Small fix for Pytorch shared memory leak

[AbhinavTuli]

New pytorch was failing for datasets larger than shared memory size. This fixes it.
self.all_shared_memory was earlier used to ensure that memory wasn't dereferenced, but the logic was changed and this is no longer required.

[codecov]

Codecov Report

Merging #1040 (6992d4c) into main (0c8b222) will decrease coverage by 0.00%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main    #1040      +/-   ##
==========================================
- Coverage   88.61%   88.61%   -0.01%     
==========================================
  Files          86       86              
  Lines        3917     3916       -1     
==========================================
- Hits         3471     3470       -1     
  Misses        446      446              

Impacted Files	Coverage Δ
hub/integrations/pytorch.py	80.24% <100.00%> (-0.13%)	⬇️

---

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 0c8b222...6992d4c. Read the comment docs.

[verbose-void]

sometimes the best solution is to make the code just a little prettier lol