A list of commonly tweaked parameters and how to tweak them in the experiment CLI

[laserkelvin]

It would be very handy to have a subsection in the experiment CLI documentation that lists out things that are regularly tweaked, and which config file to modify. Some parameters that come to mind are:

• Learning rate (sits in model/task definition)
• Callbacks (trainer config)
• DDP training
• Data loader workers and batch size

It is redundant since some of the things are already covered in the full text, but would be nice to have a very short and snappy section that lets you know at a glance. Feel free to add to the list.

@lory-w if you have things to add as you discover them, you can omnibus them in if you beat @melo-gonzo 😸

[laserkelvin]

Closed by #295.