Update squeezenet weights urls

[NicolasHug]

Addresses part of #2068 - weights are unchanged, but the files can now be properly unpickled.

The new pth files have been uploaded to the S3 bucket and to the FB internal db.

For ref, here's the script that was used: https://nbviewer.jupyter.org/gist/NicolasHug/fb2c28d758951959da94d3db326ce21b

[codecov]

Codecov Report

Merging #3581 (a7f348f) into master (c00c4d3) will decrease coverage by 0.09%.
The diff coverage is n/a.

❗ Current head a7f348f differs from pull request most recent head f8b1cd1. Consider uploading reports for the commit f8b1cd1 to get more accurate results

@@            Coverage Diff             @@
##           master    #3581      +/-   ##
==========================================
- Coverage   79.07%   78.98%   -0.10%     
==========================================
  Files         105      105              
  Lines        9786     9786              
  Branches     1572     1572              
==========================================
- Hits         7738     7729       -9     
- Misses       1567     1572       +5     
- Partials      481      485       +4     

Impacted Files	Coverage Δ
torchvision/models/squeezenet.py	86.79% <ø> (ø)
torchvision/extension.py	54.09% <0.00%> (-13.12%)	⬇️
torchvision/ops/deform_conv.py	69.23% <0.00%> (-3.08%)	⬇️
torchvision/models/detection/anchor_utils.py	93.33% <0.00%> (-1.34%)	⬇️
torchvision/transforms/functional_tensor.py	79.28% <0.00%> (-0.57%)	⬇️
torchvision/datasets/utils.py	75.10% <0.00%> (+2.18%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update db1787f...f8b1cd1. Read the comment docs.

[datumbox]

LGTM.

@NicolasHug Just to confirm, you have also updated the manifold bucket for the internal users?

[NicolasHug]

you have also updated the manifold bucket for the internal users?

yes