Add size to dataset, added uv and example of how to run examples

[eamag]

Hi, big datasets like reddit crash with OOM, I've added a param so we can load only a part of them, and also added UV so it's easier to run examples. Running

PYTHONPATH=. /home/eamag/TGB/.venv/bin/python examples/nodeproppred/tgbn-reddit/moving_average.py

with 0.1 reddit dataset returns

12077151it [00:28, 418731.17it/s]
{'ndcg': 0.6369556274406059}
Persistant forecast on Training takes--- 64.35990107500402 seconds ---
{'ndcg': 0.5457993247224994}
Persistant forecast on validation takes--- 14.123350016001496 seconds ---
{'ndcg': 0.5252439908370944}
Persistant forecast on Test takes--- 14.341881345993897 seconds ---

The PR isn't actually big, it's just a uv.lock file