Generic feed forward neural network

Is this error screenshot correct? Looks different from the one you shared in other post.

yes it is the correct error message

I inserted a debug statement where error is occuring:

The output shows:

debug: (1500, 1) (2, 2) (1, 2)

Dimension of H (1500, 1)and W (2, 2) are not compatible for matrix multiplication. You will need to check your code to see what should be the correct dimension and initialise appropriately.

i dont get it, it was working properly for the previous data

i cant figure out what should be done to change the shape of self.W[i+1]

Screenshot 2021-01-12 at 7.58.23 AM
Do you want to call self.grad with X, Y or x, y ?

Which video is it? Can you compare your code with the code provided at the end of video lectures in download section?

i can find the downloads section where is it exactly

what does this error mean and how do i correct it?

Every section of related videos has a download section at the end where the code and other helpful files are available for download.
. You can compare your code with the code in download section for quick resolution.

. Also, whenever there is an error, there is SEARCH STACK OVERFLOW button (see the error screenshot above). You should also use that to get some idea about the error (its a popular way of debugging errors).

i did try it out couldnt really understand much

also can you send the link to the page cuz your interface is quite different from what im using

this the site im using coudnt find the download section here

@Ishvinder
Can you point out how to download code from Guvi Deep Learning Course?

Sure @Sanjayk,
Hi @christy_martin,
You can refer to Summary/ Description Tab in the last few viedos of that module to look for download materials.
In case you find it difficult, please refer to the drive links shared in FAQs DL Course

1 Like