Trouble importing Dataset from MLDataset using VScode

New to Julia
,
1

I am trying to use the CIFAR10 Dataset using the MLDataset pakage of Julia, but I confirm for confirmation to download, the REPL gets stuck(i guess), and via VS code notifications I just see
this (attached below)

Screenshot 2024-12-08 at 3.27.36 PM
Screenshot 2024-12-08 at 3.27.36 PM2336×586 82.7 KB

what could be the source for it taking such a long time , 45 mins till now,? Is it related to setup of julia in vscode , or something else?
Thank you for the help

2

Hi, and welcome to the Julia community!

I tested downloading CIFAR10 from within VS Code and encountered no issues. There was no output until everything finished, at which point I got

7-Zip 21.07 (x64) : Copyright (c) 1999-2021 Igor Pavlov : 2021-12-26


Extracting archive:
--
Path =
Type = tar
Code Page = UTF-8
Characteristics = ASCII

Everything is Ok

Folders: 1
Files: 8
Size:       184380149
Compressed: 9216
dataset CIFAR10:
  metadata  =>    Dict{String, Any} with 2 entries
  split     =>    :train
  features  =>    32×32×3×50000 Array{Float32, 4}
  targets   =>    50000-element Vector{Int64}

Some questions:

  • Can you post some more info about your system? I.e. what is your versioninfo, the version of MLDatasets.jl (via ] status), the version of VS Code and the Julia plugin, your OS, …?
  • Do you get the same problems when you work in a non-VS code terminal?
  • Have you tried different download directories (CIFAR10(dir=...))?
  • If you monitor your network traffic, is it downloading?
    • (For me the download speed was around 1 MB/s.)
  • Are you able to download the data directly (https://www.cs.toronto.edu/~kriz/cifar-10-binary.tar.gz)?