Text Generation
Transformers
Safetensors
English
Chinese
llama
minicpm
minicpm5
long-context
tool-calling
on-device
edge-ai
conversational
text-generation-inference
4-bit precision
Instructions to use openbmb/MiniCPM5-1B-MLX with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use openbmb/MiniCPM5-1B-MLX with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="openbmb/MiniCPM5-1B-MLX") messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("openbmb/MiniCPM5-1B-MLX") model = AutoModelForCausalLM.from_pretrained("openbmb/MiniCPM5-1B-MLX", device_map="auto") messages = [ {"role": "user", "content": "Who are you?"}, ] inputs = tokenizer.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", ).to(model.device) outputs = model.generate(**inputs, max_new_tokens=40) print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use openbmb/MiniCPM5-1B-MLX with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "openbmb/MiniCPM5-1B-MLX" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "openbmb/MiniCPM5-1B-MLX", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/openbmb/MiniCPM5-1B-MLX
- SGLang
How to use openbmb/MiniCPM5-1B-MLX with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "openbmb/MiniCPM5-1B-MLX" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "openbmb/MiniCPM5-1B-MLX", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "openbmb/MiniCPM5-1B-MLX" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "openbmb/MiniCPM5-1B-MLX", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use openbmb/MiniCPM5-1B-MLX with Docker Model Runner:
docker model run hf.co/openbmb/MiniCPM5-1B-MLX
Update README-cn.md
Browse files- README-cn.md +318 -7
README-cn.md
CHANGED
|
@@ -21,19 +21,330 @@ datasets:
|
|
| 21 |
- openbmb/UltraData-SFT-2605
|
| 22 |
---
|
| 23 |
|
| 24 |
-
|
| 25 |
-
|
| 26 |
-
|
| 27 |
<div align="center">
|
| 28 |
<img src="https://raw.githubusercontent.com/OpenBMB/MiniCPM/main/assets/minicpm_logo.png" width="500em" />
|
| 29 |
</div>
|
| 30 |
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
|
| 34 |
<p align="center">
|
| 35 |
<a href="https://arxiv.org/pdf/2506.07900" target="_blank">MiniCPM 技术报告</a> |
|
| 36 |
<a href="https://github.com/OpenBMB/MiniCPM" target="_blank">GitHub 仓库</a> |
|
| 37 |
<a href="https://ultradata.openbmb.cn/" target="_blank">UltraData</a> |
|
| 38 |
<a href="https://github.com/OpenBMB/MiniCPM-Desk-Pet" target="_blank">MiniCPM 桌宠</a> |
|
| 39 |
-
<a href="https://huggingface.co/spaces/openbmb/MiniCPM5-1B-Demo" target="_blank">在线 Demo</a>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 21 |
- openbmb/UltraData-SFT-2605
|
| 22 |
---
|
| 23 |
|
|
|
|
|
|
|
|
|
|
| 24 |
<div align="center">
|
| 25 |
<img src="https://raw.githubusercontent.com/OpenBMB/MiniCPM/main/assets/minicpm_logo.png" width="500em" />
|
| 26 |
</div>
|
| 27 |
|
|
|
|
|
|
|
|
|
|
| 28 |
<p align="center">
|
| 29 |
<a href="https://arxiv.org/pdf/2506.07900" target="_blank">MiniCPM 技术报告</a> |
|
| 30 |
<a href="https://github.com/OpenBMB/MiniCPM" target="_blank">GitHub 仓库</a> |
|
| 31 |
<a href="https://ultradata.openbmb.cn/" target="_blank">UltraData</a> |
|
| 32 |
<a href="https://github.com/OpenBMB/MiniCPM-Desk-Pet" target="_blank">MiniCPM 桌宠</a> |
|
| 33 |
+
<a href="https://huggingface.co/spaces/openbmb/MiniCPM5-1B-Demo" target="_blank">在线 Demo</a>
|
| 34 |
+
</p>
|
| 35 |
+
|
| 36 |
+
<p align="center">
|
| 37 |
+
<a href="https://huggingface.co/openbmb/MiniCPM5-1B/blob/main/README.md" target="_blank">English</a> |
