cleaned up docker/docker-compose stuff a bit

[ppetermann]

I've run into a few smaller issues when running autogpt in docker through docker-compose, mostly related to weirdness with paths and the Dockerfile copying things while the docker-compose file mounted volumes oddly.

Changes

Dockerfile

• using full path for autogpt module, so the directory is copied to the place where its needed
• adding plugins folder to be copied, so that plugins can be installed as well

docker-compose.yml

Overhauled volumes

• mounting .env shouldn't be necessary, as env_file should already take care of making the environment variables available at run time (dropped)
• mounting autogpt into /home/appuser made autogpt the home, while the dockerfile made it a subdirectory of home (it should be the later), a commented out example how to mount it as such was added
• mounting plugins can save time when working on plugins, a commented out example was provided
• mounting the workspace can be extremly usefull, an example has been provided here as well, although i suspect people are using other folders / pathes

Documentation

changes are documented in changes above, and have code comments if context requires

Test Plan

Run autogpt through docker or docker-compose

PR Quality Checklist

• [x] My pull request is atomic and focuses on a single change.
• [x] I have thoroughly tested my changes with multiple different prompts.
• [x] I have considered potential risks and mitigations for my changes.
• [x] I have documented my changes clearly and comprehensively.
• [x] I have not snuck in any "extra" small tweaks changes

I have not added any new tests, as i've not created any real code