package owl-rllib

Build not yet checked
  1. Overview
  2. No Docs
Reinforcement learning on top of the Owl numerical library

Install

dune-project
 Dependency

Authors

Maintainers

Sources

owl-rllib-0.1.tbz
sha256=7f21880386e8dfad71fa7be32819e437b0e555db6ef4fb457f6824ec516422d8
sha512=f49617d1f6faee0b3e32cef9565879503836b46a6eb130513f63cc598d0669fbaa8c535b43070be3cc281abbe19047e6a547fec713058f02194a519818623347

Description

Tabular methods, dynamic programming, DQN, REINFORCE, A2C and PPO on Owl's neural networks, classic control environments checked against Gymnasium, RL algorithms for language-model fine-tuning (REINFORCE, RLOO, PPO, GRPO and its variants) on a toy model, and an optional bridge to Gymnasium and VizDoom running in a Python child process.

Tags

reinforcement-learning machine-learning owl

Added to opam-repository:

Dependencies (3)

  1. owl >= "1.3"
  2. ocaml >= "5.1"
  3. dune >= "3.16"

Dev Dependencies (2)

  1. odoc with-doc
  2. alcotest with-test & >= "1.0.0"

Optional Dependencies

None

Used by

None

Conflicts

None

Build status

This version of owl-rllib was published after the check.ci.ocaml.org run of against opam-repository 8d8d144, so it has no build results yet. check.ci runs about once a week.