|
| 38 |
+
中文
|
| 39 |
+
</p>
|
| 40 |
+
|
| 41 |
+
## 亮点
|
| 42 |
+
|
| 43 |
+
我们正式发布 **MiniCPM5-1B**,这是 **MiniCPM5** 系列的首个模型。它是一款面向端侧、本地部署和资源受限场景的 1B 稠密 Transformer,能够达到同尺寸开源模型 SOTA 水平。
|
| 44 |
+
|
| 45 |
+
🏆 **同尺寸开源模型 SOTA**:与同尺寸优秀开源模型相比,MiniCPM5-1B 在该对比范围内达到 SOTA 水平,优势主要体现在 Agentic 工具调用、代码生成和高难推理。
|
| 46 |
+
|
| 47 |
+

|
| 48 |
+
|
| 49 |
+
🧠 **双模式推理**:内置 `<think>` chat template,可通过 `enable_thinking` 在思考模式和非思考模式之间切换。同一份权重既可以作为快速助手,也可以承担更复杂的推理任务。
|
| 50 |
+
|
| 51 |
+
🛠️ **部署 / 微调资源**:MiniCPM GitHub 仓库提供面向主要推理后端和微调框架的单页 cookbook,并配套 Agent Skills,方便复现部署和微调流程。
|
| 52 |
+
|
| 53 |
+
🐱 **桌宠**:我们也提供了由 MiniCPM5-1B 本地驱动的桌宠应用。
|
| 54 |
+
|
| 55 |
+
## 模型列表
|
| 56 |
+
|
| 57 |
+
你可以按运行环境选择对应模型格式:
|
| 58 |
+
|
| 59 |
+
- **[MiniCPM5-1B](https://huggingface.co/openbmb/MiniCPM5-1B)** · [ModelScope](https://www.modelscope.cn/models/OpenBMB/MiniCPM5-1B) · BF16 正式版(经 RL + OPD 后训练)
|
| 60 |
+
- **[MiniCPM5-1B-SFT](https://huggingface.co/openbmb/MiniCPM5-1B-SFT)** · [ModelScope](https://www.modelscope.cn/models/OpenBMB/MiniCPM5-1B-SFT) · BF16 SFT 单独 checkpoint(RL / OPD 之前)
|
| 61 |
+
- **[MiniCPM5-1B-Base](https://huggingface.co/openbmb/MiniCPM5-1B-Base)** · [ModelScope](https://www.modelscope.cn/models/OpenBMB/MiniCPM5-1B-Base) · BF16 base checkpoint(仅预训练)
|
| 62 |
+
- **[MiniCPM5-1B-GGUF](https://huggingface.co/openbmb/MiniCPM5-1B-GGUF)** · [ModelScope](https://www.modelscope.cn/models/OpenBMB/MiniCPM5-1B-GGUF) · GGUF,适用于 llama.cpp / Ollama / LM Studio
|
| 63 |
+
- **[MiniCPM5-1B-MLX](https://huggingface.co/openbmb/MiniCPM5-1B-MLX)** · [ModelScope](https://www.modelscope.cn/models/OpenBMB/MiniCPM5-1B-MLX) · MLX / 4bit,适用于 Apple Silicon **👈 当前页面**
|
| 64 |
+
|
| 65 |
+
## 模型信息
|
| 66 |
+
|
| 67 |
+
MiniCPM5-1B 具有以下特性:
|
| 68 |
+
|
| 69 |
+
- **类型**:Causal Language Model
|
| 70 |
+
- **架构**:标准 `LlamaForCausalLM`
|
| 71 |
+
- **参数数量**:1,080,632,832
|
| 72 |
+
- **非嵌入参数数量**:679,552,512
|
| 73 |
+
- **层数**:24
|
| 74 |
+
- **注意力头(GQA)**:16 个 Q heads / 2 个 KV heads
|
| 75 |
+
- **上下文长度**:131,072
|
| 76 |
+
|
| 77 |
+
## 简介
|
| 78 |
+
|
| 79 |
+
MiniCPM5-1B 是 MiniCPM5 系列的首个模型,面向本地助手、coding agent、工具调用流程以及需要紧凑模型的推理场景。它在较小部署成本下提供原生长上下文能力,并通过同一份权重支持 Think / No Think 两种对话模式。
|
| 80 |
+
|
| 81 |
+
## 评测结果
|
| 82 |
+
|
| 83 |
+
我们选取 **LFM2.5-1.2B-Thinking**、**Qwen3-0.6B/think**、**Qwen3.5-0.8B/think** 等同尺寸优秀开源模型进行横向比较。这些模型本身已经很强;在这组对比中,MiniCPM5-1B 达到同尺寸开源模型 SOTA 水平,优势主要体现在工具调用、代码生成和高难推理上,也更适合承担本地 coding agent、工具助手和推理助手的角色。
|
| 84 |
+
|
| 85 |
+

|
| 86 |
+
|
| 87 |
+
## 训练流程
|
| 88 |
+
|
| 89 |
+
MiniCPM5-1B 的训练过程是 **[UltraData 分级数据管理体系](https://arxiv.org/pdf/2602.09003)** 的一次完整实践,覆盖 base training、mid-training 与后训练三个阶段。
|
| 90 |
+
|
| 91 |
+
**Base training** 采用逐级推进的训练配方,包含 stable training 与 decay training,用于建立基础语言能力与训练稳定性。随后进入 **mid-training**,进一步强化目标能力并适配数据分布。训练语料来自我们同步开源的 [Ultra-FineWeb](https://huggingface.co/datasets/openbmb/Ultra-FineWeb)、[Ultra-FineWeb-L3](https://huggingface.co/datasets/openbmb/Ultra-FineWeb-L3) 与 [UltraData-Math](https://huggingface.co/datasets/openbmb/UltraData-Math)。
|
| 92 |
+
|
| 93 |
+
**后训练阶段**分为 **SFT**、**RL** 与 **OPD** 三步。我们先使用 **200B tokens deep-thinking SFT** 与 **200B tokens hybrid-thinking SFT** 建立深度思考、混合思考和通用对话能力,相关 SFT 数据已同步开源为 [UltraData-SFT-2605](https://huggingface.co/datasets/openbmb/UltraData-SFT-2605)。随后针对数学、代码、闭卷问答和写作等方向训练专用 **RL teacher**,并通过 **On-Policy Distillation (OPD)** 将这些 teacher 的能力蒸馏回同一个发布模型。
|
| 94 |
+
|
| 95 |
+

