Context

At the moment, we explicitly do not allow installing ado on Python 3.13. This is in part due to SFTTrainer as it was requiring torch<2.5.0, which didn't support Python 3.13 and, being used in our tests, wasn't straightforward to handle.

As PR #38 removed torch from SFTTrainer's dependencies, we can now install both and test on Python 3.13 as well.