ERROR with ReLU, inplace_unfold cause the bad shape.

[charnameRed]

Describe the bug
I tried to use auto_LiRPA to verify the robustness of my audio processing model. Unfortunately, when I use model.compute_bounds, a RuntimeError: The following operation failed in the TorchScript interpreter. occurred.

The situation directly caused by the unfitted shape of A_pos([64, 4, 25, 20, 64, 9, 9]) and d_pos([1, 4, 24, 19, 64, 9, 9]) in A_new = d_pos * A_pos + d_neg * A_neg, which happens during clamp_mutiply_forward operation. I tried to trace back the cause of this difference, and find something happened in the inplace_unfold step of maybe_unfold_patches, in the relu.py.

Here follows my code. I make a emulation dataset to provide random data with the same attributes of real data, which are Mel Spectrograms of some audio, to avoid unnecessary uploadings. And I used an adjusted Resnet18 to extract the features, to make it work correctly on [B, 1, 100, 80] shaped data.

To Reproduce

Test Code
Adjusted Resnet

System configuration:

• OS: Ubuntu 22.04
• Python version: Python 3.10
• Pytorch Version: Pytorch 2.2
• Hardware: RTX 4090
• Have you tried to reproduce the problem in a cleanly created conda/virtualenv environment using official installation instructions and the latest code on the main branch?: Yes

[shizhouxing]

Hi @charnameRed, could you please post your complete error message?

[charnameRed]

THX for replying. The error message is as follows:

The following operation failed in the TorchScript interpreter.
Traceback of TorchScript (most recent call last):
RuntimeError: The following operation failed in the TorchScript interpreter.
Traceback of TorchScript (most recent call last):
  File "/home/cnred/anaconda3/envs/realForen/lib/python3.11/site-packages/auto_LiRPA/operators/clampmult.py", line 44, in <fused code>
            A_new = d_pos * A
        else:  
            A_new = d_pos * A_pos + d_neg * A_neg
                    ~~~~~~~~~~~~~ <--- HERE
        
        bias_pos = bias_neg = torch.zeros(
RuntimeError: The size of tensor a (19) must match the size of tensor b (20) at non-singleton dimension 3

  File "/home/cnred/projects/RealForensics/stage2/chR/test.py", line 87, in <module>
    lb_a,ub_a = a_submodel.compute_bounds(x=(aud_ptbd,))
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: The following operation failed in the TorchScript interpreter.
Traceback of TorchScript (most recent call last):
RuntimeError: The following operation failed in the TorchScript interpreter.
Traceback of TorchScript (most recent call last):
  File "/home/cnred/anaconda3/envs/realForen/lib/python3.11/site-packages/auto_LiRPA/operators/clampmult.py", line 44, in <fused code>
            A_new = d_pos * A
        else:  
            A_new = d_pos * A_pos + d_neg * A_neg
                    ~~~~~~~~~~~~~ <--- HERE
        
        bias_pos = bias_neg = torch.zeros(
RuntimeError: The size of tensor a (19) must match the size of tensor b (20) at non-singleton dimension 3