|
| 96 |
+
|
| 97 |
+
### RL + OPD 带来了什么?
|
| 98 |
+
|
| 99 |
+
**RL + OPD** 是 MiniCPM5-1B 后训练中的关键环节。在数学、代码、指令跟随三类任务上,RL + OPD 将平均分提升 **↑16 分**,同时将回复触顶 max-tokens 预算的比例降低 **↓29 个百分点**。下方图示展示 Reasoning RL 两阶段流程、分数提升和超长率下降。
|
| 100 |
+
|
| 101 |
+
**RL** 阶段组合了推理、闭卷问答、写作、指令跟随、长上下文理解和通用对话等多类互补训练信号。Reasoning RL 基于 [DAPO-Math-17k](https://huggingface.co/datasets/BytedTsinghua-SIA/DAPO-Math-17k),在遵循 [JustRL](https://arxiv.org/pdf/2512.16649) 极简配方的基础上,进一步加入了两阶段长度调度,逐步降低超长率并提升推理准确率。我们还使用 [TriviaQA](https://huggingface.co/datasets/mandarjoshi/trivia_qa)、[NQ-Open](https://huggingface.co/datasets/google-research-datasets/nq_open)、[LongWriter-Zero-RLData](https://huggingface.co/datasets/THU-KEG/LongWriter-Zero-RLData)、合成可验证 RLVR 数据与 pair-wise RLHF 信号,提升可靠性、指令跟随和用户体验。
|
| 102 |
+
|
| 103 |
+

|
| 104 |
+
|
| 105 |
+
**OPD** 阶段参考 Thinking Machines Lab 的 [On-Policy Distillation](https://thinkingmachines.ai/blog/on-policy-distillation/) 思路,并结合 [Rethinking On-Policy Distillation](https://arxiv.org/pdf/2604.13016) 做了实现改进。我们在强化学习框架中使用反向 KL 散度作为优势估计值,替代原有的 verification-based advantage;同时在 response 序列的每个位置分别对学生模型和教师模型 logits 做双边 top-k 采样,取并集后计算反向 KL 散度,以平衡监督信号准确性和训练效率。OPD 直接复用各 RL teacher 训练时的同分布 prompt 作为蒸馏数据,无需额外构造语料。
|
| 106 |
+
|
| 107 |
+

|
| 108 |
+
|
| 109 |
+

|
| 110 |
+
|
| 111 |
+
## 快速上手
|
| 112 |
+
|
| 113 |
+
### vLLM
|
| 114 |
+
|
| 115 |
+
```bash
|
| 116 |
+
pip install "vllm>=0.21"
|
| 117 |
+
vllm serve openbmb/MiniCPM5-1B --port 8000
|
| 118 |
+
```
|
| 119 |
+
|
| 120 |
+
```bash
|
| 121 |
+
curl http://localhost:8000/v1/chat/completions \
|
| 122 |
+
-H "Content-Type: application/json" \
|
| 123 |
+
-d '{
|
| 124 |
+
"model": "openbmb/MiniCPM5-1B",
|
| 125 |
+
"messages": [{"role": "user", "content": "你是谁?可以简单介绍一下自己吗?"}],
|
| 126 |
+
"max_tokens": 128,
|
| 127 |
+
"temperature": 0.7
|
| 128 |
+
}'
|
| 129 |
+
```
|
| 130 |
+
|
| 131 |
+
### SGLang
|
| 132 |
+
|
| 133 |
+
```bash
|
| 134 |
+
pip install "sglang[srt]>=0.5.12"
|
| 135 |
+
python -m sglang.launch_server --model-path openbmb/MiniCPM5-1B --port 30000
|
| 136 |
+
```
|
| 137 |
+
|
| 138 |
+
```bash
|
| 139 |
+
curl http://localhost:30000/v1/chat/completions \
|
| 140 |
+
-H "Content-Type: application/json" \
|
| 141 |
+
-d '{
|
| 142 |
+
"model": "openbmb/MiniCPM5-1B",
|
| 143 |
+
"messages": [{"role": "user", "content": "你是谁?可以简单介绍一下自己吗?"}],
|
| 144 |
+
"max_tokens": 128,
|
| 145 |
+
"temperature": 0.7
|
| 146 |
+
}'
|
| 147 |
+
```
|
| 148 |
+
|
| 149 |
+
### Transformers
|
| 150 |
+
|
| 151 |
+
```bash
|
| 152 |
+
pip install -U "transformers>=5.6" accelerate torch
|
| 153 |
+
```
|
| 154 |
+
|
| 155 |
+
```python
|
| 156 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
