Hi
Is there any online course about Ray?
I know that there is a tutorial, but I think that an course would be excellent.
Do the creators of Ray have any plans on making a course?
Rodney
1 Like
Hey sorry I missed this notification. @Rodney_N_Silva you are right that we don’t have a course yet. It is on our plan. Meanwhile, we plan to roll out more tutorials in Ray Summit. Please stay tuned!
1 Like
@zhz I hope there will be some ray course in the feature. Every new feature makes ray ecosystem more powerful but in the same time more complex to understand. Huge number of options to configure makes dependencies more unclear. Personally I use rllib and tune with custom gym environments but after new features I’m not sure when to use tune.run, when tuner.fit() when experiment, trainer, AIR etc.
All issues become more complex with using scheduler, bayessian search custom models with variable number of layers and activation functions, masking outputs etc.
Of course all things have examples but it’s not easy to use its if you don’t know how exactly combine all components.
My conclusion is that ray course with methodical introduction and modules dependencies explanation is very necessary.
The second course can be focused on development topics like clusters, docker usage and model serving.
1 Like