aviralkumar2907 / CQL

Code for conservative Q-learning

aviralkumar2907/CQL Issues

About hyperparameters for D4RL Kitchen
Updated 3 months ago
Why substract entropy from Q-values? ("min_q_version == 3")
Updated 7 months ago1
ResolvePackageNotFound
Updated 2 years ago1
Make checkpoints public
Updated 2 years ago
Logsumexp calculation in CQL(H) for continuous action space
Closed 4 years ago2
Why "alpha_prime" and "min_q_weight" are not the same thing when using Lagrange?
Updated 2 years ago
Potential mismatch between math and code for CQL(rho)
Updated 2 years ago2
`rlkit/torch/sac/cql.py` not found
Closed 2 years ago1
ImportError: cannot import name 'elem_or_tuple_to_numpy' from 'rlkit.torch.core'
Closed 3 years ago3
About the derivation in paper
Updated 2 years ago1
Error when creating quantile agent for using CQL for Atari
Closed 4 years ago2
seed is not actually used in the script
Updated 3 years ago1
Unable to open file: 'batch_rl/fixed_replay/configs/quantile_pong.gin'
Updated 3 years ago1
QF_Loss backprops policy network
Updated 3 years ago17
cannot reproduce results of adroit task hammer-cloned, relocate-human and relocate-cloned
Updated 3 years ago1
Examples for D4RL Experiments
Updated 3 years ago
code bugs
Closed 3 years ago2
SAC CQL: Potential mismatch between observations and actions fed to the Q network in CQL computation.
Closed 4 years ago4
About the readability
Updated 4 years ago2
Function argument problem about expl_path_collector.collect_new_paths()
Closed 4 years ago1