search

tinkoff-ai / CORL

High-quality single-file implementations of SOTA Offline and Offline-to-Online RL algorithms: AWAC, BC, CQL, DT, EDAC, IQL, SAC-N, TD3+BC, LB-SAC, SPOT, Cal-QL, ReBRAC
https://arxiv.org/abs/2210.07105
Apache License 2.0
1.06k stars 124 forks source link

Finetune: IQL #46

Closed DT6A closed 1 year ago

DT6A commented 1 year ago

Wandb report

DT6A commented 1 year ago

@vkurenkov typings are fixed and lr params are added (to both offline and finetune versions) but https://github.com/tinkoff-ai/CORL/pull/52 need to be merged first

DT6A commented 1 year ago

@vkurenkov @Howuhh @Scitator