tensorflow / tfjs

tfjs/tfjs-layers/src/layers/convolutional.ts

Line 108 in 63250ec

`${kernel.shape.length} instead`);

Shouldn't bias.shape.length be output as an error message here?

Thank you for bringing this issue to our attention and as per my current understanding, The if statement checks if bias is not null and its shape.length (number of dimensions) is not equal to 1. This is the correct check to ensure a valid bias tensor for a 1D convolution.However, the error message incorrectly references kernel.shape.length when it should be referencing bias.shape.length.

The following error message would be more accurate and clearly indicates that the issue lies with the number of dimensions in the bias tensor, not the kernel.

if (bias != null && bias.shape.length !== 1) {
      throw new ValueError(
          `The bias for a conv1dWithBias operation should be 1, but is ` +
          `${bias.shape.length} instead`);
    }

Would you like to submit a PR to take care of this issue from your end now so our developer team will review your PR and will take appropriate action from their end ? if not I'll go ahead and submit a PR from my end to take care of this error message issue.

Thank you for your cooperation and patience.

Could you please submit a PR from your end?

Alright, I'll go ahead submit PR to take care of this issue. Thank you.

Hi, @set5u

I have submitted a PR #8248 and this issue will be taken care once that PR got merged. Thank you.

Hi, @set5u

I see that Pull Request #8248 has now been merged, which addresses the issue reported here. Therefore, I'm closing this ticket. Should you encounter any further problems or require additional assistance, please don't hesitate to create a new issue. We're always happy to help!

Thank you for your cooperation and patience.

Are you satisfied with the resolution of your issue?
Yes
No

The if statement refers to `bias.shape.length`, but the error message refers to `kernel.shape.length`