perf(utils): better implementation of repeat_expand_2d

[BlueAmulet]

After doing some profiling I found that repeat_expand_2d is relatively slow. Rather than manually repeating the array, take advantage of torch.nn.functional.interpolate which is a couple orders of magnitude faster.

I have tested the new and old methods and found the results to be the same.

[codecov-commenter]

Codecov Report

Merging #216 (b989a6d) into main (1a6c021) will increase coverage by 0.03%.
The diff coverage is 25.00%.

📣 This organization is not using Codecov’s GitHub App Integration. We recommend you install it so Codecov can continue to function properly for your repositories. Learn more

@@            Coverage Diff             @@
##             main     #216      +/-   ##
==========================================
+ Coverage   19.93%   19.96%   +0.03%     
==========================================
  Files          38       38              
  Lines        3241     3235       -6     
  Branches      410      409       -1     
==========================================
  Hits          646      646              
+ Misses       2579     2573       -6     
  Partials       16       16              

Impacted Files	Coverage Δ
src/so_vits_svc_fork/utils.py	21.02% <0.00%> (+0.57%)	⬆️
src/so_vits_svc_fork/__init__.py	100.00% <100.00%> (ø)

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[34j]

I'm just impressed that you can touch that code...
By the way, I'm really worried about f0.interpolate_f0. Like it obviously shouldn't be called in training.