Reverb is used at DeepMind and @GoogleResearch to manage RL experience at scale and is now open sourced.
— DeepMind (@DeepMind) June 1, 2020
Reverb can be used in a Colab or scale up to thousands of machines for on and off-policy RL: DQN, D4PG, PPO, IMPALA... see here: https://t.co/8qWRbtVAGn
from Twitter https://twitter.com/DeepMind
June 01, 2020 at 09:37AM
via IFTTT
No comments:
Post a Comment