| 157 |
+
|
| 158 |
+
model_id = "openbmb/MiniCPM5-1B"
|
| 159 |
+
tokenizer = AutoTokenizer.from_pretrained(model_id)
|
| 160 |
+
model = AutoModelForCausalLM.from_pretrained(
|
| 161 |
+
model_id,
|
| 162 |
+
torch_dtype="auto",
|
| 163 |
+
device_map="auto",
|
| 164 |
+
)
|
| 165 |
+
|
| 166 |
+
messages = [{"role": "user", "content": "你是谁?可以简单介绍一下自己吗?"}]
|
| 167 |
+
inputs = tokenizer.apply_chat_template(
|
| 168 |
+
messages,
|
| 169 |
+
tokenize=True,
|
| 170 |
+
add_generation_prompt=True,
|
| 171 |
+
enable_thinking=False,
|
| 172 |
+
return_tensors="pt",
|
| 173 |
+
).to(model.device)
|
| 174 |
+
|
| 175 |
+
outputs = model.generate(inputs, max_new_tokens=128)
|
| 176 |
+
print(tokenizer.decode(outputs[0][inputs.shape[-1]:], skip_special_tokens=True))
|
| 177 |
+
```
|
| 178 |
+
|
| 179 |
+
推荐的 chat template 采样参数:
|
| 180 |
+
|
| 181 |
+
| 模式 | 推荐采样参数 | 启用方式 |
|
| 182 |
+
| --- | --- | --- |
|
| 183 |
+
| **Think** | `temperature=0.9, top_p=0.95` | `enable_thinking=True` |
|
| 184 |
+
| **No Think** | `temperature=0.7, top_p=0.95` | `enable_thinking=False` |
|
| 185 |
+
|
| 186 |
+
## 工具调用
|
| 187 |
+
|
| 188 |
+
工具调用**推荐使用 SGLang**。MiniCPM5-1B 以 XML 格式产出工具调用,SGLang 内置的 `minicpm5` parser 会自动将其转换为 OpenAI 兼容的 `tool_calls` 字段。
|
| 189 |
+
|
| 190 |
+
```bash
|
| 191 |
+
python -m sglang.launch_server --model-path openbmb/MiniCPM5-1B --port 30000 \
|
| 192 |
+
--tool-call-parser minicpm5 # 或:--tool-call-parser auto
|
| 193 |
+
```
|
| 194 |
+
|
| 195 |
+
## GitHub Cookbooks 与 Agent Skills
|
| 196 |
+
|
| 197 |
+
MiniCPM5-1B 使用**标准 `LlamaForCausalLM` 架构**,主流推理引擎可直接加载,**无需自定义算子,也无模型代码 fork**。逐步部署和微调说明请参考下方 GitHub cookbooks;Agent Skills 作为 GitHub 资源提供给使用 Cursor / Claude Code 类 coding agent 的用户。
|
| 198 |
+
|
| 199 |
+
### 部署
|
| 200 |
+
|
| 201 |
+
| 后端 | 模型格式 / 适用场景 | Cookbook | Agent Skill |
|
| 202 |
+
| --- | --- | --- | --- |
|
| 203 |
+
| Transformers | BF16 / FP16,本地 Python 推理,GPU + CPU | [transformers.md](https://github.com/OpenBMB/MiniCPM/blob/main/docs/deployment/transformers.md) | [minicpm5-deploy-transformers](https://github.com/OpenBMB/MiniCPM/blob/main/skills/minicpm5-deploy-transformers/SKILL.md) |
|
| 204 |
+
| vLLM | BF16 / FP16 OpenAI server | [vllm.md](https://github.com/OpenBMB/MiniCPM/blob/main/docs/deployment/vllm.md) | [minicpm5-deploy-vllm](https://github.com/OpenBMB/MiniCPM/blob/main/skills/minicpm5-deploy-vllm/SKILL.md) |
|
| 205 |
+
| SGLang | BF16 / FP16 OpenAI server,推荐用于 tool calling | [sglang.md](https://github.com/OpenBMB/MiniCPM/blob/main/docs/deployment/sglang.md) | [minicpm5-deploy-sglang](https://github.com/OpenBMB/MiniCPM/blob/main/skills/minicpm5-deploy-sglang/SKILL.md) |
