Increase tol in DeformConv2d test for more robustness

[NicolasHug]

Closes #3060

As discussed in the original issue, slightly increasing the tolerance of the equality check for float16 will help make the test less flaky.

[datumbox]

LGTM! thanks.

[codecov]

Codecov Report

Merging #3320 (bd0a003) into master (9aa0722) will increase coverage by 0.08%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #3320      +/-   ##
==========================================
+ Coverage   73.84%   73.93%   +0.08%     
==========================================
  Files         104      104              
  Lines        9606     9594      -12     
  Branches     1537     1531       -6     
==========================================
- Hits         7094     7093       -1     
+ Misses       2033     2024       -9     
+ Partials      479      477       -2     

Impacted Files	Coverage Δ
torchvision/transforms/transforms.py	83.69% <0.00%> (-0.42%)	⬇️
torchvision/transforms/functional_tensor.py	79.23% <0.00%> (+0.19%)	⬆️
torchvision/models/detection/anchor_utils.py	94.66% <0.00%> (+1.33%)	⬆️
torchvision/ops/deform_conv.py	72.30% <0.00%> (+3.07%)	⬆️
torchvision/extension.py	67.21% <0.00%> (+13.11%)	⬆️

---

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 9aa0722...1091dd9. Read the comment docs.