Lstm_backward(da, caches)

ValueError: could not broadcast input array from shape (5,10) into shape (3,10)

dxt gradient error

please see the attachment, what I feel there might be issue with the da_next but I am not understanding if I need to go back in lstm and do the correction or in this backward propagation.

sorry I got the answer and mistake where I did. dx was not right, I missed a part of instruction. I don’t know how to delete this topic here now :frowning:

1 Like

Hi @Deepti_Prasad,

Thank you for letting us know you are able to fix the problem yourself. I will close this topic for you.

Cheers,
Raymond

1 Like

Hello @Deepti_Prasad!

It’s great to know that you solved your problem by yourself. However, I would like to highlight just one thing. You posted this in DLS Course 3 but I guess this query belongs to DLS Course 5. For now, it may not be necessary to change the post category but in the future, please post your query in the correct course category. So, you can get immediate responses from mentors.

Read this post regarding the course category.

Best,
Saif.