|
| 206 |
+
| llama.cpp | GGUF,CPU/GPU 本地推理 | [llama_cpp.md](https://github.com/OpenBMB/MiniCPM/blob/main/docs/deployment/llama_cpp.md) | [minicpm5-deploy-llama-cpp](https://github.com/OpenBMB/MiniCPM/blob/main/skills/minicpm5-deploy-llama-cpp/SKILL.md) |
|
| 207 |
+
| Ollama | GGUF,本地端侧运行 | [ollama.md](https://github.com/OpenBMB/MiniCPM/blob/main/docs/deployment/ollama.md) | [minicpm5-deploy-ollama](https://github.com/OpenBMB/MiniCPM/blob/main/skills/minicpm5-deploy-ollama/SKILL.md) |
|
| 208 |
+
| LM Studio | GGUF,Mac 桌面应用与 OpenAI server | [lmstudio.md](https://github.com/OpenBMB/MiniCPM/blob/main/docs/deployment/lmstudio.md) | [minicpm5-deploy-lmstudio](https://github.com/OpenBMB/MiniCPM/blob/main/skills/minicpm5-deploy-lmstudio/SKILL.md) |
|
| 209 |
+
| MLX | MLX / 4bit,Apple Silicon 本地推理 | [mlx.md](https://github.com/OpenBMB/MiniCPM/blob/main/docs/deployment/mlx.md) | [minicpm5-deploy-mlx](https://github.com/OpenBMB/MiniCPM/blob/main/skills/minicpm5-deploy-mlx/SKILL.md) |
|
| 210 |
+
| ArcLight | GGUF 本地端侧 / CPU / 桌面 / 服务器 | [arclight.md](https://github.com/OpenBMB/MiniCPM/blob/main/docs/deployment/arclight.md) | [minicpm5-deploy-arclight](https://github.com/OpenBMB/MiniCPM/blob/main/skills/minicpm5-deploy-arclight/SKILL.md) |
|
| 211 |
+
|
| 212 |
+
### 微调
|
| 213 |
+
|
| 214 |
+
| 框架 | 适用场景 | Cookbook | Agent Skill |
|
| 215 |
+
| --- | --- | --- | --- |
|
| 216 |
+
| TRL + PEFT | LoRA / SFT 微调 | [trl.md](https://github.com/OpenBMB/MiniCPM/blob/main/docs/finetune/trl.md) | [minicpm5-finetune-trl](https://github.com/OpenBMB/MiniCPM/blob/main/skills/minicpm5-finetune-trl/SKILL.md) |
|
| 217 |
+
| LLaMA-Factory | 微调 | [llamafactory.md](https://github.com/OpenBMB/MiniCPM/blob/main/docs/finetune/llamafactory.md) | [minicpm5-finetune-llamafactory](https://github.com/OpenBMB/MiniCPM/blob/main/skills/minicpm5-finetune-llamafactory/SKILL.md) |
|
| 218 |
+
| ms-swift | 微调 | [ms_swift.md](https://github.com/OpenBMB/MiniCPM/blob/main/docs/finetune/ms_swift.md) | [minicpm5-finetune-ms-swift](https://github.com/OpenBMB/MiniCPM/blob/main/skills/minicpm5-finetune-ms-swift/SKILL.md) |
|
| 219 |
+
| unsloth | 微调 | [unsloth.md](https://github.com/OpenBMB/MiniCPM/blob/main/docs/finetune/unsloth.md) | [minicpm5-finetune-unsloth](https://github.com/OpenBMB/MiniCPM/blob/main/skills/minicpm5-finetune-unsloth/SKILL.md) |
|
| 220 |
+
| xtuner | 微调 | [xtuner.md](https://github.com/OpenBMB/MiniCPM/blob/main/docs/finetune/xtuner.md) | [minicpm5-finetune-xtuner](https://github.com/OpenBMB/MiniCPM/blob/main/skills/minicpm5-finetune-xtuner/SKILL.md) |
|
| 221 |
+
|
| 222 |
+
### 其他支持的框架
|
| 223 |
+
|
| 224 |
+
