Use float instead of integers for durations in Tween documentation

[timkrief]

I think it could be cleaner to use floats instead of integers, since duration is supposed to be floating numbers. In this case it doesn't make a big difference, but if you use integers instead of floats as target value in tweens, it can create unexpected results, so I think it could be cleaner to use floats when floats are expected, at least in the documentation itself.

[Calinou]

Looks good to me.

In general, I prefer docs (and sample code) having explicit float literals whenever the value being set is a float.

[akien-mga]

As this will invalidate translations, let's merge for 4.5-dev1, so translators have enough time to handle the fuzzy strings before 4.5-stable.

[tetrapod00]

Suggested change

	tween.tween_method(look_at.bind(Vector3.UP), Vector3(-1, 0, -1), Vector3(1, 0, -1), 1.0) # The look_at() method takes up vector as second argument.
	tween.tween_method(look_at.bind(Vector3.UP), Vector3(-1.0, 0.0, -1.0), Vector3(1.0, 0.0, -1.0), 1.0) # The look_at() method takes up vector as second argument.

If we are trying to use explicit float literals, why are we not using them for the Vector2/Vector3 constructors too, for consistency?

I don't necessarily think we should do so. But if we choose to keep the inconsistency we should do so intentionally and have some reason even if it's just "it's longer and doesn't increase clarity"

[Repiteo]

We can handle this in a separate PR/discussion, no need to keep this one on ice in the meantime

[Repiteo]

Thanks! Congratulations on your first merged contribution! 🎉