About DQN replay dataset
sglucas opened this issue · 2 comments
sglucas commented
Hi
Thank you so much for your contribution. This is a really great repo for students.
I think it will be very nice if we can try the atari offline training with some recently proposed methods.
Could you please recommend some recent papers about offline rl training on atari?
Thank you very much!
Best
agarwl commented
Here's a biased list:
- Conservative Q-learning
- DR3: Value-Based Deep Reinforcement Learning Requires Explicit Regularization
- Decision Transformer / Multi-game Decision Transformer
- Offline Q-Learning on Diverse Multi-Task Data Both Scales And Generalizes
- Online and Offline Reinforcement Learning by Planning with a Learned Model
- Efficient Offline Policy Optimization with a Learned Model
- Pretraining representations for data-efficient reinforcement learning