Haochen Tian, Tianyu Li, Haochen Liu, Jiazhi Yang, Yihang Qiu, Guang Li, Junli Wang, Yinfeng Gao, Zhang Zhang, Liang Wang, Hangjun Ye, Tieniu Tan, Long Chen, Hongyang Li

• 📧 Primary Contact: Haochen Tian (tianhaochen2023@ia.ac.cn)
• 📜 Materials: 🌐 𝕏 | 📰 Media | 🗂️ Slides | 🪧 Poster | 🎬 CVPR Oral Pre / Talk (in Chinese)
• 🖊️ Joint effort by CASIA, OpenDriveLab at HKU, and Xiaomi EV.

• 🏗️ A scalable simulation pipepline that synthesizes diverse and high-fidelity reactive driving scenarios with pseudo-expert demonstrations.
• 🚀 An effective sim-real co-training strategy that improves robustness and generalization synergistically across various end-to-end planners.
• 🔬 A comprehensive recipe that reveals crucial insights into the underlying scaling properties of sim-real learning systems for end-to-end autonomy.

• [2026/8/24] We released the model eval scripts in AlpaSim E2E Challenge.
• [2026/4/9] 🎉🎉🎉 Selected as CVPR 2026 Oral Presentation!
• [2026/2/21] 🎉 Accepted to CVPR 2026.
• [2026/1/16] We released the data and models on 👾 ModelScope to better serve users in China.
• [2026/1/6] We released the code v1.0.
• [2025/12/31] We released the data and models v1.0 on 🤗 Hugging Face. Happy New Year ! 🎄
• [2025/12/1] We released our paper on arXiv.

• More Visualization Results.
• Future Sensors Data.
• Sim-Real Co-training Code release (Jan. 2026).
• Simulation Data release (Dec. 2025).
• Checkpoints release (Dec. 2025).

• 🏛️ Model Zoo
• 🎯 Getting Started
• 📦 Data Preparation
  • Download Dataset
  • Set Up Configuration
• ⚙️ Sim-Real Co-Training
  • Co-Training with Pseudo-Expert
  • Co-Training with Rewards Only
• 🔍 Inference
  • NAVSIM v2 navhard
  • NAVSIM v2 navtest
• 🤝 Awesome Community Projects
• ⭐ License and Citation

git clone https://github.com/OpenDriveLab/SimScale.git
cd SimScale

conda env create --name simscale -f environment.yml
conda activate simscale
pip install -e .

Our released simulation data is based on nuPlan and NAVSIM. We recommend first preparing the real-world data by following the instructions in Download NAVSIM. If you plan to use GTRS, please directly refer Download NAVSIM.

We provide 🤗 Script (Hugging Face) and 👾 Script (ModelScope) (users in China) for downloading the simulation data .

Our simulation data format follows that of OpenScene, with each clip/log has a fixed temporal horizon of 6 seconds at 2 Hz (2 s history + 4 s future), which are stored separately in sensor_blobs_hist and sensor_blobs_fut, respectively. For policy training, sensor_blobs_hist alone is sufficient.

5c9694f15f9c5537	367cfa28901257ee	d37c49db3dcd59fa
Sim. 1 Sim. 2 Sim. 3	Sim. 1 Sim. 2 Sim. 3	Sim. 1 Sim. 2 Sim. 3

We provide a Script for moving the download simulation data to create the following structure.

navsim_workspace/
├── simscale/
├── exp/
└── dataset/
    ├── maps/
    ├── navsim_logs/
    │   ├── test/
    │   ├── trainval/
    │   ├── synthetic_reaction_pdm_v1.0-*/
    │   │   ├── [log]-00*.pkl
    │   │   └── ...
    │   └── synthetic_reaction_recovery_v1.0-*/
    ├── sensor_blobs/
    │   ├── test/
    │   ├── trainval/
    │   ├── synthetic_reaction_pdm_v1.0-*/
    │   │   └── [token]-00*/
    │   │       ├── CAM_B0/
    │   │       └── ...
    │   └── synthetic_reaction_recovery_v1.0-*/
    └── navhard_two_stage/

1. Refer the Script to cache the real-world and simulation data.
2. Download pretrained image backbone weight, ResNet34 or V2_99.

