dick
52c0c4a32f Rename latent.py -> denoise_latents.py. 2024-06-07 09:28:42 -04:00
Ryan Dick
8f1afc032a Move SchedulerInvocation to a new file. No functional changes. 2024-06-07 09:28:42 -04:00
Ryan Dick
854bca668a Move CreateDenoiseMaskInvocation to its own file. No functional changes. 2024-06-07 09:28:42 -04:00
Ryan Dick
fea9013cad Move CreateGradientMaskInvocation to its own file. No functional changes. 2024-06-07 09:28:42 -04:00
Ryan Dick
045caddee1 Move LatentsToImageInvocation to its own file. No functional changes. 2024-06-07 09:28:42 -04:00
Ryan Dick
58697141bf Move ImageToLatentsInvocation to its own file. No functional changes. 2024-06-07 09:28:42 -04:00
Ryan Dick
5e419dbb56 Move ScaleLatentsInvocation and ResizeLatentsInvocation to their own file. No functional changes. 2024-06-07 09:28:42 -04:00
Ryan Dick
595096bdcf Move BlendLatentsInvocation to its own file. No functional changes. 2024-06-07 09:28:42 -04:00
Ryan Dick
ed03d281e6 Move CropLatentsCoreInvocation to its own file. No functional changes. 2024-06-07 09:28:42 -04:00
Ryan Dick
0b37496c57 Move IdealSizeInvocation to its own file. No functional changes. 2024-06-07 09:28:42 -04:00
psychedelicious
fde58ce0a3 Merge remote-tracking branch 'origin/main' into lstein/feat/simple-mm2-api 2024-06-07 14:23:41 +10:00
Lincoln Stein
dc134935c8 replace load_and_cache_model() with load_remote_model() and load_local_odel() 2024-06-07 14:12:16 +10:00
Lincoln Stein
2871676f79
LoRA patching optimization (#6439) 
* allow model patcher to optimize away the unpatching step when feasible

* remove lazy_offloading functionality

* allow model patcher to optimize away the unpatching step when feasible

* remove lazy_offloading functionality

* do not save original weights if there is a CPU copy of state dict

* Update invokeai/backend/model_manager/load/load_base.py

Co-authored-by: Ryan Dick <ryanjdick3@gmail.com>

* documentation fixes added during penultimate review

---------

Co-authored-by: Lincoln Stein <lstein@gmail.com>
Co-authored-by: Kent Keirsey <31807370+hipsterusername@users.noreply.github.com>
Co-authored-by: Ryan Dick <ryanjdick3@gmail.com>
 2024-06-06 13:53:35 +00:00
psychedelicious
14372e3818 fix(nodes): blend latents with weight=0 with DPMSolverSDEScheduler 
- Pass the seed from `latents_a` to the output latents. Fixed an issue where using `BlendLatentsInvocation` could result in different outputs during denoising even when the alpha or slerp weight was 0.

## Explanation

`LatentsField` has an optional `seed` field. During denoising, if this `seed` field is not present, we **fall back to 0 for the seed**. The seed is used during denoising in a few ways:

1. Initializing the scheduler.

The seed is used in two places in `invokeai/app/invocations/latent.py`.

The `get_scheduler()` utility function has special handling for `DPMSolverSDEScheduler`, which appears to need a seed for deterministic outputs.

`DenoiseLatentsInvocation.init_scheduler()` has special handling for schedulers that accept a generator - the generator needs to be seeded in a particular way. At the time of this commit, these are the Invoke-supported schedulers that need this seed:
  - DDIMScheduler
  - DDPMScheduler
  - DPMSolverMultistepScheduler
  - EulerAncestralDiscreteScheduler
  - EulerDiscreteScheduler
  - KDPM2AncestralDiscreteScheduler
  - LCMScheduler
  - TCDScheduler

2. Adding noise during inpainting.

If a mask is used for denoising, and we are not using an inpainting model, we add noise to the unmasked area. If, for some reason, we have a mask but no noise, the seed is used to add noise.

I wonder if we should instead assert that if a mask is provided, we also have noise.

