Failed to load GSM8K dataset in M1_G1_Inspecting_Finetuned_vs_Base_Model

In the lab “M1_G1_Inspecting_Finetuned_vs_Base_Model“, the gsm8k data cannot be loaded sucessfully.

This is not the first time to reach error in this lab, previously the whole lab was not loaded, now the data is not loaded. Appericate if the platform can be managed a bit more stable, thanks.

For what it’s worth, I am having the same issue. It seems that the Dataset metadata was loaded instead of the Dataset itself, and the notebook prevents you from pulling the source from HuggingFace.

Exactly, the connection to internet is blocked, so we cannot directly load from public source.

is this issue occuring on platform? or locally?

based on recent updates there is some issue going on with the labs in deeplearning.ai learning platform.

Can you confirm if your kernel remains connected when this error appears?

I was doing the lab directly on the deeplearning.ai platform with kernal connected.

@mystho

looks like there is server error, staff had notified about this that labs are not opening. they probably must be working on resolving the issue.

please try after sometime or wait for my response because when I am trying to run this lab I am getting 502 kernel response.

I will also notify staff about your issue, thank you for informing.

I have notified the staff.

1 Like

I am seeing same issue currently - not able to proceed with the lab.

That’s because it hasn’t been fixed yet.

Sorry you had the issue. It is solved now. thanks for reporting it.

thanks for the support