Upgrade to PyTorch 1.10

[ant0nsc]

Package upgrade. By default, models are now non-deterministic. With the deterministic flag set to True, training times for the Prostate model went from 280 sec/epoch to 360 sec/epoch. All PR models are still built deterministically.

Please follow the guidelines for PRs contained here. Checklist:

• Ensure that your PR is small, and implements one change.
• Add unit tests for all functions that you introduced or modified.
• Run PyCharm's code cleanup tools on your Python files.
• Link the correct GitHub issue for tracking.
• Update the Changelog file: Describe your change in terms of
  Added/Changed/Removed/... in the "Upcoming" section.
• When merging your PR, replace the default merge message with a description of your PR,
  and if needed a motivation why that change was required.

[ant0nsc]

Putting that on hold. Making this change leads to a 3% Dice drop for the Prostate model

[ant0nsc]

Results on Prostate fold are virtually identical to present main.

Results on Prostate 5fold: HD_55a5fd94-c2d5-4d08-b70e-78c73e91c7af

Structure	25%	50%	75%	count	max	mean	min	std
external	0.992	0.996	0.997	109.0	0.999	0.994743	0.969	0.003663
femur_r	0.982	0.984	0.986	109.0	0.990	0.983725	0.969	0.003464
femur_l	0.982	0.983	0.985	109.0	0.991	0.983431	0.967	0.003181
bladder	0.960	0.972	0.980	109.0	0.991	0.967220	0.829	0.022785
rectum	0.860	0.901	0.922	109.0	0.976	0.889303	0.644	0.050344
prostate	0.823	0.870	0.903	109.0	0.969	0.854349	0.617	0.069380
seminalvesicles	0.753	0.799	0.843	109.0	0.914	0.792128	0.360	0.079068

Main: HD_ce452db4-5049-4393-a1ee-55d234cf2651

Structure	25%	50%	75%	count	max	mean	min	std
external	0.992	0.996	0.997	109.0	0.999	0.994734	0.969	0.003688
femur_l	0.983	0.984	0.985	109.0	0.992	0.983982	0.969	0.003037
femur_r	0.983	0.984	0.985	109.0	0.990	0.983734	0.970	0.003360
bladder	0.960	0.972	0.980	109.0	0.991	0.966211	0.791	0.025511
rectum	0.865	0.904	0.922	109.0	0.974	0.889266	0.635	0.051430
prostate	0.820	0.866	0.901	109.0	0.966	0.853092	0.613	0.068902
seminalvesicles	0.757	0.797	0.841	109.0	0.913	0.790826	0.339	0.080626