We provide Scripts for sim–real co-training, e.g., run_diffusiondrive_training_syn.sh.

The main configuration options are as follows:

export SYN_IDX=0   # 0, 1, 2, 3, 4
export SYN_GT=pdm  # pdm, recovery

• SYN_IDX specifies which rounds of simulation data are included; e.g., SYN_IDX=2 means that rounds 0, 1, and 2 will be used.
• SYN_GT specifies the type of pseudo-expert used for supervision.

Simulation feature and target caches are loaded from ${NAVSIM_EXP_ROOT}/cache/synthetic_reaction_${SYN_GT}_v1.0-${idx}_cache in dataset.py#136, where idx ranges from 0 through SYN_IDX. Generate these caches with run_dataset_cache.sh, setting TRAIN_TEST_SPLIT to the corresponding synthetic_reaction_* split. If you already have caches under the old split names, rename those cache directories to match this convention before training.

• Regression-based Policy | LTF

We provide a Script to train LTF with 8 GPUs for 100 epochs.

• Diffusion-based Policy | DiffusionDrive

We provide a Script to train DiffusionDrive with 8 GPUs for 100 epochs.

• Scoring-based Policy | GTRS-Dense

We provide a Script to train GTRS_Dense on 4 nodes, each with 8 GPUs, for 50 epochs.

We also provide 🤗 Reward Files (Hugging Face) and 👾 Reward Files (ModelScope) (users in China) for rewards in simulation data. Please download correspending files first and move them to NAVSIM_TRAJPDM_ROOT/sim. The reward files path is hard-coded in gtrs_agent.py#223. Check it before training.

• Scoring-based Policy | GTRS-Dense

It uses the same training Script, to train GTRS_Dense on 4 nodes, each with 8 GPUs, for 50 epochs.

The main configuration option is as follows:

syn_imi=false  # true, false

• syn_imi: When set to false, the imitation learning loss is disabled for simulation data, while it remains enabled for real-world data.

Refer the Script to cache metric first.

We provide Scripts to evaluate three policies on navhard using GPU inference.

We provide Scripts to evaluate three policies on navtest using GPU inference.

We acknowledge all the open-source contributors for the following projects to make this work possible:

• NAVSIM | MTGS | GTRS | DiffusionDrive

The SimScale community continues to grow, with projects building on and extending its real-world simulation and sim-real co-training framework for end-to-end autonomous driving. Here are some awesome community projects:

• AlpaSim | NVIDIA | An open-source platform for developing and testing e2e driving policies.
• World Engine | Huawei & HKU | A post-training framework for e2e driving.
• DrivoR | Valeo.ai | A driving policy using register tokens to compress visual features and score candidate trajs.
• DriveZero | Xiaomi EV | A driving policy trained without human by distilling RL teacher and vision models.
• AutoScale | Li Auto | An efficient data mixture strategy for selecting high-value synthetic data.
• DriveVLA-M0 | Chang'an Tech & CASIA | A failure-aware memory augmentation method for e2e driving.
• TE-E2E | Bosch & THU | A plug-and-play traffic-element awareness method for e2e driving.

All content in this repository is under the Apache-2.0 license. The released data is based on nuPlan and is under the CC-BY-NC-SA 4.0 license.

If any parts of our paper and code help your research, please consider citing us and giving a star to our repository.

@inproceedings{tian2026simscale,
  title={SimScale: Learning to Drive via Real-World Simulation at Scale},
  author={Tian, Haochen and Li, Tianyu and Liu, Haochen and Yang, Jiazhi and Qiu, Yihang and Li, Guang and Wang, Junli and Gao, Yinfeng and Zhang, Zhang and Wang, Liang and Ye, Hangjun and Chen, Long and Li, Hongyang},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  year={2026}