search

google-research / recsim

A Configurable Recommender Systems Simulation Platform
https://github.com/google-research/recsim
Apache License 2.0
741 stars 129 forks source link

Tensorflow 2.0 and more complex agents examples #3

Open pkorobov opened 4 years ago

pkorobov commented 4 years ago

Hello, thank you for this great work!

As I understand, RecSim is made for use with TF 1, because, if I'm not mistaken, we must provide a session to an agent. Do you plan to adapt interface for the new version of TF?

And it would be great to have some more examples of more complex deep RL agents, besides DQN.

Thanks!

mmladenov-google commented 4 years ago

Hey! Thanks for the kind words and your interest in RecSim! TF2 support is in progress and will definitely happen in one of the next releases. It's definitely worthwhile to pursue more complex agents, although I'm not sure how high this is on the priority list. Perhaps one way to get exposure to more advanced agents quickly is to take advantage of the fact that environment.Environment conforms (more or less) to the OpenAI Gym API, so in principle a RecSim environment could be used directly within an RL framework such as TF-Agents. I'll take some time to see if there are any blockers to that and report back.

getsanjeevdubey commented 3 years ago

@mmladenov-google hi, do we have any update on this?