Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

MNIST dataset download failes with 503 using torchvision 0.8.2 #298

Closed
Imipenem opened this issue Mar 6, 2021 · 8 comments · Fixed by #305 or #321
Closed

MNIST dataset download failes with 503 using torchvision 0.8.2 #298

Imipenem opened this issue Mar 6, 2021 · 8 comments · Fixed by #305 or #321
Assignees
@Imipenem
Labels
bug Something isn't working high-priority
Milestone
1.10.0

Comments

@Imipenem
Copy link
Contributor

Imipenem commented Mar 6, 2021 •
edited by Zethson
Loading

Describe the bug

In our pytorch template, when trying to run a project with the MNIST dataset, it will fail to download with a 503 error. See pytorch/vision#1938.

To Reproduce

Steps to reproduce the behavior:

  1. Create mlf-core pytorch project
  2. Push it to GH
  3. See train WF fail

Expected behavior

Training should start and run without any errors with the downloaded MNIST dataset.

Solution
See pytorch/vision#1938. We need to update torchvision to 0.9.0. However, this results in many package conflicts (mainly pytorch (which should be upgraded to 1.8.0), pycuda/cudatoolkit).
@Imipenem Imipenem added the bug Something isn't working label Mar 6, 2021
@Imipenem Imipenem self-assigned this Mar 6, 2021
@Zethson

This comment has been minimized.

Sign in to view
@Zethson Zethson added this to the 1.10.0 milestone Mar 6, 2021
@Imipenem

This comment has been minimized.

Sign in to view
@Zethson

This comment has been minimized.

Sign in to view
@Imipenem

This comment has been minimized.

Sign in to view
@Zethson

This comment has been minimized.

Sign in to view
@Imipenem

This comment has been minimized.

Sign in to view
@Imipenem Imipenem linked a pull request Mar 8, 2021 that will close this issue
Fix Torchvision download, upgrade mlflow-pytorch dependencies and fix GA space issues #305
Merged
@Zethson Zethson closed this as completed Mar 11, 2021
@Zethson Zethson mentioned this issue Mar 11, 2021
Merged
@Zethson
Copy link
Member

Zethson commented Mar 18, 2021

Broken again.

@Zethson Zethson reopened this Mar 18, 2021
@Zethson Zethson changed the title MNIST dataset download failes with 403 using torchvision 0.8.2 MNIST dataset download failes with 503 using torchvision 0.8.2 Mar 18, 2021
@Zethson
Copy link
Member

Zethson commented Mar 19, 2021

fmassa commented 4 days ago

The release of 0.9.1 is expected to be out in around 10 days

pytorch/vision#3549 (comment)

@Zethson Zethson mentioned this issue Mar 26, 2021
Merged
@Zethson Zethson linked a pull request Mar 30, 2021 that will close this issue
Fix MNIST loading & remove mlflow-xgboost_dask #321
Merged
@Zethson Zethson closed this as completed Mar 30, 2021
@Zethson Zethson mentioned this issue Apr 12, 2021
Merged
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@Imipenem Imipenem

Labels
bug Something isn't working high-priority
Projects
None yet
Milestone
1.10.0
2 participants
@Zethson @Imipenem