search

edbeeching / godot_rl_agents

An Open Source package that allows video game creators, AI researchers and hobbyists the opportunity to learn complex behaviors for their Non Player Characters or agents
MIT License
901 stars 63 forks source link

Suggestion for docs: Graphing progress with Tensorboard #64

Open Jetpackjules opened 1 year ago

Jetpackjules commented 1 year ago

Hello,

I just have a quick suggestion for the documentation:

Running the command

Tensorboard --logdir=~\ray_results\PPO\editor\ --port=8008

in a terminal while/after the program is training can provide some useful insight that helps with debugging and fine-tuning the model.

For example: image

This site updates whenever the page is reloaded, and works great for trakcing the success of a model. I think this would be a great addition for new people trying to train their models, as it certainly helped me!

Ivan-267 commented 9 months ago

Thanks for this request.

It seems we have it here now: https://github.com/edbeeching/godot_rl_agents/blob/main/docs/TRAINING_STATISTICS.md

Although adding this link to the quick-start guide or one of the advanced tutorials might be useful.