vikhyatk
/

moondream2

Image-Text-to-Text

text-generation

Inference Endpoints

Model card Files Files and versions Community

Aspect ratio bug

#29

by sachin - opened Jul 17

base: refs/heads/main

←

from: refs/pr/29

Discussion Files changed

Files changed (1) hide show

vision_encoder.py +1 -1

vision_encoder.py CHANGED Viewed

@@ -233,7 +233,7 @@ class VisionEncoder(nn.Module):
             im_size = min(
                 self.supported_sizes,
                 key=lambda size: (
-                    abs((size[1] / size[0]) - aspect_ratio),
                     abs(size[0] - width) + abs(size[1] - height),
                 ),
             )

             im_size = min(
                 self.supported_sizes,
                 key=lambda size: (
+                    abs((size[0] / size[1]) - aspect_ratio),
                     abs(size[0] - width) + abs(size[1] - height),
                 ),
             )