ln2697/transfuser-carla-123d

TransFuser CARLA 123D

End-to-end driving checkpoints trained with kesai-labs/lead on the LEAD 123D dataset for closed-loop driving on CARLA Leaderboard 2.0 routes.

Website · Blog · Paper · Code

Contents

Folders are named <model>_v<lead release>. Each folder's README.md has the model's training details and results, and each seed<N>/ inside holds a config.yaml and the weights, which is all the code needs to rebuild the model.

Usage

Follow the setup in kesai-labs/lead (1.5.1 or newer), then point the evaluation at a seed directory:

user@host:~/lead$ hf download ln2697/transfuser-carla-123d --local-dir checkpoints
user@host:~/lead$ scripts/cli/start_carla
user@host:~/lead$ python -m lead \
    --checkpoint checkpoints/visiononly_resnet34_v1.5.1/seed0 \
    --routes src/lead/routes/benchmark_routes/longest6/00.xml

Citation

If these checkpoints are useful to you, please cite:

@inproceedings{Nguyen2026CVPR,
  author    = {Long Nguyen and Micha Fauth and Bernhard Jaeger and Daniel Dauner and Maximilian Igl and Andreas Geiger and Kashyap Chitta},
  title     = {LEAD: Minimizing Learner-Expert Asymmetry in End-to-End Driving},
  booktitle = {Conference on Computer Vision and Pattern Recognition (CVPR)},
  year      = {2026},
}

@article{Dauner2026ARXIV,
  author  = {Dauner, Daniel and Charraut, Valentin and Berle, Bastian and Li, Tianyu and Nguyen, Long and Wang, Jiabao and Jing, Changhui and Igl, Maximilian and Caesar, Holger and Ivanovic, Boris and Geiger, Andreas and Chitta, Kashyap},
  title   = {123D: Unifying Multi-Modal Autonomous Driving Data at Scale},
  journal = {arXiv preprint arXiv:2605.08084},
  year    = {2026},
}