By: sina.amirrajab on Aug. 8, 2025, 1:03 p.m.
My docker submission failed with the error:
The algorithm failed on one or more cases.
How can I get the log? I tested the docker and works locally. Thanks and regards, Sina
By: vlm3dchallenge on Aug. 11, 2025, 9:04 a.m.
Hi @sina.amirrajab,
Your image has a precision issue:
Text Conditional MAISI Inference: 5%|▌ | 5/100 [05:15<1:39:52, 63.08s/it] Traceback (most recent call last): File "/opt/app/process.py", line 263, in <module> main() File "/opt/app/process.py", line 211, in main data = run_inference( ^^^^^^^^^^^^^^ File "/opt/app/TextMAISI/model_utils.py", line 175, in run_inference raise e File "/opt/app/TextMAISI/model_utils.py", line 142, in run_inference model_output = unet(**unet_inputs) ^^^^^^^^^^^^^^^^^^^ File "/output/venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1739, in _wrapped_call_impl return self._call_impl(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/output/venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1750, in _call_impl return forward_call(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/output/venv/lib/python3.12/site-packages/monai/apps/generation/maisi/networks/diffusion_model_unet_maisi.py", line 400, in forward h, _updated_down_block_res_samples = self._apply_down_blocks(h, emb, context, down_block_additional_residuals) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/output/venv/lib/python3.12/site-packages/monai/apps/generation/maisi/networks/diffusion_model_unet_maisi.py", line 344, in _apply_down_blocks h, res_samples = downsample_block(hidden_states=h, temb=emb, context=context) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/output/venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1739, in _wrapped_call_impl return self._call_impl(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/output/venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1750, in _call_impl return forward_call(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/output/venv/lib/python3.12/site-packages/monai/networks/nets/diffusion_model_unet.py", line 764, in forward hidden_states = attn(hidden_states, context=context).contiguous() ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/output/venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1739, in _wrapped_call_impl return self._call_impl(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/output/venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1750, in _call_impl return forward_call(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/output/venv/lib/python3.12/site-packages/monai/networks/nets/diffusion_model_unet.py", line 232, in forward x = block(x, context=context) ^^^^^^^^^^^^^^^^^^^^^^^^^ File "/output/venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1739, in _wrapped_call_impl return self._call_impl(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/output/venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1750, in _call_impl return forward_call(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/output/venv/lib/python3.12/site-packages/monai/networks/nets/diffusion_model_unet.py", line 120, in forward x = self.attn2(self.norm2(x), context=context) + x ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/output/venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1739, in _wrapped_call_impl return self._call_impl(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/output/venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1750, in _call_impl return forward_call(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/output/venv/lib/python3.12/site-packages/monai/networks/blocks/crossattention.py", line 152, in forward k = self.input_rearrange(self.to_k(kv)) ^^^^^^^^^^^^^ File "/output/venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1739, in _wrapped_call_impl return self._call_impl(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/output/venv/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1750, in _call_impl return forward_call(*args, **kwargs) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/output/venv/lib/python3.12/site-packages/torch/nn/modules/linear.py", line 125, in forward return F.linear(input, self.weight, self.bias) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ RuntimeError: expected mat1 and mat2 to have the same dtype, but got: double != c10::Half
Could you please fix the issue and resend your image?
Best, Sezgin
By: sina.amirrajab on Aug. 12, 2025, 11:11 a.m.
Thanks for sharing the log. I fixed the issue and submitted a new Docker image, which is queued now.
