2098 Commits

Author	SHA1	Message	Date
Ryan Dick	faee79dc95	Distinguish between restricted and unrestricted attn masks in FLUX regional prompting.	2024-11-26 16:55:52 +00:00
Ryan Dick	e01f66b026	Apply regional attention masks in the single stream blocks in addition to the double stream blocks.	2024-11-25 22:40:08 +00:00
Ryan Dick	53abdde242	Update Flux RegionalPromptingExtension to prepare both a mask with restricted image self-attention and a mask with unrestricted image self attention.	2024-11-25 22:04:23 +00:00
Ryan Dick	94c088300f	Be smarter about selecting the global CLIP embedding for FLUX regional prompting.	2024-11-25 20:15:04 +00:00
Ryan Dick	3741a6f5e0	Fix device handling for regional masks and apply the attention mask in the FLUX double stream block.	2024-11-25 16:02:03 +00:00
Ryan Dick	2c23b8414c	Use a single global CLIP embedding for FLUX regional guidance.	2024-11-22 23:01:43 +00:00
Ryan Dick	20356c0746	Fixup the logic for preparing FLUX regional prompt attention masks.	2024-11-21 22:46:25 +00:00
Ryan Dick	bad1149504	WIP - add rough logic for preparing the FLUX regional prompting attention mask.	2024-11-20 22:29:36 +00:00
Ryan Dick	fda7aaa7ca	Pass RegionalPromptingExtension down to the CustomDoubleStreamBlockProcessor in FLUX.	2024-11-20 19:48:04 +00:00
Ryan Dick	85c616fa34	WIP - Pass prompt masks to FLUX model during denoising.	2024-11-20 18:51:43 +00:00
Mary Hipp	1948ffe106	make sure Soft Edge Detection has preprocessor applied	2024-11-20 08:46:02 -05:00
Brandon Rising	a69c5ff9ef	Add copyright notice for CIELab_to_UPLab.icc	2024-11-19 07:02:37 -08:00
Brandon Rising	1fd80d54a4	Run Ruff	2024-11-19 07:02:37 -08:00
Brandon Rising	991f63e455	Store CIELab_to_UPLab.icc within the repo	2024-11-19 07:02:37 -08:00
Brandon Rising	0eadc0dd9e	feat: Support a subset of composition nodes within base invokeai	2024-11-19 07:02:37 -08:00
Ryan Dick	1a0885bfb1	Update FLUX IP-Adapter starter model from XLabs v1 to XLabs v2.	2024-11-18 17:06:53 -08:00
Ryan Dick	c6fc82f756	Infer the clip_extra_context_tokens param from the state dict for FLUX XLabs IP-Adapter V2 models.	2024-11-18 17:06:53 -08:00
Ryan Dick	1cf7600f5b	Merge branch 'main' into ryan/sd3-image-to-image	2024-11-14 09:25:23 -08:00
Ryan Dick	50897ba066	Add flag to optionally allow missing layer keys in FLUX lora loader.	2024-11-14 16:53:49 +00:00
Ryan Dick	3510643870	Support FLUX LoRAs without .proj_mlp layers.	2024-11-14 16:53:49 +00:00
Ryan Dick	b89caa02bd	Merge branch 'main' into flux_vae_fp16_broke	2024-11-13 15:33:43 -08:00
Darrell	fb19621361	Updated link to flux ip adapter model	2024-11-12 08:11:40 -05:00
Mary Hipp	13cb5f0ba2	Merge remote-tracking branch 'origin/main' into ryan/sd3-image-to-image	2024-11-08 20:29:56 -05:00
Vargol	83609d5008	Merge branch 'invoke-ai:main' into flux_vae_fp16_broke	2024-11-08 10:37:31 +00:00
David Burnett	bb3cedddd5	Rework change based on comments	2024-11-08 10:27:47 +00:00
Ryan Dick	0e6cb91863	Update SD3 InpaintExtension with gradient adjustment to match FLUX.	2024-11-07 22:55:30 +00:00
Ryan Dick	a0fefcd43f	Switch to using a custom scheduler implementation for SD3 rather than the diffusers FlowMatchEulerDiscreteScheduler. It is easier to work with and enables us to re-use the clip_timestep_schedule_fractional() utility from FLUX.	2024-11-07 22:46:52 +00:00
psychedelicious	29d63d5dea	fix(app): silence pydantic protected namespace warning ... Closes #7287	2024-11-08 07:36:50 +11:00
Ryan Dick	a5f8c23dee	Add inpainting support for SD3.	2024-11-07 20:21:43 +00:00
Brandon Rising	f4b0b6a93d	fix: Look in known subfolders for configs for clip variants	2024-11-07 14:02:20 +00:00
Brandon Rising	1337c33ad3	fix: Avoid downloading unsafe .bin files if a safetensors file is available	2024-11-07 14:02:20 +00:00
Brandon Rising	c064257759	fix: Look in known subfolders for configs for clip variants	2024-11-07 12:01:02 +11:00
Brandon Rising	8a4c629576	fix: Avoid downloading unsafe .bin files if a safetensors file is available	2024-11-06 19:31:18 -05:00
David Burnett	496b02a3bc	Same issue affects image2image, so do the same again	2024-11-06 17:47:22 -05:00
David Burnett	7b5efc2203	Flux Vae broke for float16, force bfloat16 or float32 were compatible	2024-11-06 17:47:22 -05:00
Brandon Rising	faa9ac4e15	fix: get_clip_variant_type should never return None	2024-11-06 09:59:50 +11:00
Brandon Rising	d07a6385ab	Always default to ClipVariantType.L instead of None	2024-11-05 12:03:40 -05:00
Brandon Rising	68df612fa1	fix: Never throw an exception when finding the clip variant type	2024-11-05 12:03:40 -05:00
Brandon Rising	22bff1fb22	Fix conditional within filter_by_variant to not read all candidates as default	2024-11-04 12:42:09 -05:00
Brandon Rising	391a24b0f6	Re-add erroniously removed hash code	2024-11-04 12:42:09 -05:00
Brandon Rising	e75903389f	Run ruff, fix bug in hf downloading code which failed to download parts of a model	2024-11-04 12:42:09 -05:00
Brandon Rising	6f447f7169	Rather than .fp16., some repos start the suffix with .fp16... for weights spread across multiple files	2024-11-04 12:42:09 -05:00
maryhipp	af583d2971	ruff format	2024-11-04 12:42:09 -05:00
maryhipp	874d67cb37	add SD3.5 to starter models	2024-11-04 12:42:09 -05:00
maryhipp	78bd605617	(nodes,api): expose the submodels on SD3 model loader as optional, add types needed for CLIP-L and CLIP-G fields	2024-11-04 12:42:09 -05:00
Brandon Rising	b87f4e59a5	Create clip variant type, create new fucntions for discerning clipL and clipG in the frontend	2024-11-04 12:42:09 -05:00
Brandon Rising	ebabf4f7a8	Setup Model and T5 Encoder selection fields for sd3 nodes	2024-11-04 12:42:09 -05:00
Brandon Rising	3970aa30fb	define submodels on sd3 models during probe	2024-11-04 12:42:09 -05:00
Ryan Dick	c620581699	Bug fixes to get SD3 text-to-image workflow running.	2024-11-04 12:42:09 -05:00
Ryan Dick	2d86298b7f	Add first draft of Sd3TextEncoderInvocation.	2024-11-04 12:42:09 -05:00