Adala
github.com
Autonomous DAta (Labeling) Agent framework
Описание
Adala is an open-source framework designed to automate data labeling and other data processing tasks using autonomous AI agents. These agents leverage large language models (LLMs) to perform tasks such as data classification, summarization, and generation. Adala's agents learn and improve over time by interacting with ground truth datasets and incorporating human feedback, ensuring high reliability and efficiency in data processing workflows.
Возможности
Autonomous learning from human feedback,
Integration with large language models (LLMs),
Dynamic memory for storing and retrieving knowledge,
Modular and extensible architecture,
Open-source framework
Сценарии использования
Automated data labeling for machine learning,
Data classification and organization,
Summarization of large datasets,
Data generation for training models,
Enhancing data quality and consistency
Характеристики
| Тип | Агент |
| Категория | Данные и аналитика / Из awesome-списка |
| Цена | бесплатно |
| Платформа | Своё развёртывание |
| Системы | cli, docker, api, web |
| Язык сайта | en |
| GitHub | humansignal/adala |
| Рейтинг | 0.00 (0 отзывов) |
| Просмотры | 566 |
| Запуск | 2024-07-19 |
