Evaluate three types of task shifting with popular continual learning algorithms.

CORe50: a new Dataset and Benchmark for Continual Learning

TRACE: A Comprehensive Benchmark for Continual Learning in Large Language Models

A library for benchmarking the Long Term Memory and Continual learning capabilities of LLM based agents. With all the tests and code you need to evaluate your own agents. See more in the blogpost:

The Continual Learning in Multimodality Benchmark

Benchmarking continual learning techniques for Human Activity Recognition data. We offer interesting insights on how the performance techniques vary with a domain other than images.

The original code for the paper "Benchmarks for Continual Few-Shot Learning".

Implementation of CLAD: A Continual Learning benchmark for Autonomous Driving. A continual classification and detectin track. Provides pure PyTorch, Avalanche and Detectron2 implementations.

CLEAR benchmark (NeurIPS 2021 Dataset & Benchmark)

COOM: Benchmarking Continual Reinforcement Learning on Doom

The Official benchmark for continual learning for deepfake audio detection

Video-Language Continual Learning Benchmark

✌ CLoG: Benchmarking Continual Learning of Image Generation Models

Benchmark framework for evaluating LLM agent continual learning in stateful environments. Features production-realistic CRM workflows with multi-turn conversations, state mutations, and cross-entity relationships. Extensible to additional domains

Continual Learning Benchmark for Spoken Keyword Spotting

The original code for the data providers and the datasets of the paper "Defining Benchmarks for Continual Few-Shot Learning".

Co-SAM Benchmark for evaluating Segment Anything Model (SAM) adaptation algorithms in a Continual Learning framework

AVQACL: A Novel Benchmark for Audio-Visual Question Answering Continual Learning

Easily generate continual learning benchmarks.

ACMMM2025-Serial Over Parallel: Learning Continual Unification for Multi-Modal Visual Object Tracking and Benchmarking

A benchmark of online learning algorithms for RNNs in a continual learning setting. The models are evaluated on the task of incrementally learning a repertoire of 20 continuous sequences.

Learderboards of Continual Learning for various benchmarks.

Continual learning benchmark using CLIP + LoRA for CIFAR-10. Includes Catastrophic Forgetting experiment and PEFT mitigation.

A distributed Continual Learning benchmark for a large sequence of tasks.

A benchmark for Continual Multi-Agent Reinforcement Learning on the Overcooked environment.

No description available

STREAMS: A Benchmark of Naturalistic Streaming Data for Online Continual Learning

[WACV'26] PoseAdapt: Sustainable Human Pose Estimation via Continual Learning Benchmarks and Toolkit

Use the CORe50 Continual Learning Benchmark Dataset in a Meta-Learning setting via the pyMeta library.

This is a benchmark for AI-generated image detection and model adaptation through continual learning