mikelma/componet
Source code of the ICML24 paper "Self-Composing Policies for Scalable Continual Reinforcement Learning" (selected for oral presentation)
This project helps reinforcement learning researchers explore how to train AI agents that can learn new skills continually without forgetting old ones. It takes in experimental settings for tasks like Atari games or robotic manipulation and outputs performance metrics and visualizations, allowing researchers to evaluate new training approaches. It is designed for academics and practitioners working on the cutting edge of AI.
No commits in the last 6 months.
Use this if you are a reinforcement learning researcher investigating methods for agents to learn tasks sequentially and retain previously acquired knowledge.
Not ideal if you are looking for a pre-trained agent or a ready-to-use solution for an existing problem, as this is a research framework.
Stars
27
Forks
3
Language
Python
License
GPL-3.0
Category
Last pushed
Jul 20, 2024
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/mikelma/componet"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
aimagelab/mammoth
An Extendible (General) Continual Learning Framework based on Pytorch - official codebase of...
LAMDA-CL/PyCIL
PyCIL: A Python Toolbox for Class-Incremental Learning
GMvandeVen/continual-learning
PyTorch implementation of various methods for continual learning (XdG, EWC, SI, LwF, FROMP, DGR,...
LAMDA-CL/LAMDA-PILOT
🎉 PILOT: A Pre-trained Model-Based Continual Learning Toolbox
mmasana/FACIL
Framework for Analysis of Class-Incremental Learning with 12 state-of-the-art methods and 3 baselines.