除上文列出的部署与微调框架外,MiniCPM5-1B 也支持通过 FlagOS 进行多芯片部署。
|
| 225 |
+
|
| 226 |
+
#### FlagOS 介绍
|
| 227 |
+
|
| 228 |
+
为解决不同 AI 芯片大规模落地应用,北京智源研究院联合众多科研机构、芯片企业、系统厂商、算法和软件相关单位等国内外机构共同发起并创立了 FlagOS 开源社区。
|
| 229 |
+
|
| 230 |
+
FlagOS 社区致力于打造面向多种 AI 芯片的统一、开源的系统软件栈,包括大型算子库、统一AI编译器、并行训推框架、统一通信库等核心开源项目,构建「模型-系统-芯片」三层贯通的开放技术生态,通过“一次开发跨芯迁移”释放硬件计算潜力,打破不同芯片软件栈之间生态隔离,有效降低开发者的迁移成本。FlagOS 社区构建人工智能软硬件生态,突破单一闭源垄断,推动AI硬件技术大范围落地发展,立足中国、拥抱全球合作。
|
| 231 |
+
|
| 232 |
+
官网速递:[https://flagos.io](https://flagos.io/)
|
| 233 |
+
|
| 234 |
+
<details>
|
| 235 |
+
<summary>FlagOS 多 AI 芯片支持与使用方式</summary>
|
| 236 |
+
|
| 237 |
+
#### FlagOS 多 AI 芯片支持
|
| 238 |
+
|
| 239 |
+
基于 FlagOS 极短时间内适配 MiniCPM5-1B 到 9 种不同的 AI 芯片,得益于众智 FlagOS 的多芯片统一 AI 系统软件栈的能力。目前,在 FlagOS 团队构建的面向多架构人工智能芯片的大模型自动迁移、适配与发布平台 FlagRelease 上,已发布 MiniCPM5-1B 的多芯片版本。细节如下:
|
| 240 |
+
|
| 241 |
+
|Vendor|ModelScope|Huggingface|
|
| 242 |
+
|---|---|---|
|
| 243 |
+
|Nvidia|[MiniCPM5-1B-nvidia-FlagOS](https://www.modelscope.cn/models/FlagRelease/MiniCPM5-1B-nvidia-FlagOS)|[MiniCPM5-1B-nvidia-FlagOS](https://huggingface.co/FlagRelease/MiniCPM5-1B-nvidia-FlagOS)|
|
| 244 |
+
|Hygon|[MiniCPM5-1B-hygon-FlagOS](https://www.modelscope.cn/models/FlagRelease/MiniCPM5-1B-hygon-FlagOS)|[MiniCPM5-1B-hygon-FlagOS](https://huggingface.co/FlagRelease/MiniCPM5-1B-hygon-FlagOS)|
|
| 245 |
+
|Metax|[MiniCPM5-1B-metax-FlagOS](https://www.modelscope.cn/models/FlagRelease/MiniCPM5-1B-metax-FlagOS)|[MiniCPM5-1B-metax-FlagOS](https://huggingface.co/FlagRelease/MiniCPM5-1B-metax-FlagOS)|
|
| 246 |
+
|Iluvatar|[MiniCPM5-1B-iluvatar-FlagOS](https://www.modelscope.cn/models/FlagRelease/MiniCPM5-1B-iluvatar-FlagOS)|[MiniCPM5-1B-iluvatar-FlagOS](https://huggingface.co/FlagRelease/MiniCPM5-1B-iluvatar-FlagOS)|
|
| 247 |
+
|Zhenwu|[MiniCPM5-1B-zhenwu-FlagOS](https://www.modelscope.cn/models/FlagRelease/MiniCPM5-1B-zhenwu-FlagOS)|[MiniCPM5-1B-zhenwu-FlagOS](https://huggingface.co/FlagRelease/MiniCPM5-1B-zhenwu-FlagOS)|
|
| 248 |
+
|Mthreads|[MiniCPM5-1B-mthreads-FlagOS](https://www.modelscope.cn/models/FlagRelease/MiniCPM5-1B-mthreads-FlagOS)|[MiniCPM5-1B-mthreads-FlagOS](https://huggingface.co/FlagRelease/MiniCPM5-1B-mthreads-FlagOS)|
|
| 249 |
+
|Kunlunxin|[MiniCPM5-1B-kunlunxin-FlagOS](https://www.modelscope.cn/models/FlagRelease/MiniCPM5-1B-kunlunxin-FlagOS)|[MiniCPM5-1B-kunlunxin-FlagOS](https://huggingface.co/FlagRelease/MiniCPM5-1B-kunlunxin-FlagOS)|
|
| 250 |
+
|Ascend|[MiniCPM5-1B-ascend-FlagOS](https://modelscope.cn/models/FlagRelease/MiniCPM5-1B-ascend-FlagOS)|[MiniCPM5-1B-ascend-FlagOS](https://huggingface.co/FlagRelease/MiniCPM5-1B-ascend-FlagOS)|
|
| 251 |
+
