搭建一个本地DeepSeek r1 14B配置低到惊喜！附超简单步骤

发现只需要11G显存！只使用一张rtx3060 12G的就搞定了。速度和答案质量还不错，太令人惊喜了！

 nvidia-smi -l
Thu Feb 13 10:22:25 2025       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 525.89.02    Driver Version: 525.89.02    CUDA Version: 12.0     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  NVIDIA GeForce ...  Off  | 00000000:02:00.0 Off |                  N/A |
| 31%   59C    P2    66W / 170W |  10899MiB / 12288MiB |     51%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|    0   N/A  N/A      2154      G   /usr/lib/xorg/Xorg                 14MiB |
|    0   N/A  N/A     23463      C   ...1_avx/ollama_llama_server    10880MiB |
+-----------------------------------------------------------------------------+

7B只需要5GB的显存

+-----------------------------------------------------------------------------+
| NVIDIA-SMI 525.89.02    Driver Version: 525.89.02    CUDA Version: 12.0     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  NVIDIA GeForce ...  Off  | 00000000:02:00.0 Off |                  N/A |
| 30%   58C    P2   168W / 170W |   5689MiB / 12288MiB |     84%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|    0   N/A  N/A      2154      G   /usr/lib/xorg/Xorg                 14MiB |
|    0   N/A  N/A     19963      C   ...1_avx/ollama_llama_server     5670MiB |
+-----------------------------------------------------------------------------+

搭建步骤：

1. 安装ollma

curl -fsSL https://ollama.com/install.sh | sh

2.下载安装14B模型

 ollama pull deepseek-r1:14B

3.安装open-webui

conda create -n open-webui python=3.11
conda activate open-webui
 pip install open-webu

4.启动open-webui

open-webui serve

5. 浏览器访问机器的8080，设置管理员密码，选择开放模型

6.完工，选择deepseek-r1:14B，开始聊！