Skip to main content
Real-ESRGAN uses a two-stage training process to achieve high-quality image super-resolution. This approach combines the stability of L1 loss with the perceptual quality improvements from GAN training.

​
Training Process

The training is divided into two distinct stages that share the same data synthesis process and training pipeline, but differ in their loss functions:
1

Stage 1: Train Real-ESRNet

Train Real-ESRNet using L1 loss from a pre-trained ESRGAN model. This stage provides a stable foundation and prevents mode collapse.
  • Uses L1 loss only
  • Starts from pre-trained ESRGAN weights
  • Results in a stable base model
2

Stage 2: Train Real-ESRGAN

Use the trained Real-ESRNet model as generator initialization and train Real-ESRGAN with a combination of losses.
  • Combines L1 loss, perceptual loss, and GAN loss
  • Improves perceptual quality
  • Produces the final Real-ESRGAN model

​
Why Two Stages?

This two-stage approach offers several advantages:
  • Stability: Starting with L1 loss provides a stable baseline before introducing adversarial training
  • Quality: The combination of losses in stage 2 improves perceptual quality while maintaining fidelity
  • Convergence: Pre-training with L1 loss helps the GAN training converge more reliably

​
Training Requirements

​
Hardware

  • Multiple GPUs recommended (examples use 4 GPUs)
  • Single GPU training is supported but slower
  • Adequate disk space for datasets

​
Datasets

Real-ESRGAN is trained on:
  • DF2K: Combination of DIV2K and Flickr2K datasets
  • OST: OpenImages subset for training
Only high-resolution (HR) images are required. Low-quality images are generated on-the-fly during training using Real-ESRGAN’s degradation process.
The degradation process simulates real-world image degradation, including blur, noise, compression artifacts, and downsampling.

​
Training Modes

​
Debug Mode

Test your configuration before full training: 
python realesrgan/train.py -opt options/train_realesrnet_x4plus.yml --debug

​
Full Training

Run the complete training process: 
python realesrgan/train.py -opt options/train_realesrnet_x4plus.yml --auto_resume
The --auto_resume flag automatically resumes training from the last checkpoint if interrupted.

​
Next Steps

Dataset Preparation

Learn how to prepare and process training datasets

Train Real-ESRNet

Start with stage 1 training

Train Real-ESRGAN

Complete stage 2 for the final model

Fine-tuning

Adapt the model to your custom dataset

Build docs developers (and LLMs) love

Get started for freeTalk to us