tlol-py VS tlol-scraper

Along with this reinforcement learning library, I was considering creating a paid service which would allow people to get the information from League of Legends replay files (*.rofl) and extract it to create data analytics / supervised learning datasets. However, along with the RL library I have open-sourced the data analytics library and it is now open-source and free for all. This library can be found at tlol-py and depends on tlol-scraper.

As for D4RL like datasets, I've tried to create and release open-source datasets for League during Season 12, however those datasets lacked scope which made them less useful for any researcher who wanted to use them. I have already created a library, [tlol-py](https://github.com/MiscellaneousStuff/tlol-py), which allows creating League of Legends datasets from .rofl replay files for data analysis and RL / supervised learning tasks. Perhaps if I created a discord and got community feedback, it would be much easier to create these types of datasets as people could say what would or would not be useful for them to contain.

The other one is a supervised learning / offline reinforcement learning [project](https://github.com/MiscellaneousStuff/tlol-py) which contains the only game playing [dataset](https://github.com/MiscellaneousStuff/tlol) for League of Legends (70 hours of gameplay).

Along with this reinforcement learning library, I was considering creating a paid service which would allow people to get the information from League of Legends replay files (*.rofl) and extract it to create data analytics / supervised learning datasets. However, along with the RL library I have open-sourced the data analytics library and it is now open-source and free for all. This library can be found at tlol-py and depends on tlol-scraper.