TIMM

This page describes how to use TIMM models with LightlyTrain.

Important

TIMM must be installed with pip install "lightly-train[timm]".

Pretrain a TIMM Model

Pretraining TIMM models with LightlyTrain is straightforward. Below we will provide the minimum scripts for pretraining using timm/resnet18 as an example:

Python

import lightly_train

if __name__ == "__main__":
    lightly_train.train(
        out="out/my_experiment",                # Output directory.
        data="my_data_dir",                     # Directory with images.
        model="timm/resnet18",                  # Pass the timm model.
    )

Command Line

lightly-train train out="out/my_experiment" data="my_data_dir" model="timm/resnet18"

You can reload the exported model by timm directly:

import timm

model = timm.create_model(
  model_name="resnet18",
  checkpoint_path="out/my_experiment/exported_models/exported_last.pt",
)

Supported Models

All timm models are supported, see timm docs for a full list.

Examples:

• timm/resnet50

• timm/convnext_base

• timm/vit_base_patch16_224