README_Windows_CUDA_Acceleration_en_US.md 2.38 KB
Newer Older
1
2
3
4
# Windows 10/11

### 1. Install CUDA and cuDNN

5
You need to install a CUDA version that is compatible with torch's requirements. Currently, torch supports CUDA 11.8/12.4/12.6.
6

7
8
9
- CUDA 11.8 https://developer.nvidia.com/cuda-11-8-0-download-archive
- CUDA 12.4 https://developer.nvidia.com/cuda-12-4-0-download-archive
- CUDA 12.6 https://developer.nvidia.com/cuda-12-6-0-download-archive
10
11
12
13
14
15
16
17
18

### 2. Install Anaconda

If Anaconda is already installed, you can skip this step.

Download link: https://repo.anaconda.com/archive/Anaconda3-2024.06-1-Windows-x86_64.exe

### 3. Create an Environment Using Conda

19
```bash
20
conda create -n mineru 'python>=3.10' -y
21
conda activate mineru
22
23
24
25
26
```

### 4. Install Applications

```
27
pip install -U magic-pdf[full]
28
29
```

30
> [!IMPORTANT]
31
> After installation, you can check the version of `magic-pdf` using the following command:
32
33
34
35
>
> ```bash
> magic-pdf --version
> ```
36

37
38
39
40
41
42
43
44
45
46

### 5. Download Models

Refer to detailed instructions on [how to download model files](how_to_download_models_en.md).

### 6. Understand the Location of the Configuration File

After completing the [5. Download Models](#5-download-models) step, the script will automatically generate a `magic-pdf.json` file in the user directory and configure the default model path.
You can find the `magic-pdf.json` file in your 【user directory】 .

47
> [!TIP]
48
49
50
51
52
53
54
> The user directory for Windows is "C:/Users/username".

### 7. First Run

Download a sample file from the repository and test it.

```powershell
55
  wget https://github.com/opendatalab/MinerU/raw/master/demo/pdfs/small_ocr.pdf -O small_ocr.pdf
56
  magic-pdf -p small_ocr.pdf -o ./output
57
58
59
60
```

### 8. Test CUDA Acceleration

61
If your graphics card has at least 6GB of VRAM, follow these steps to test CUDA-accelerated parsing performance.
62

63
1. **Overwrite the installation of torch and torchvision** supporting CUDA.(Please select the appropriate index-url based on your CUDA version. For more details, refer to the [PyTorch official website](https://pytorch.org/get-started/locally/).)
64
65

   ```
66
   pip install --force-reinstall torch torchvision --index-url https://download.pytorch.org/whl/cu124
67
68
69
70
71
72
73
74
75
76
   ```

2. **Modify the value of `"device-mode"`** in the `magic-pdf.json` configuration file located in your user directory.

   ```json
   {
     "device-mode": "cuda"
   }
   ```

77

78
79
80
3. **Run the following command to test CUDA acceleration**:

   ```
81
   magic-pdf -p small_ocr.pdf -o ./output
82
   ```