Upgrade Playground

[dmytropolityka]

Motivation and Context

Playground was not extended to support self-learning feedback generation, although this feature exists.
Also, this PR #280 changed some internals and changed Authentication.
Playground was not affected. This PR fixes that.

Description

1. Renamed all URL fields into uri fields to stay consistent with Artemis.
2. Added deployment information about Playground to enable support for new Authentication.
3. Updated nodejs version for docker images.
4. Added lms(client) URL field to the playground header.
5. Chaned athena fetcher to modify the authorization headers in Playground client.
6. Chaned Playground server to forward these headers.
7. Removed unnecessary yarn lock file.

Steps for Testing

Deploy the testing setup on the ts or locally. Submit new self-learning-feedback requests and old graded requests. Verify everything works.
For local, the key is empty. For athena ts3, the key is here: https://confluence.ase.in.tum.de/pages/viewpage.action?pageId=25252245

Screenshots

[FelixTJDietrich]

Why is the prospector linting check failing? Can we fix this in a follow-up, I will create an issue.

Run echo "$HOME/.local/bin" >> $GITHUB_PATH
Creating virtualenv athena-modularized-tH5MAedG-py3.11 in /home/runner/.cache/pypoetry/virtualenvs
Installing dependencies from lock file

pyproject.toml changed significantly since poetry.lock was last generated. Run `poetry lock [--no-update]` to fix the lock file.

[FelixTJDietrich]

Code LGTM, works as expected on the test server. I will merge it now.

[dmytropolityka]

Why is the prospector linting check failing? Can we fix this in a follow-up, I will create an issue.

Run echo "$HOME/.local/bin" >> $GITHUB_PATH
Creating virtualenv athena-modularized-tH5MAedG-py3.11 in /home/runner/.cache/pypoetry/virtualenvs
Installing dependencies from lock file

pyproject.toml changed significantly since poetry.lock was last generated. Run `poetry lock [--no-update]` to fix the lock file.

This is because we use different versions of langchain for different llm modules. We cannot fix it now but this issue will disappear when we decouple modules from the top-level module(I will create a PR by tomorrow).