This is a user guide for the MiniCPM and MiniCPM-V series of small language models (SLMs) developed by ModelBest. “面壁小钢炮” focuses on achieving exceptional performance on the edge.

Official completion of “Training on the Benchmark Is Not All You Need”.

MiniCPM3-4B: An edge-side LLM that surpasses GPT-3.5-Turbo.

An open-source framework for collaborative AI agents, enabling diverse, distributed agents to team up and tackle complex tasks through internet-like connectivity.

A Toolkit for Running On-device Large Language Models (LLMs) in APP

MiniCPM-V 2.6: A GPT-4V Level MLLM for Single Image, Multi Image and Video on Your Phone

Ouroboros: Speculative Decoding with Large Model Enhanced Drafting (EMNLP 2024 main)

Open Platform for Embodied Agents

The code of our paper "InfLLM: Unveiling the Intrinsic Capacity of LLMs for Understanding Extremely Long Sequences with Training-Free Memory"

An Autonomous LLM Agent for Complex Task Solving

Create Customized Software using Natural Language Idea (through LLM-powered Multi-Agent Collaboration)

Codes for our paper "ChatEval: Towards Better LLM-based Evaluators through Multi-Agent Debate"

[ICLR'24 spotlight] Chinese and English Multimodal Large Model Series (Chat and Paint) | 基于CPM基础模型的中英双语多模态大模型系列

[ICLR'24 spotlight] An open platform for training, serving, and evaluating large language model for tool learning.

A collection of phenomenons observed during the scaling of big foundation models, which may be developed into consensus, principles, or laws in the future

百亿参数的中英文双语基座大模型

Official codes for ACL 2023 paper "WebCPM: Interactive Web Search for Chinese Long-form Question Answering"

🤖 AgentVerse 🪐 is designed to facilitate the deployment of multiple LLM-based agents in various applications, which primarily provides two frameworks: task-solving and simulation

Tool Learning for Big Models, Open-Source Solutions of ChatGPT-Plugins

Large-scale, Informative, and Diverse Multi-round Chat Data (and Models)

Source code for "A Deep-learning System Bridging Molecule Structure and Biomedical Text with Comprehension Comparable to Human Professionals"

Live Training for Open-source Big Models

An Open-Source Package for Information Retrieval

A comprehensive, unified and modular event extraction toolkit.

A List of Big Models

Must-read Papers of Parameter-Efficient Tuning (Delta Tuning) Methods on Pre-trained Models.

Source code and checkpoints for legal pre-trained language models.