Add FLUX image-to-image and inpainting by RyanJDick · Pull Request #6798 · invoke-ai/InvokeAI

Add FLUX VAE support to ImageToLatentsInvocation.

RyanJDick

committed

Add FLUX VAE decoding support to LatentsToImageInvocation.

RyanJDick

committed

Get FLUX non-masked image-to-image working - still rough.

RyanJDick

committed

Split VAE decoding out from the FLUXTextToImageInvocation.

RyanJDick

committed

Bump FLUX node versions after splitting out VAE encode/decode.

RyanJDick

committed

Remove unused vae field from FLUXTextToImageInvocation.

RyanJDick

committed

Update MaskTensorToImageInvocation to support input mask tensors with or without a channel dimension.

RyanJDick

committed

Get a rough version of FLUX inpainting working.

RyanJDick

committed

Split FLUX VAE encoding out into its own node from ImageToLatentsInvocation.

RyanJDick

committed

Split FLUX VAE decoding out into its own node from LatentsToImageInvocation.

RyanJDick

committed

Code cleanup and documentation around FLUX inpainting.

RyanJDick

committed

Use the existence of initial latents to decide whether we are doing image-to-image in the FLUX denoising node. Previously we were using the denoising_start value, but in some cases with an inpainti…

RyanJDick

committed

Rename FluxTextToImageInvocation -> FluxDenoiseInvocation.

RyanJDick

committed

Add denoise_end param to FluxDenoiseInvocation.

RyanJDick

committed

Rename flux_text_to_image.py -> flex_denoise.py

RyanJDick

committed

Update default workflows for FLUX.

RyanJDick

committed

Update schema.ts.

RyanJDick

committed

Restore line that was accidentally removed during development.

RyanJDick

committed

Rename 'FLUX VAE Encode'/'FLUX VAE Decode' to 'FLUX Image to Latents'/'FLUX Latents to Image'

RyanJDick

committed

Expose all FLUX model params in the default FLUX models.

RyanJDick

committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Add FLUX image-to-image and inpainting#6798

Add FLUX image-to-image and inpainting#6798
RyanJDick merged 20 commits intomainfrom
ryan/flux-split-vae

Commits on Sep 2, 2024

Add FLUX VAE support to ImageToLatentsInvocation.

Add FLUX VAE decoding support to LatentsToImageInvocation.

Get FLUX non-masked image-to-image working - still rough.

Split VAE decoding out from the FLUXTextToImageInvocation.

Bump FLUX node versions after splitting out VAE encode/decode.

Remove unused vae field from FLUXTextToImageInvocation.

Update MaskTensorToImageInvocation to support input mask tensors with or without a channel dimension.

Get a rough version of FLUX inpainting working.

Split FLUX VAE encoding out into its own node from ImageToLatentsInvocation.

Split FLUX VAE decoding out into its own node from LatentsToImageInvocation.

Code cleanup and documentation around FLUX inpainting.

Use the existence of initial latents to decide whether we are doing image-to-image in the FLUX denoising node. Previously we were using the denoising_start value, but in some cases with an inpainti…

Rename FluxTextToImageInvocation -> FluxDenoiseInvocation.

Add denoise_end param to FluxDenoiseInvocation.

Rename flux_text_to_image.py -> flex_denoise.py

Update default workflows for FLUX.

Update schema.ts.

Restore line that was accidentally removed during development.

Rename 'FLUX VAE Encode'/'FLUX VAE Decode' to 'FLUX Image to Latents'/'FLUX Latents to Image'

Expose all FLUX model params in the default FLUX models.

Add FLUX image-to-image and inpainting#6798RyanJDick merged 20 commits intomaininvoke-ai/InvokeAI:mainfrom ryan/flux-split-vaeinvoke-ai/InvokeAI:ryan/flux-split-vaeCopy head branch name to clipboard

Commits

Commits on Sep 2, 2024

Add FLUX image-to-image and inpainting#6798
RyanJDick merged 20 commits intomainfrom
ryan/flux-split-vae