nvidia-smi
If you see information similar to the following, it means that the NVIDIA drivers are already installed, and you can skip Step 2.
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 537.34 Driver Version: 537.34 CUDA Version: 12.2 |
|-----------------------------------------+----------------------+----------------------+
| GPU Name TCC/WDDM | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|=========================================+======================+======================|
| 0 NVIDIA GeForce RTX 3060 Ti WDDM | 00000000:01:00.0 On | N/A |
| 0% 51C P8 12W / 200W | 1489MiB / 8192MiB | 5% Default |
| | | N/A |
+-----------------------------------------+----------------------+----------------------+
If no driver is installed, use the following command:
sudo apt-get update
sudo apt-get install nvidia-driver-545
Install the proprietary driver and restart your computer after installation.
reboot
If Anaconda is already installed, skip this step.
wget https://repo.anaconda.com/archive/Anaconda3-2024.06-1-Linux-x86_64.sh
bash Anaconda3-2024.06-1-Linux-x86_64.sh
In the final step, enter yes
, close the terminal, and reopen it.
Specify Python version 3.10.
conda create -n MinerU python=3.10
conda activate MinerU
pip install -U magic-pdf[full] --extra-index-url https://wheels.myhloli.com
❗ After installation, make sure to check the version of magic-pdf
using the following command:
magic-pdf --version
If the version number is less than 0.7.0, please report the issue.
Refer to detailed instructions on how to download model files.
After downloading, move the models
directory to an SSD with more space.
❗ After downloading the models, ensure they are complete:
- Check that the file sizes match the description on the website.
- If possible, verify the integrity using SHA256.
Obtain the configuration template file magic-pdf.template.json
from the root directory of the repository.
❗ Execute the following command to copy the configuration file to your home directory, otherwise the program will not run:
wget https://github.com/opendatalab/MinerU/raw/master/magic-pdf.template.json
cp magic-pdf.template.json ~/magic-pdf.json
Find the magic-pdf.json
file in your home directory and configure "models-dir"
to be the directory where the model weights from Step 6 were downloaded.
❗ Correctly specify the absolute path of the directory containing the model weights; otherwise, the program will fail due to missing model files.
{
"models-dir": "/tmp/models"
}
Download a sample file from the repository and test it.
wget https://github.com/opendatalab/MinerU/raw/master/demo/small_ocr.pdf
magic-pdf -p small_ocr.pdf
If your graphics card has at least 8GB of VRAM, follow these steps to test CUDA acceleration:
- Modify the value of
"device-mode"
in themagic-pdf.json
configuration file located in your home directory.{ "device-mode": "cuda" }
- Test CUDA acceleration with the following command:
magic-pdf -p small_ocr.pdf
❗ The following operations require a graphics card with at least 16GB of VRAM; otherwise, the program may crash or experience reduced performance.
- Download
paddlepaddle-gpu
. Installation will automatically enable OCR acceleration.python -m pip install paddlepaddle-gpu==3.0.0b1 -i https://www.paddlepaddle.org.cn/packages/stable/cu118/
- Test OCR acceleration with the following command:
magic-pdf -p small_ocr.pdf