install monkyocr - zhongguogu/Linux GitHub Wiki

安装驱动

` sudo apt update

sudo apt install nvidia-driver-535 # 535

下载anaconda

wget https://repo.anaconda.com/archive/Anaconda3-2024.06-1-Linux-x86_64.sh

conda create -n MonkeyOCR python=3.10

conda activate MonkeyOCR

git clone https://github.com/Yuliang-Liu/MonkeyOCR.git //如果下载不了，可以本地下载

cd MonkeyOCR

遇到了下面的问题，是没有安装CUDA 开发工具包

` ERROR: Operation cancelled by user (MonkeyOCR) ubuntu@10-60-51-1:~/project/MonkeyOCR-main$ pip install flash-attn==2.7.4.post1 --no-build-isolation -i https://pypi.tuna.tsinghua.edu.cn/simple Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple Collecting flash-attn==2.7.4.post1 Downloading https://pypi.tuna.tsinghua.edu.cn/packages/11/34/9bf60e736ed7bbe15055ac2dab48ec67d9dbd088d2b4ae318fd77190ab4e/flash_attn-2.7.4.post1.tar.gz (6.0 MB) ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 6.0/6.0 MB 28.8 MB/s eta 0:00:00 Preparing metadata (setup.py) ... error error: subprocess-exited-with-error

× python setup.py egg_info did not run successfully. │ exit code: 1 ╰─> [19 lines of output] /tmp/pip-install-xmjmyejk/flash-attn_883f013ce0da43608d5bedd0669881f3/setup.py:106: UserWarning: flash_attn was requested, but nvcc was not found. Are you sure your environment has nvcc available? If you're installing within a container from https://hub.docker.com/r/pytorch/pytorch, only images whose names contain 'devel' will provide nvcc. warnings.warn( Traceback (most recent call last): File "", line 2, in File "", line 35, in File "/tmp/pip-install-xmjmyejk/flash-attn_883f013ce0da43608d5bedd0669881f3/setup.py", line 198, in CUDAExtension( File "/home/ubuntu/anaconda3/envs/MonkeyOCR/lib/python3.10/site-packages/torch/utils/cpp_extension.py", line 1078, in CUDAExtension library_dirs += library_paths(cuda=True) File "/home/ubuntu/anaconda3/envs/MonkeyOCR/lib/python3.10/site-packages/torch/utils/cpp_extension.py", line 1209, in library_paths if (not os.path.exists(_join_cuda_home(lib_dir)) and File "/home/ubuntu/anaconda3/envs/MonkeyOCR/lib/python3.10/site-packages/torch/utils/cpp_extension.py", line 2416, in _join_cuda_home raise OSError('CUDA_HOME environment variable is not set. ' OSError: CUDA_HOME environment variable is not set. Please set it to your CUDA install root.

  torch.__version__  = 2.5.1+cu124
 `

执行python demo/demo_gradio.py

WARNING ⚠️ model_weight/Structure/doclayout_yolo_docstructbench_imgsz1280_2501.pt appears to require 'dill', which is not in doclayout_yolo requirements.

AutoInstall will run now for 'dill' but this feature will be removed in the future.

Recommend fixes are to train a new model using the latest 'doclayout_yolo' package or to run a command with an official YOLOv8 model, i.e. 'yolo predict model=yolov8n.pt'

requirements: Ultralytics requirement ['dill'] not found, attempting AutoUpdate...

你遇到的警告信息表明，doclayout_yolo 模型需要 dill 库，但这个库不在它的标准依赖列表中。Ultralytics 框架正在自动尝试安装这个缺失的依赖，但这是一个临时解决方案，未来版本可能会移除这种自动安装的功能。 解决方案 pip install dill -i https://pypi.tuna.tsinghua.edu.cn/simple

  1. 安装 CUDA 开发工具包 首先需要安装 CUDA Toolkit：

bash

安装 CUDA 开发工具包

sudo apt install nvidia-cuda-toolkit

验证安装

nvcc --version

Install pytorch, see https://pytorch.org/get-started/previous-versions/ for your cuda version

pip install torch==2.5.1 torchvision==0.20.1 torchaudio==2.5.1 --index-url https://download.pytorch.org/whl/cu124 -i https://pypi.tuna.tsinghua.edu.cn/simple

pip install . -i https://pypi.tuna.tsinghua.edu.cn/simple

安装模型

pip install huggingface_hub -i https://pypi.tuna.tsinghua.edu.cn/simple

python download_model.py

pip install modelscope -i https://pypi.tuna.tsinghua.edu.cn/simple

python download_model.py -t modelscope

Gradio Demo页面

Prepare your env for gradio

pip install gradio==5.23.3 -i https://pypi.tuna.tsinghua.edu.cn/simple

pip install pdf2image==1.17.0 -i https://pypi.tuna.tsinghua.edu.cn/simple

sudo apt-get update

sudo apt-get install poppler-utils #依赖poppler

`

1、安装NVIDIA 方法 1：使用 Ubuntu 软件源安装 添加官方 NVIDIA 驱动 PPA：

bash sudo apt update sudo apt install nvidia-driver-535 # 535是当前最新的稳定版本，可能需要根据你的显卡兼容性选择其他版本

2、安装conda ubuntu@10-60-51-1:~/software$ wget https://repo.anaconda.com/archive/Anaconda3-2024.06-1-Linux-x86_64.sh --2025-06-11 09:26:58-- https://repo.anaconda.com/archive/Anaconda3-2024.06-1-Linux-x86_64.sh Resolving repo.anaconda.com (repo.anaconda.com)... 100.65.138.175 Connecting to repo.anaconda.com (repo.anaconda.com)|100.65.138.175|:443... connected. HTTP request sent, awaiting response... 200 OK Length: 1056829859 (1008M) [application/octet-stream] Saving to: ‘Anaconda3-2024.06-1-Linux-x86_64.sh’

Anaconda3-2024.06-1-Linux-x86_64.sh 100%[=================================================================================================================================================================>] 1008M 2.69MB/s in 5m 42s

2025-06-11 09:32:41 (2.95 MB/s) - ‘Anaconda3-2024.06-1-Linux-x86_64.sh’ saved [1056829859/1056829859]

删除 404/无效的 channel 全部换成清华或中科大等完善的国内镜像，尤其是 main/free/r/msys2/cloud 不要有 repo.anaconda.com 和 anaconda/cloud/pytorch 这种不存在的路径 清理缓存再试 如还有问题，把 conda config --show channels 的输出发出来

(base) ubuntu@10-60-51-1:~$ conda config --show channels channels:

bash conda config --remove channels defaults

⚠️ **GitHub.com Fallback** ⚠️