|ARM-v9|[MiniCPM5-1B-Armv9-FlagOS](https://modelscope.cn/models/FlagRelease/MiniCPM5-1B-Armv9-FlagOS)|[MiniCPM5-1B-Armv9-FlagOS](https://huggingface.co/FlagRelease/MiniCPM5-1B-Armv9-FlagOS)|
|
| 252 |
+
|
| 253 |
+
#### FlagOS 使用方式
|
| 254 |
+
|
| 255 |
+
##### 使用 FlagOS 在 Nvidia 体验性能加速
|
| 256 |
+
|
| 257 |
+
###### From FlagRelease(**推荐**)
|
| 258 |
+
|
| 259 |
+
FlagRelease是FlagOS团队构建的一套面向多架构人工智能芯片的大模型自动迁移、适配与发布平台,已发布MiniCPM-1B的多芯片版本。FlagRelase已内置相关软件包,无需用户安装。
|
| 260 |
+
|
| 261 |
+
###### FlagRelease 镜像关键版本信息
|
| 262 |
+
|
| 263 |
+
###### FlagRelease 使用速递
|
| 264 |
+
|
| 265 |
+
|Vendor|ModelScope|Huggingface|
|
| 266 |
+
|---|---|---|
|
| 267 |
+
|Nvidia|[MiniCPM5-1B-nvidia-FlagOS](https://www.modelscope.cn/models/FlagRelease/MiniCPM5-1B-nvidia-FlagOS)|[MiniCPM5-1B-nvidia-FlagOS](https://huggingface.co/FlagRelease/MiniCPM5-1B-nvidia-FlagOS)|
|
| 268 |
+
|Hygon|[MiniCPM5-1B-hygon-FlagOS](https://www.modelscope.cn/models/FlagRelease/MiniCPM5-1B-hygon-FlagOS)|[MiniCPM5-1B-hygon-FlagOS](https://huggingface.co/FlagRelease/MiniCPM5-1B-hygon-FlagOS)|
|
| 269 |
+
|Metax|[MiniCPM5-1B-metax-FlagOS](https://www.modelscope.cn/models/FlagRelease/MiniCPM5-1B-metax-FlagOS)|[MiniCPM5-1B-metax-FlagOS](https://huggingface.co/FlagRelease/MiniCPM5-1B-metax-FlagOS)|
|
| 270 |
+
|Iluvatar|[MiniCPM5-1B-iluvatar-FlagOS](https://www.modelscope.cn/models/FlagRelease/MiniCPM5-1B-iluvatar-FlagOS)|[MiniCPM5-1B-iluvatar-FlagOS](https://huggingface.co/FlagRelease/MiniCPM5-1B-iluvatar-FlagOS)|
|
| 271 |
+
|Zhenwu|[MiniCPM5-1B-zhenwu-FlagOS](https://www.modelscope.cn/models/FlagRelease/MiniCPM5-1B-zhenwu-FlagOS)|[MiniCPM5-1B-zhenwu-FlagOS](https://huggingface.co/FlagRelease/MiniCPM5-1B-zhenwu-FlagOS)|
|
| 272 |
+
|Mthreads|[MiniCPM5-1B-mthreads-FlagOS](https://www.modelscope.cn/models/FlagRelease/MiniCPM5-1B-mthreads-FlagOS)|[MiniCPM5-1B-mthreads-FlagOS](https://huggingface.co/FlagRelease/MiniCPM5-1B-mthreads-FlagOS)|
|
| 273 |
+
|Kunlunxin|[MiniCPM5-1B-kunlunxin-FlagOS](https://www.modelscope.cn/models/FlagRelease/MiniCPM5-1B-kunlunxin-FlagOS)|[MiniCPM5-1B-kunlunxin-FlagOS](https://huggingface.co/FlagRelease/MiniCPM5-1B-kunlunxin-FlagOS)|
|
| 274 |
+
|Ascend|[MiniCPM5-1B-ascend-FlagOS](https://modelscope.cn/models/FlagRelease/MiniCPM5-1B-ascend-FlagOS)|[MiniCPM5-1B-ascend-FlagOS](https://huggingface.co/FlagRelease/MiniCPM5-1B-ascend-FlagOS)|
|
| 275 |
+
|ARM-v9|[MiniCPM5-1B-Armv9-FlagOS](https://modelscope.cn/models/FlagRelease/MiniCPM5-1B-Armv9-FlagOS)|[MiniCPM5-1B-Armv9-FlagOS](https://huggingface.co/FlagRelease/MiniCPM5-1B-Armv9-FlagOS)|
|
| 276 |
+
|
| 277 |
+
###### 从零开始
|
| 278 |
+
|
| 279 |
+
- 依赖Python3.12, GLIBC_2.39, GLIBCXX_3.4.33, CXXABI_1.3.15 环境
|
| 280 |
+
|
| 281 |
+
