thedotmack/claude-mem
Claude-Mem is a plugin for Claude Code that captures, compresses, and preserves coding session context to enable persistent memory and continuity across sessions.
Awesome AI Agents › Continual Learning Agents
allenai/lumos
Lumos is an advanced open-source project developed by AllenAI that focuses on creating language agents with a unified data format, modular design, and integration of open-source large language models (LLMs). The project aims to unify a variety of complex interactive tasks into a single framework, enabling the development of versatile and high-performing language agents. Lumos is built on a modular architecture that includes planning, grounding, and execution modules, primarily leveraging LLAMA-2 models and off-the-shelf APIs. This modularity allows for flexible and efficient handling of diverse tasks. A key feature of Lumos is its use of a unified data format that supports multiple task types, facilitating the training and evaluation of agents across different interactive scenarios. The training data comprises approximately 56,000 high-quality subgoal and action annotations derived from ground-truth reasoning steps in existing benchmarks, enhanced with GPT-4 annotations. This rich dataset supports the development of agents capable of complex reasoning and task execution. Lumos demonstrates competitive performance, often matching or surpassing GPT-3.5/4-based agents and other larger open-source agents on various benchmarks, including web-based question answering (Mind2Web, HotpotQA), math, and multimodal tasks. It outperforms contemporaneous fine-tuned agents like FiReAct, AgentLM, and AutoAct, as well as baseline methods such as chain-of-thought and integrated training approaches. Additionally, Lumos excels on unseen tasks like WebShop and InterCode_SQL, showcasing its generalization capabilities. The project provides comprehensive resources, including code for annotation generation, training, and evaluation, as well as pre-trained model checkpoints and datasets available on Huggingface and Google Drive. It also offers a demo illustrating the planning and grounding processes. Lumos is well-documented and actively maintained, making it a valuable resource for researchers and developers interested in building sophisticated, open-source language agents for complex interactive tasks.
https://github.com/allenai/lumos
Also in Continual Learning Agents
vectorize-io/hindsight
Hindsight is an agent memory system designed to create smarter, learning AI agents, addressing limitations of traditional memory techniques like RAG and knowledge graphs.
langchain-ai/langchainjs
LangChain.js is a TypeScript framework for building context-aware, reasoning applications powered by large language models, offering modular components, chains, agents, and tools for production and deployment.
aden-hive/hive
Hive is an outcome-driven agent development framework designed for evolving agents with AI evaluation, human-in-the-loop, and self-improvement capabilities.
aiming-lab/MetaClaw
Proxy-based layer that turns everyday conversations with a personal agent into learning signals, injecting skills per turn and running scheduled reinforcement learning updates without a local GPU.
kayba-ai/agentic-context-engine
Agentic Context Engine (ACE) is a framework that enables AI agents to autonomously learn from their successes and failures, continuously improving their performance across various tasks without additional training data or fine-tuning.
AgentDock/AgentDock
AgentDock is an open-source framework for building sophisticated AI agents with configurable determinism, enabling reliable and creative AI applications through a node-based architecture and multi-stage workflows.
modelscope/AgentEvolver
Self-evolving agent training framework from ModelScope that combines automatic task generation, experience-guided exploration and attribution-based credit assignment into one reinforcement learning stack.