Fix VAEEncodeForInpaint to support WAN VAE tuple downscale_ratio (#11572)

Use vae.spacial_compression_encode() instead of directly accessing
downscale_ratio to handle both standard VAEs (int) and WAN VAEs (tuple).

Addresses reviewer feedback on PR #11259.

Co-authored-by: ChrisFab16 <christopher@fabritius.dk>
This commit is contained in:
rattus 2026-01-08 20:34:48 -08:00 committed by GitHub
parent 1dc3da6314
commit 6207f86c18
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194

View File

@ -378,14 +378,15 @@ class VAEEncodeForInpaint:
CATEGORY = "latent/inpaint"
def encode(self, vae, pixels, mask, grow_mask_by=6):
x = (pixels.shape[1] // vae.downscale_ratio) * vae.downscale_ratio
y = (pixels.shape[2] // vae.downscale_ratio) * vae.downscale_ratio
downscale_ratio = vae.spacial_compression_encode()
x = (pixels.shape[1] // downscale_ratio) * downscale_ratio
y = (pixels.shape[2] // downscale_ratio) * downscale_ratio
mask = torch.nn.functional.interpolate(mask.reshape((-1, 1, mask.shape[-2], mask.shape[-1])), size=(pixels.shape[1], pixels.shape[2]), mode="bilinear")
pixels = pixels.clone()
if pixels.shape[1] != x or pixels.shape[2] != y:
x_offset = (pixels.shape[1] % vae.downscale_ratio) // 2
y_offset = (pixels.shape[2] % vae.downscale_ratio) // 2
x_offset = (pixels.shape[1] % downscale_ratio) // 2
y_offset = (pixels.shape[2] % downscale_ratio) // 2
pixels = pixels[:,x_offset:x + x_offset, y_offset:y + y_offset,:]
mask = mask[:,:,x_offset:x + x_offset, y_offset:y + y_offset]