LightlyTrain Documentation¶

Train Better Models, Faster

LightlyTrain is the leading framework for transforming your data into state-of-the-art computer vision models. It covers the entire model development lifecycle from pretraining DINOv2/v3 vision foundation models on your unlabeled data to fine-tuning transformer and YOLO models on detection and segmentation tasks for edge deployment.

News¶

[0.12.0] - 2025-11-06: 💡 New DINOv3 Object Detection: Run inference or fine-tune DINOv3 models for object detection! 💡
[0.11.0] - 2025-08-15: 🚀 New DINOv3 Support: Pretrain your own model with distillation from DINOv3 weights. Or fine-tune our SOTA EoMT semantic segmentation model with a DINOv3 backbone! 🚀
[0.10.0] - 2025-08-04: 🔥 Train state-of-the-art semantic segmentation models with our new DINOv2 semantic segmentation fine-tuning method! 🔥
[0.9.0] - 2025-07-21: DINOv2 pretraining is now officially available!

Workflows¶

Object Detection

Train LTDETR detection models with DINOv2 or DINOv3 backbones.

object_detection.html

Instance Segmentation

Train EoMT segmentation models with DINOv3 backbones.

instance_segmentation.html

Semantic Segmentation

Train EoMT segmentation models with DINOv2 or DINOv3 backbones.

semantic_segmentation.html

Distillation

Distill knowledge from DINOv2 or DINOv3 into any model architecture.

methods/distillation.html

Pretraining

Pretrain DINOv2 foundation models on your domain data.

methods/dinov2.html

Autolabeling

Generate high-quality pseudo labels for detection and segmentation tasks.

predict_autolabel.html

How It Works ¶

Install LightlyTrain on Python 3.8+ for Windows, Linux or MacOS.

pip install lightly-train

Then train an object detection model with:

import lightly_train

if __name__ == "__main__":
    lightly_train.train_object_detection(
        out="out/my_experiment",
        model="dinov3/convnext-tiny-ltdetr-coco",
        data={
            # ... Data configuration
        }
      )

And run inference like this:

import lightly_train

# Load the model from the best checkpoint
model = lightly_train.load_model("out/my_experiment/exported_models/exported_best.pt")
# Or load one of the models hosted by LightlyTrain
model = lightly_train.load_model("dinov3/convnext-tiny-ltdetr-coco")
results = model.predict("image.jpg")

See the full quick start guide for more details.

Features¶

Python, Command Line, and Docker support
Built for high performance including multi-GPU and multi-node support
Monitor training progress with MLflow, TensorBoard, Weights & Biases, and more
Runs fully on-premises with no API authentication
Export models in their native format for fine-tuning or inference
Export models in ONNX or TensorRT format for edge deployment

Models¶

LightlyTrain supports the following model and workflow combinations.

Fine-tuning¶

Model	Object Detection	Instance Segmentation	Semantic Segmentation
DINOv3	✅ 🔗	✅ 🔗	✅ 🔗
DINOv2	✅ 🔗		✅ 🔗

Distillation & Pretraining¶

Model	Distillation	Pretraining
DINOv3	✅ 🔗
DINOv2	✅ 🔗	✅ 🔗
Torchvision ResNet, ConvNext, ShuffleNetV2	✅ 🔗	✅ 🔗
TIMM models	✅ 🔗	✅ 🔗
Ultralytics YOLOv5–YOLO12	✅ 🔗	✅ 🔗
RT-DETR, RT-DETRv2	✅ 🔗	✅ 🔗
RF-DETR	✅ 🔗	✅ 🔗
YOLOv12	✅ 🔗	✅ 🔗
Custom PyTorch Model	✅ 🔗	✅ 🔗

Usage Events¶

LightlyTrain collects anonymous usage events to help us improve the product. We only track training method, model architecture, and system information (OS, GPU). To opt-out, set the environment variable: export LIGHTLY_TRAIN_EVENTS_DISABLED=1

License¶

LightlyTrain offers flexible licensing options to suit your specific needs:

AGPL-3.0 License: Perfect for open-source projects, academic research, and community contributions. Share your innovations with the world while benefiting from community improvements.
Commercial License: Ideal for businesses and organizations that need proprietary development freedom. Enjoy all the benefits of LightlyTrain while keeping your code and models private.

We’re committed to supporting both open-source and commercial users. Please contact us to discuss the best licensing option for your project!