This is done in `invokeai/backend/stable_diffusion/diffusers_pipeline.py` in `StableDiffusionGeneratorPipeline.latents_from_embeddings()`.

When we create noise to be used in denoising, we are expected to set `LatentsField.seed` to the seed used to create the noise. This introduces some awkwardness when we manipulate any "latents" that will be used for denoising. We have to pass the seed along for every operation.

If the wrong seed or no seed is passed along, we can get unexpected outputs during denoising. One notable case relates to blending latents (slerping tensors).

If we slerp two noise tensors (`LatentsField`s) _without_ passing along the seed from the source latents, when we denoise with a seed-dependent scheduler*, the schedulers use the fallback seed of 0 and we get the wrong output. This is most obvious when slerping with a weight of 0, in which case we expect the exact same output after denoising.

*It looks like only the DPMSolver* schedulers are affected, but I haven't tested all of them.

Passing the seed along in the output fixes this issue.
 2024-06-05 00:02:52 +10:00
Lincoln Stein
756108f6bd Update invokeai/app/invocations/latent.py 
Co-authored-by: Ryan Dick <ryanjdick3@gmail.com>
 2024-06-03 11:41:47 -07:00
Lincoln Stein
68d628dc14 use zip to iterate over image prompts and adapters 2024-06-03 11:41:47 -07:00
Lincoln Stein
93c9852142 fix ruff 2024-06-03 11:41:47 -07:00
Lincoln Stein
493f81788c added a few comments to document design choices 2024-06-03 11:41:47 -07:00
Lincoln Stein
f13427e3f4 refactor redundant code and fix typechecking errors 2024-06-03 11:41:47 -07:00
Lincoln Stein
e28737fc8b add check for congruence between # of ip_adapters and image_prompts 2024-06-03 11:41:47 -07:00
Lincoln Stein
7391c126d3 handle case of no IP adapters requested 2024-06-03 11:41:47 -07:00
Lincoln Stein
1c59fce6ad reduce peak VRAM memory usage of IP adapter 2024-06-03 11:41:47 -07:00
psychedelicious
521f907f58 tidy(nodes): infill 
- Set `self._context=context` instead of passing it as an arg
 2024-06-03 09:43:25 +10:00
psychedelicious
ccdecf21a3 tidy(nodes): cnet processors 
- Set `self._context=context` instead of changing the type signature of `run_processor`
- Tidy a few typing things
 2024-06-03 09:41:17 +10:00
Lincoln Stein
2276f327e5
Merge branch 'main' into lstein/feat/simple-mm2-api 2024-06-02 09:45:31 -04:00
psychedelicious
50d3030471 feat(app): dynamic type adapters for invocations & outputs 
Keep track of whether or not the typeadapter needs to be updated. Allows for dynamic invocation and output unions.
 2024-05-30 12:03:38 +10:00
psychedelicious
2f9ebdec69 fix(app): openapi schema generation 
Some tech debt related to dynamic pydantic schemas for invocations became problematic. Including the invocations and results in the event schemas was breaking pydantic's handling of ref schemas. I don't really understand why - I think it's a pydantic bug in a remote edge case that we are hitting.

After many failed attempts I landed on this implementation, which is actually much tidier than what was in there before.

- Create pydantic-enabled types for `AnyInvocation` and `AnyInvocationOutput` and use these in place of the janky dynamic unions. Actually, they are kinda the same, but better encapsulated. Use these in `Graph`, `GraphExecutionState`, `InvocationEventBase` and `InvocationCompleteEvent`.
- Revise the custom openapi function to work with the new models.
- Split out the custom openapi function to a separate file. Add a `post_transform` callback so consumers can customize the output schema.
- Update makefile scripts.
 2024-05-30 12:03:03 +10:00
Lincoln Stein
34e1eb19f9 merge with main and resolve conflicts 2024-05-27 22:20:34 -04:00
Lincoln Stein
532f82cb97
Optimize RAM to VRAM transfer (#6312) 
* avoid copying model back from cuda to cpu

