RuntimeError: The size of tensor a (9782) must match the size of tensor b (7680) at non-singleton dimension 1

[wclemonade]

debugpy 1.8.20
decorator 5.2.1
diffusers 0.38.0.dev0
distro 1.9.0
tokenizers 0.22.2
torch 2.6.0+cu124
torchvision 0.21.0+cu124
tornado 6.5.5
tqdm 4.67.3
traitlets 5.14.3
transformers 5.6.0.dev0

File "/home/tianye/tianye/code/AIImageOpensource/GLM-Image/test.py", line 9, in
image = pipe(
^^^^^
File "/home/tianye/.conda/envs/omnigen2/lib/python3.11/site-packages/torch/utils/_contextlib.py", line 116, in decorate_context
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/home/tianye/.conda/envs/omnigen2/lib/python3.11/site-packages/diffusers/pipelines/glm_image/pipeline_glm_image.py", line 906, in call
_ = self.transformer(
^^^^^^^^^^^^^^^^^
File "/home/tianye/.conda/envs/omnigen2/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1739, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/tianye/.conda/envs/omnigen2/lib/python3.11/site-packages/torch/nn/modules/module.py", line 1750, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/tianye/.conda/envs/omnigen2/lib/python3.11aa/site-packages/diffusers/models/transformers/transformer_glm_image.py", line 629, in forward
hidden_states = hidden_states + prior_hidden_states
~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~
RuntimeError: The size of tensor a (9782) must match the size of tensor b (7680) at non-singleton dimension 1

[JaredforReal]

can u show us the details of your test.py?

[wclemonade]

can u show us the details of your test.py?

[wclemonade]

T2I pipeline runs normally, but the I2I pipeline reports the error mentioned above.

[wclemonade]

"The width and height must be within the range of 512px to 2048px, and both width and height must be integer multiples of 32 ", as mentioned in your API doc, resizing the image according to the above rules will resolve the error.