Skip to content

The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface. The fastest local inference engine in the world.

Fix potential issue with batch size and clip vision.

comfyanonymousae270f79bcb0
diff --git a/comfy/clip_vision.py b/comfy/clip_vision.py
index 8d04faf71..c408a732d 100644
--- a/comfy/clip_vision.py
+++ b/comfy/clip_vision.py
@@ -25,6 +25,8 @@ class ClipVisionModel():
def encode_image(self, image):
img = torch.clip((255. * image), 0, 255).round().int()
+ if len(img.shape) == 3:
+ img = [img]
inputs = self.processor(images=img, return_tensors="pt")
outputs = self.model(**inputs)
return outputs