Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

MiniCPM-2B-DPO

Model basics

Reasoning traces

Not supported

Context length

2K tokens

Max output length

No data

Model type

聊天大模型

Release date

2024-01-30

Model file size

10.9GB

MoE architecture

Total params / Active params

24.0B / N/A

Knowledge cutoff

No data

Inference modes

No mode data

MiniCPM-2B-DPO

Open source & experience

Code license

Apache 2.0

Weights license

OpenBMB通用模型许可协议-来源说明-宣传限制-商业授权- 免费商用授权

GitHub repo

https://github.com/OpenBMB/MiniCPM

Hugging Face

https://huggingface.co/openbmb/MiniCPM-2B-dpo-fp32

Live demo

https://www.datalearner.com/llm-tutorials/pretrained-model-tutorials/how-to-deploy-minicpm-2b-pretrained-large-language-model

MiniCPM-2B-DPO

Official resources

Paper

MiniCPM: Unveiling the Potential of End-side Large Language Models

DataLearnerAI blog

面壁智能开源迄今位置最强的20亿参数规模的大语言模型：MiniCPM-20B，评测结果超过阿里巴巴Qwen-1.8B，多项评测结果接近GPT-3和众多70亿参数规模大模型

MiniCPM-2B-DPO

API details

API speed

No data

No public API pricing yet.

MiniCPM-2B-DPO

Benchmark Results

No benchmark data to show.

MiniCPM-2B-DPO

Publisher

面壁智能

View publisher details

MiniCPM-2B-DPO

Model Overview

MiniCPM-2B-DPO是由面壁智能开源的一个24亿参数规模的大语言模型，是一个参数规模不大的语言模型。但是官方的评测结果中显示，该模型的评测水平超过了很多70亿参数规模的大语言模型，最高与130亿参数的Llama2-13B差不多。

DataLearner 官方微信

欢迎关注 DataLearner 官方微信，获得最新 AI 技术推送