Skip to main content
Real-ESRGAN uses the DF2K (DIV2K + Flickr2K) and OST datasets for training. This guide walks through downloading and preparing these datasets.

​
Download Datasets

Only high-resolution (HR) images are required. Download the following datasets:

DIV2K

High-quality diverse imagesDownload DIV2K_train_HR.zip

Flickr2K

Flickr image collectionDownload Flickr2K.tar

OST

OpenImages subsetDownload OST_dataset.zip
Extract all datasets to your datasets/ directory. Combine DIV2K and Flickr2K into a DF2K folder.

​
Preparation Steps

The dataset preparation involves three steps. Steps 1 and 2 are optional but recommended for optimal training performance.

​
Step 1: Generate Multi-scale Images (Optional)

For the DF2K dataset, use a multi-scale strategy to create ground-truth images at different scales. This improves the model’s ability to handle various image sizes.
This step can be omitted for a quick trial, but is recommended for full training.
The script generates images at multiple scales (0.75x, 0.5x, 0.33x) and a version where the shortest edge is 400 pixels: 
python scripts/generate_multiscale_DF2K.py --input datasets/DF2K/DF2K_HR --output datasets/DF2K/DF2K_multiscale
The generate_multiscale_DF2K.py script:
  • Downsamples HR images to 75%, 50%, and 33% of original size
  • Creates a version with shortest edge = 400 pixels
  • Uses LANCZOS resampling for high quality
  • Saves images with suffixes like T0, T1, T2, T3
This provides training data at multiple resolutions, helping the model generalize better.

​
Step 2: Crop to Sub-images (Optional)

Crop images into sub-images for faster IO and processing during training. This is especially helpful when training on slower storage.
Skip this step if you have fast IO or limited disk space.
python scripts/extract_subimages.py --input datasets/DF2K/DF2K_multiscale --output datasets/DF2K/DF2K_multiscale_sub --crop_size 400 --step 200

​
Parameters

​
input
string
required
Input folder containing images to crop
​
output
string
required
Output folder for cropped sub-images
​
crop_size
int
default:"480"
Size of each cropped sub-image (e.g., 400x400 pixels)
​
step
int
default:"240"
Step size for overlapped sliding window. Smaller values create more sub-images with more overlap.
​
thresh_size
int
default:"0"
Minimum size threshold. Patches smaller than this are dropped.
​
n_thread
int
default:"20"
Number of parallel threads for processing
  • crop_size: The dimension of each sub-image patch (e.g., 400 means 400x400)
  • step: How much the sliding window moves. A step of 200 with crop_size 400 means 50% overlap
  • More overlap = more training data but requires more disk space
Example: For a 1200x1200 image with crop_size=400 and step=200:
  • Creates 5x5 = 25 overlapping sub-images
  • Each sub-image is 400x400 pixels

​
Step 3: Generate Meta Info File (Required)

Create a text file containing all image paths. This file tells the training script which images to use. 
python scripts/generate_meta_info.py --input datasets/DF2K/DF2K_HR datasets/DF2K/DF2K_multiscale --root datasets/DF2K datasets/DF2K --meta_info datasets/DF2K/meta_info/meta_info_DF2Kmultiscale.txt

​
Parameters

​
input
list
required
List of input folders containing images. Can specify multiple folders separated by spaces.
​
root
list
required
Corresponding root paths for each input folder. Must have same length as --input.
​
meta_info
string
required
Output path for the generated meta info text file
​
check
boolean
default:"false"
Read each image to verify it can be loaded without errors

​
Meta Info Format

The generated file contains relative paths, one per line: 
DF2K_HR_sub/000001_s001.png
DF2K_HR_sub/000001_s002.png
DF2K_HR_sub/000001_s003.png
DF2K_multiscale/000001T0.png
DF2K_multiscale/000001T1.png
...
Each user’s meta info file will be different based on how images were cropped and processed. You must generate your own file - do not use example files from the repository.

​
Combining Multiple Datasets

You can merge multiple folders into a single meta info file: 
python scripts/generate_meta_info.py \
  --input datasets/DF2K/DF2K_multiscale_sub datasets/OST \
  --root datasets/DF2K datasets/OST \
  --meta_info datasets/meta_info/meta_info_DF2Kmultiscale+OST_sub.txt
This creates a combined dataset from both DF2K and OST images.

​
Validation Dataset (Optional)

If you want to validate during training, prepare a separate validation dataset with paired low-quality and high-quality images.
  1. Prepare validation image pairs:
    • Ground-truth (GT) folder: High-resolution images
    • Low-quality (LQ) folder: Corresponding low-resolution images
  2. Uncomment validation sections in your training config:
val:
  name: validation
  type: PairedImageDataset
  dataroot_gt: path_to_gt
  dataroot_lq: path_to_lq
  io_backend:
    type: disk
  1. Configure validation frequency and metrics:
val:
  val_freq: !!float 5e3
  save_img: true
  metrics:
    psnr:
      type: calculate_psnr
      crop_border: 4
      test_y_channel: false

​
Directory Structure

After preparation, your directory structure should look like: 
datasets/
├── DF2K/
│   ├── DF2K_HR/                    # Original HR images
│   ├── DF2K_multiscale/            # Multi-scale images
│   ├── DF2K_multiscale_sub/        # Cropped sub-images
│   └── meta_info/
│       └── meta_info_DF2Kmultiscale.txt
├── OST/                            # OST dataset
└── meta_info/
    └── meta_info_DF2Kmultiscale+OST_sub.txt

​
Next Steps

Train Real-ESRNet

Begin stage 1 training with your prepared dataset

Build docs developers (and LLMs) love

Get started for freeTalk to us