prunaai/p-image-trainer 🔢🖼️❓📝 → 🖼️

⭐ Official ▶️ 299 runs 📅 Dec 2025 ⚙️ Cog 0.16.9
image-lora-training

Performance

266.9sTypical run time
299Total runs

About

Fast LoRA trainer for p-image, a super fast text-to-image model developed by Pruna AI. Use LoRAs here: https://replicate.com/prunaai/p-image-lora. Find or contribute LoRAs here: https://huggingface.co/collections/PrunaAI/p-image

Example Output

Output

Example output

Performance Metrics

266.90s Prediction Time
266.91s Total Time
All Input Parameters 
{
  "steps": 100,
  "image_data": "https://replicate.delivery/pbxt/OZnSAfjPaTkql5frLsdDnHnCKO1NrKuCw41iW7eZSeIZPMFI/input.zip",
  "learning_rate": 0.0001,
  "training_type": "style"
}
Input Parameters
steps Type: integerDefault: 1000Range: 100 - 5000
Total number of training steps
image_data (required) Type: string
Zip archive with images of a consistent style. Use at least 10 images. The zip can also contain a text file for each image (e.g., photo.txt for photo.jpg) to specify edit instructions for the image pair. You can also pass a URL pointing to the zip file instead of uploading a file directly.
learning_rate Type: numberDefault: 0.0001Range: 0.00001 - 0.01
Learning rate applied to trainable parameters
training_type Default: balanced
Type of training to perform
default_caption Type: string
Default caption to use when caption files are missing. If not provided and captions are missing, training will fail.
Output Schema

Output

Type: string • Format: uri

Example Execution Logs 
Uploading training data...
Upload complete
Starting training (steps=100, lr=0.0001, type=style)...
Training complete
Downloading LoRA weights...
Encrypting LoRA weights...
Zipping LoRA weights...
Done!
Version Details
Version ID
4db88013a63930daab0cd7007a0c9683cd58cfac6768e3ca244de99b30921010
Version Created
February 12, 2026
Run on Replicate →