MiniCPM 代码库 | MiniCPM 论文 | MiniCPM-V 代码库 | 欢迎加入我们的 Discord 和 微信交流群
MiniCPM3-4B 是 MiniCPM 系列的第三代模型。其综合性能超越了 Phi-3.5-mini-Instruct 和 GPT-3.5-Turbo-0125,可与众多最新的 7B~9B 模型相媲美。
与 MiniCPM1.0/MiniCPM2.0 相比,MiniCPM3-4B 具备更强大、更多样化的技能组合,以支持更广泛的通用场景。MiniCPM3-4B 支持函数调用和代码解释器功能。使用指南请参见进阶功能。
MiniCPM3-4B 拥有 32k 的上下文窗口。借助 LLMxMapReduce 技术,MiniCPM3-4B 理论上可处理无限长上下文,且无需占用大量内存。
from mindnlp.transformers import AutoModelForCausalLM, AutoTokenizer
path = "openbmb/MiniCPM3-4B"
tokenizer = AutoTokenizer.from_pretrained(path, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(path, trust_remote_code=True)
messages = [
{"role": "user", "content": "推荐5个北京的景点。"},
]
model_inputs = tokenizer.apply_chat_template(messages, return_tensors="pt", add_generation_prompt=True).to(device)
model_outputs = model.generate(
model_inputs,
max_new_tokens=1024,
top_p=0.7,
temperature=0.7
)
output_token_ids = [
model_outputs[i][len(model_inputs[i]):] for i in range(len(model_inputs))
]
responses = tokenizer.batch_decode(output_token_ids, skip_special_tokens=True)[0]
print(responses)目前,你需要安装我们的 vLLM 分支版本。
pip install git+https://github.com/OpenBMB/vllm.git@minicpm3from mindnlp.transformers import AutoTokenizer
from vllm import LLM, SamplingParams
model_name = "openbmb/MiniCPM3-4B"
prompt = [{"role": "user", "content": "推荐5个北京的景点。"}]
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
input_text = tokenizer.apply_chat_template(prompt, tokenize=False, add_generation_prompt=True)
llm = LLM(
model=model_name,
trust_remote_code=True,
tensor_parallel_size=1
)
sampling_params = SamplingParams(top_p=0.7, temperature=0.7, max_tokens=1024, repetition_penalty=1.02)
outputs = llm.generate(prompts=input_text, sampling_params=sampling_params)
print(outputs[0].outputs[0].text)| 评测基准 | Qwen2-7B-Instruct | GLM-4-9B-Chat | Gemma2-9B-it | Llama3.1-8B-Instruct | GPT-3.5-Turbo-0125 | Phi-3.5-mini-Instruct(3.8B) | MiniCPM3-4B | |||||||
| 英文能力 | ||||||||||||||
| MMLU(多任务语言理解) | 70.5 | 72.4 | 72.6 | 69.4 | 69.2 | 68.4 | 67.2 | |||||||
| BBH(大语言模型推理能力) | 64.9 | 76.3 | 65.2 | 67.8 | 70.3 | 68.6 | 70.2 | |||||||
| MT-Bench(多轮对话质量) | 8.41 | 8.35 | 7.88 | 8.28 | 8.17 | 8.60 | 8.41 | |||||||
| IFEVAL(指令遵循严格准确率) | 51.0 | 64.5 | 71.9 | 71.5 | 58.8 | 49.4 | 68.4 | |||||||
| 中文能力 | ||||||||||||||
| CMMLU(中文多任务语言理解) | 80.9 | 71.5 | 59.5 | 55.8 | 54.5 | 46.9 | 73.3 | |||||||
| CEVAL(中文基础模型评估) | 77.2 | 75.6 | 56.7 | 55.2 | 52.8 | 46.1 | 73.6 | |||||||
| AlignBench v1.1(中文对齐基准) | 7.10 | 6.61 | 7.10 | 5.68 | 5.82 | 5.73 | 6.74 | |||||||
| FollowBench-zh(中文指令遵循SSR指标) | 63.0 | 56.4 | 57.0 | 50.6 | 64.6 | 58.1 | 66.8 | |||||||
| 数学能力 | ||||||||||||||
| MATH(数学问题解决) | 49.6 | 50.6 | 46.0 | 51.9 | 41.8 | 46.4 | 46.6 | |||||||
| GSM8K(小学数学问题) | 82.3 | 79.6 | 79.7 | 84.5 | 76.4 | 82.7 | 81.1 | |||||||
| MathBench(数学综合能力) | 63.4 | 59.4 | 45.8 | 54.3 | 48.9 | 54.9 | 65.6 | |||||||
| 代码能力 | ||||||||||||||
| HumanEval+(代码生成评估) | 70.1 | 67.1 | 61.6 | 62.8 | 66.5 | 68.9 | 68.3 | |||||||
| MBPP+(代码生成与执行) | 57.1 | 62.2 | 64.3 | 55.3 | 71.4 | 55.8 | 63.2 | |||||||
| LiveCodeBench v3(实时代码基准) | 22.2 | 20.2 | 19.2 | 20.4 | 24.0 | 19.6 | 22.6 | |||||||
| 函数调用能力 | ||||||||||||||
| BFCL v2(函数调用基准) | 71.6 | 70.1 | 19.2 | 73.3 | 75.4 | 48.4 | 76.0 | |||||||
| 综合能力 | ||||||||||||||
| 平均值 | 65.3 | 65.0 | 57.9 | 60.8 | 61.0 | 57.2 | 66.3 | |||||||
@article{hu2024minicpm,
title={MiniCPM: Unveiling the Potential of Small Language Models with Scalable Training Strategies},
author={Hu, Shengding and Tu, Yuge and Han, Xu and He, Chaoqun and Cui, Ganqu and Long, Xiang and Zheng, Zhi and Fang, Yewei and Huang, Yuxiang and Zhao, Weilin and others},
journal={arXiv preprint arXiv:2404.06395},
year={2024}
}