Fix NaN in gradients

[SebastianM-C]

The initial conditions for the UDE are the same as the reference solution, leading to taking the derivative of sqrt at 0, which is NaN. Changing the loss function to squared l2loss fixes this.

Checklist

• [x] Appropriate tests were added
• [x] Any code changes were done in a way that does not break public API
• [ ] ~All documentation related to code changes were updated~ No docs yet
• [x] The new code follows the contributor guidelines, in particular the SciML Style Guide and COLPRAC.
• [ ] ~Any new documentation only uses public API~ No docs yet

Additional context

Fixes #6

[codecov[bot]]

Welcome to Codecov :tada:

Once you merge this PR into your default branch, you're all set! Codecov will compare coverage reports and display results in all future pull requests.

Thanks for integrating Codecov - We've got you covered :open_umbrella:

[ChrisRackauckas]

We should remove nightly from the tests, I don't know how that got on there.

[SebastianM-C]

I think PkgTemplates added it, I'll drop that.

[SebastianM-C]

Other than that the CI failures are unrelated to this PR.