###### Vllm 版本
|
| 282 |
+
|
| 283 |
+
###### 安装 FlagOS 算子库
|
| 284 |
+
|
| 285 |
+
官方仓库:https://github.com/flagos-ai/FlagGems
|
| 286 |
+
|
| 287 |
+
```PowerShell
|
| 288 |
+
pip install flag-gems==4.2.1rc0
|
| 289 |
+
pip install triton==3.5.1
|
| 290 |
+
```
|
| 291 |
+
|
| 292 |
+
###### 开启加速
|
| 293 |
+
|
| 294 |
+
通过在vllm执行推理的源码中增加flagGems的导入即可开启flagGems加速
|
| 295 |
+
|
| 296 |
+
```Bash
|
| 297 |
+
import flag_gems
|
| 298 |
+
flag_gems.enable(record=True, once=True, path="/root/gems.txt")
|
| 299 |
+
```
|
| 300 |
+
|
| 301 |
+
```Bash
|
| 302 |
+
vllm serve ${model_path} \
|
| 303 |
+
--trust-remote-code \
|
| 304 |
+
--dtype bfloat16 \
|
| 305 |
+
--enforce-eager \
|
| 306 |
+
--port ${Port} \
|
| 307 |
+
--served-model-name ${model_name} \
|
| 308 |
+
--gpu-memory-utilization 0.85
|
| 309 |
+
```
|
| 310 |
+
|
| 311 |
+
##### 使用 FlagOS 统一多芯片后端插件
|
| 312 |
+
|
| 313 |
+
**[vllm-plugin-FL](https://github.com/flagos-ai/vllm-plugin-FL)** 是一个为 **vLLM** 推理/服务框架构建的插件,它基于 **FlagOS 的统一多芯片后端**开发,旨在扩展 vLLM 在多种硬件环境下的功能和性能表现。
|
| 314 |
+
|
| 315 |
+
###### vllm-plugin-FL 使用
|
| 316 |
+
|
| 317 |
+
|厂商|从零开始|从 FlagRelease 开始||
|
| 318 |
+
|---|---|---|---|
|
| 319 |
+
|英伟达|[vllm-plugin-FL/MiniCPM5-1B](https://github.com/flagos-ai/vllm-plugin-FL/blob/main/examples/minicpm/README.md)|[MiniCPM5-1B-ModelScope](https://www.modelscope.cn/models/FlagRelease/MiniCPM5-1B-nvidia-FlagOS)|[MiniCPM5-1B-nvidia-FlagOS](https://huggingface.co/FlagRelease/MiniCPM5-1B-nvidia-FlagOS)|
|
| 320 |
+
|
| 321 |
+
</details>
|
| 322 |
+
|
| 323 |
+
## 桌宠
|
| 324 |
+
|
| 325 |
+
我们也发布了 **[OpenBMB/MiniCPM-Desk-Pet](https://github.com/OpenBMB/MiniCPM-Desk-Pet)**,一个由 MiniCPM5-1B 本地驱动的桌宠应用。它支持 Apple Silicon / NVIDIA GPU / CPU 路线,可以与 Cursor、Claude Code、Codex 等 coding agent 联动,并支持 LoRA 人格切换。
|
| 326 |
+
|
| 327 |
+
<a href="https://youtu.be/Ee0slMW8SEk"><img src="https://img.youtube.com/vi/Ee0slMW8SEk/0.jpg" alt="MiniCPM Desk Pet video demo" width="720"></a>
|
| 328 |
+
|
| 329 |
+
## 局限性与负责任使用
|
| 330 |
+
|
| 331 |
+
MiniCPM5-1B 是一个基于训练数据统计规律生成文本的语言模型,可能生成不准确、有偏见或不安全的内容。在高风险场景中使用前,应对模型输出进行审查和验证。
|
| 332 |
+
|
| 333 |
+
用户需要自行评估模型输出,配置必要的安全防护,并遵守适用法律法规和平台政策。
|
| 334 |
+
|
| 335 |
+
## 开源协议
|
| 336 |
+
|
| 337 |
+
MiniCPM 模型权重与相关代码依照 [Apache-2.0](https://github.com/OpenBMB/MiniCPM/blob/main/LICENSE) 协议发布。
|
| 338 |
+
|
| 339 |
+
## 引用
|
| 340 |
+
|
| 341 |
+
如果觉得我们的工作有帮助,请引用:
|
| 342 |
+
|
| 343 |
+
```bibtex
|
| 344 |
+
@article{minicpm4,
|
| 345 |
+
title={Minicpm4: Ultra-efficient llms on end devices},
|
| 346 |
+
author={MiniCPM, Team},
|
| 347 |
+
journal={arXiv preprint arXiv:2506.07900},
|
| 348 |
+
year={2025}
|
| 349 |
+
}
|
| 350 |
+
```
|