Dispose implicitly converted TorchSharp.Scalar and torch.Tensor

[hiyuh]

• SSIA.
  • included commits from #1434.
  • newly introduced TorchSharp.{Scalar,Tensor}LeakDetector can throw exceptions on implicit conversion.
    • if enabled, it would show stack trace for possible missing {TorchSharp.Scalar,torch.Tensor}.Dispose.
    • i'm not sure whether these leak detectors need retain even after all fixes in TorchSharp have been made though.
  • real fix for missing {TorchSharp.Scalar,torch.Tensor}.Dispose in TorchSharp is on going.
• CC: @ds5678

[hiyuh]

• nuked torch.Tensor.add(torch.Tensor, Scalar) w/ implicit conversion to TorchSharp.Scalar.
• i'll continue same kind fix based on "Find All References" by Visual Studio.

[hiyuh]

• nuked torch.Tensor.add(Scalar) w/ implicit conversion to TorchSharp.Scalar.
  • intentionally, i won't touch any unit tests if possible, since leaking due to test code incorrectness is not critical, as long as CI/CD passed.

[hiyuh]

• ditto for torch.Tensor.add_(torch.Tensor, Scalar).

[hiyuh]

• ditto for torch.Tensor.add_(Scalar).
• CI/CD errors on Azure DevOps are irrelevant to this MR.
  • Windows_x64_NetFX Release_Build fails on "Run Tests" due to network connection closed by remote host.
  • MacOS_arm64 Release_Build fails on "Build" due to cmake_minimum_required doesn't met.

[hiyuh]

• ditto for torch.Tensor.addcmul_(Tensor, Tensor, Scalar).

[hiyuh]

• ditto for torch.Tensor.div_(Scalar target, RoundingMode).

[hiyuh]

• ditto for torch.Tensor.mul(Scalar).

[hiyuh]

• ditto for torch.Tensor.mul_(Scalar).

[hiyuh]

• prepared before nuking torch.Tensor.pow(Scalar); after migrated to torch.Tensor.square if possible.

[hiyuh]

• ditto for torch.Tensor.pow(Scalar).

[hiyuh]

• ditto for torch.Tensor.fill_(Scalar).

[hiyuh]

• ditto for torch.Tensor.index_put_(Scalar, params Tensor[]).

[hiyuh]

• ditto for torch.Tensor.threshold{,_}(Scalar, Scalar).

[hiyuh]

• ditto for torch.Tensor.softplus(double, double).
  • this is an exceptional change since it calls private torch.Tensor.softplus1(Scalar, Scalar).

[hiyuh]

• ditto for torch.Tensor.celu{,_}(Scalar).

[hiyuh]

• ditto fot toch.Tensor.elu{,_}(double).

[hiyuh]

• ditto for torch.Tensor.hardtanh{,_}(Scalar, Scalar).

[hiyuh]

• ditto for torch.Tensor.leaky_relu{,_}(Scalar).

[hiyuh]

• ditto for torch.Tensor.clamp(Scalar?, Scalar?).

[hiyuh]

• ditto for torch.Tensor.clip(Scalar?, Scalar?).

[hiyuh]

• ditto for torch.Tensor.clamp_(Scalar?, Scalar?).

[hiyuh]

• ditto for torch.Tensor.clamp_max(Scalar).

[hiyuh]

• ditto for torch.Tensor.clamp_min(Scalar).

[hiyuh]

• ditto for torch.Tensor.eq(Scalar).

[hiyuh]

• ditto for torch.Tensor.ge(Scalar).

[hiyuh]

• ditto for torch.Tensor.le(Scalar).

[hiyuh]

• ditto for torch.Tensor.masked_fill(torch.Tensor, Scalar).

[hiyuh]

• introducing overloads on callee side, rather than fix on caller side.
  • although it might be suboptimal (sometimes, caller side could be reorganized with explicitly cached scalars/tensors), changeset would be consist (lower review cost) and conservative (more robust against uncareful caller side).