ilql.rst

.. _ilql:


Implicit Language Q-Learning (ILQL)
===================================

ILQL is an extension of Implicit Q-learning that can be used to finetune large language models (LLMs) with
reinforcement leaning from human feedback (RLHF).

* ILQL paper: https://arxiv.org/pdf/2206.11871.pdf

Parameters
-----------

.. autoclass:: agilerl.algorithms.ilql.ILQL
  :members:
  :inherited-members: