Cannot progress past # EXERCISE: Split the dataset …
without the message:
The kernel appears to have died. It will restart automatically. I keep getting dead kernel message.
THere is a line of code that splits 1 dataset into 3 using tfds.load
& split on ‘cats_vs_dogs:4..’
Any thoughts, or tips?
Many thanks
C
if it is not grade cell codes then post the screenshot here or DM the screenshot which is causing kernel to die down.
Are you sure it is completely disconnecting or get disconnected and reconnects??
Also there is some issue going through Coursera lab for some courses, although none reported till now for the course you are working, so I am not sure if that is also the reason.
Regards
DP
on a different PC, i have the identical issue:
The kernel appears to have died. It will restart automatically.
I cannot progress past cell 3 - which splits the cats_vs_dogs dataset into train_examples, validation_examples, test_examples.
Its likely a code issue, but impossible to troubleshoot as there is always a kernel crash. I will DM you the code cell. Many thanks 
I need to also see the complete error. so post the error here and codes by personal DM.
You will get through!!
you probably going to remember this specialisation the most 
as I remember the toughest specialisation I did the most.
If the kernel is dying automatically then surely there is an issue, coursera lab did have a issue litter while ago but even what I remember in cats vs dog, whenever people were incorrectly writing codes to the directory or hard coding the path, those also resulted in the same issue.
Sincerely I recommend you to get again a fresh copy first, then only write codes where it is ask to do so, then work upon. the type(init) op and message error usually means the initiation to the code given is throwing error. it can be also cell#3 doesn’t mean the cell you are looking the error into.
Also if you writing codes in vs code and then copy pasting in coursera environment, then that also could cause such issue.
Ok now addressing your week 2 assignment.
First below code
filePath = f"{getcwd()}/data"
is not mention in exercise 2 split the dataset, but is mentioned in 1. Load the mobilnet model and it’s features.
by any chance if you have don’t too much editing, then re-start by getting a fresh copy.
Then mentions codes only where it is mentioned ###YOUR CODE HERE###
in the split dataset, your first code line is incorrect, recheck again. It is almost correct, one of the dataset split is incorrect it should [-10% and -5%]
then where you load the split and info, don’t use supervised=true argument
please update me the progress, so we can shift to your other post.