Draft: Feat/initialization component

[le1nux]

What does this PR do?

This PR introduces the components for weight initialisation and is based on PR #161.
In PR #161 the differenct initialization methods plain, scaled and scaled_embed (see https://arxiv.org/abs/2312.16903) were implemented and added to the abstract NNModel class.
Due to some design concerns (e.g., some GPT2 internals were called from the parent), we decided to introduce a weight initialisation component that modifies the model weights in place.

General changes

• Components and factories for plain, scaled and scaled_embed initialisation.

Breaking Changes

• The raw model (i.e., the model with random weights) must be initialised with a weight initialiser, as shown here.

Checklist before submitting final PR

• My PR is minimal and addresses one issue / enhancement in isolation
• I have merged the target branch into this feature branch
• I have reviewed my own code w.r.t. correct implementation, missing type hints, proper documentation, etc.
• I have run a sample config for model training
• I have fixed all failing tests (python tests/tests.py)

[flxst]

Looks good to me generally, although I am a bit concerned about the complexity of this whole implementation. Added a few comments. The ones that contain something like "std can be auto also for initialization types other than plain" are the most important ones, I think this should definitely be fixed.

[flxst]

I think this is another thing that needs to be fixed.