ComfyUI搭建与文生图运行

原创

平常心

发布于 2024-08-06 18:06:07

2400

发布于 2024-08-06 18:06:07

文章被收录于专栏：个人总结系列

一、ComfyUI本地环境搭建

1、git仓库下载

gpt:~/llms_dir$ git clone https://github.com/comfyanonymous/ComfyUI.git

2、python环境准备

gpt:~/llms_dir$ sudo apt update
gpt:~/llms_dir$ sudo apt install python3 python3-pip git

3、启动ComfyUI

gpt:~/llms_dir/ComfyUI$ python3 -m venv venv
gpt:~/llms_dir/ComfyUI$ source venv/bin/activate
(venv) gpt:~/llms_dir/ComfyUI$ pip install -r requirements.txt
(venv) gpt:~/llms_dir/ComfyUI$ python main.py localhost

Starting server
To see the GUI go to: http://127.0.0.1:8188

通过http://localhost:8188/ 就可以成功访问。

二、文生图模型运行

1、模型下载

如https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0 模型

目录下载到指定目录
gpt:~/llms_dir/ComfyUI/models/checkpoints$ wget https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/resolve/main/sd_xl_base_1.0.safetensors?download=true .
②、comfyui加载对应模型

加载模型并点击Queue Prompt

got prompt
model_type EPS
Using pytorch attention in VAE
loaded straight to GPU
Requested to load SDXL
Loading 1 new model
Requested to load SDXLClipModel
Loading 1 new model
Requested to load SDXL
Loading 1 new model
100%|███████████████████████████████████████████████████████████████████████████████████| 20/20 [00:03<00:00,  5.09it/s]
Requested to load AutoencoderKL
Loading 1 new model
Prompt executed in 102.37 seconds

界面显示文生图结果，更改提示词为 A boy running on the beach，对应生成效果为：