Hi Team,
Can you please check below issue, I think a still not understand Con and even not able to debug is there any possibility to run this task in my Laptop using VScode ?if Yes please guiding us how to download the whole project and make it work in Local environment
Dear @Mohamad_Omar,
Click on Lab files icon (top-right), Select the folder/files then click Download all files
–
Keep Learning AI with DeepLearning.AI - Girijesh
Installing all of the correct library versions on your local platform is no easy task.
Please check your personal messages for instructions.
Correct but we will do it once for all
please sort out all the libraries need and their versions ?
ex:Pyton → which version ?
because is not easy !!
Sorry, I’m not able to do that.
The other alternative would be to run a debugger in the Jupyter notebook. I think it is possible to do that.
But the key is that writing the code is just an exercise in frustration unless you understand the algorithm, meaning what the code is supposed to actually be doing. My guess is that your problem is on that level. Debugging locally won’t help if you don’t have a clear picture of what is supposed to be happening.
Fortunately this issue has come up before. Please have a look at this thread and I hope it will give you a better picture of what it is the code needs to do.
There are no official instructions for running locally. That is outside the scope of the course. But with that said, there have been some valuable threads from fellow students who have explored this space. E.g. here’s one that will get you started down that road and explain things exactly like the question you are asking: how to figure out the versions of everything.
But as I commented above, I think this is wasted effort at this point. The real problem is you need to understand how the algorithm works and a debugger will not help with that. And if you need to debug, the “old fashioned” method of adding print statements is the quickest and easiest way to see what is happening in your code.
please your support to guide me in how to activate debugger because i can’t seeing in conv_forward function how the defined filters walk/pass inside all image matrix .and why we have 4D array of Tensor..I read a little bit using chat GPT is this is normal when we dealing with CNN ?because it’s very hard to know how many axis we have and visualize it in real world .
For that reason i want to simulate all the batches in their images in VScode and explore the filters shapes and how the for loop pass trough images ,these things can’t achieved in notebook or maybe we can do it if you support us how to activate debugger.As you know this exercise is very important for the career and not good to only understand the algorithm , i think is better to memorize in our brain for later use.
For the Shared thread i will give it a look..
The shared link explained in how to make the project working in Pycharm, for me i have VScode
two different approach .
Yes, the right first step here is to read that thread that I linked above. That should give you a much clearer picture of what needs to happen here.
I used Pdb in jupyter and i can see what’s going on with little bit support from gpt .
now i will continue debugging and digging in code more and more .i understood the reasons for nested loops, one for batching of images second/third for vertical/horizontal image slicing and last one for numbers of filters .
we have in output: 8 filters * 4 channels * 3 height *3 width .
Thanks