* handle models that don't have state dicts

* add assertions that models need a `device()` method

* do not rely on torch.nn.Module having the device() method

* apply all patches after model is on the execution device

* fix model patching in latents too

* log patched tokenizer

* closes #6375

---------

Co-authored-by: Lincoln Stein <lstein@gmail.com>
 2024-05-24 17:06:09 +00:00
psychedelicious
281bd31db2 feat(nodes): make ModelIdentifierInvocation a prototype 2024-05-19 20:14:01 +10:00
psychedelicious
2680ef52c2 feat(nodes): add ModelIdentifierInvocation 
This node allows a user to select _any_ model, outputting a `ModelIdentifierField` for that model.
 2024-05-19 20:14:01 +10:00
psychedelicious
fe7ed72c9c feat(nodes): make all ModelIdentifierField inputs accept connections 2024-05-19 20:14:01 +10:00
Lincoln Stein
987ee704a1
Merge branch 'main' into lstein/feat/simple-mm2-api 2024-05-17 22:54:03 -04:00
Lincoln Stein
d968c6f379 refactor multifile download code 2024-05-17 22:29:19 -04:00
psychedelicious
ef89c7e537 feat(nodes): add LoRASelectorInvocation, LoRACollectionLoader, SDXLLoRACollectionLoader 
These simplify loading multiple LoRAs. Instead of requiring chained lora loader nodes, configure each LoRA (model & weight) with a selector, collect them, then send the collection to the collection loader to apply all of the LoRAs to the UNet/CLIP models.

The collection loaders accept a single lora or collection of loras.
 2024-05-15 14:09:44 +10:00
psychedelicious
93da75209c feat(nodes): use new blur_if_nsfw method 2024-05-14 07:23:38 +10:00
blessedcoolant
da61396b1c cleanup: seamless unused older code cleanup 2024-05-13 08:11:08 +10:00
Lincoln Stein
e9a20051bd refactor DWOpenPose and add type hints 2024-05-03 18:08:53 -04:00
Lincoln Stein
38df6f3702 fix ruff error 2024-05-02 21:22:33 -04:00
Lincoln Stein
3b64e7a1fd
Merge branch 'main' into lstein/feat/simple-mm2-api 2024-05-02 21:20:35 -04:00
psychedelicious
33a9f9a4dc fix(nodes): fix constraints in cnet processors 
There were some invalid constraints with the processors - minimum of 0 for resolution or multiple of 64 for resolution.

Made minimum 1px and no multiple ofs.
 2024-05-02 12:24:04 +10:00
blessedcoolant
dce8b88aaf fix: change eta only for TCD Scheduler 2024-05-01 12:47:46 +05:30
blessedcoolant
1bdcbe3284 cleanup: use dict update to actually update the scheduler keyword args 2024-05-01 12:22:39 +05:30
Lincoln Stein
49c84cd423
Merge branch 'main' into lstein/feat/simple-mm2-api 2024-04-30 18:13:42 -04:00
blessedcoolant
2ddb82200c fix: Manually update eta(gamma) to 1.0 for TCDScheduler 
seems to work best with invoke at 4 steps
 2024-05-01 01:20:53 +05:30
psychedelicious
e822897b1c feat(nodes): add prototype heuristic image resize node 
Uses the fancy cnet resize that retains edges.
 2024-04-30 08:10:59 -04:00
psychedelicious
1fe90c357c feat(backend): lift managed model loading out of depthanything class 2024-04-29 08:56:00 +10:00
psychedelicious
fcb071f30c feat(backend): lift managed model loading out of lama class 2024-04-29 08:12:51 +10:00
Lincoln Stein
f65c7e2bfd
Merge branch 'main' into lstein/feat/simple-mm2-api 2024-04-28 13:42:54 -04:00
dunkeroni
f262b9032d fix: changed validation to not error on connection 2024-04-28 12:48:56 -04:00
dunkeroni
71c3197eab fix: denoise latents accepts CFG lists as input 2024-04-28 12:48:56 -04:00
Lincoln Stein
bb04f496e0 Merge branch 'main' into lstein/feat/simple-mm2-api 2024-04-28 11:33:26 -04:00
Lincoln Stein
70903ef057 refactor load_ckpt_from_url() 2024-04-28 11:33:23 -04:00