--- library_name: pytorch license: other tags: - bu_auto - android pipeline_tag: other --- ![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/statetransformer/web-assets/model_demo.png) # StateTransformer: Optimized for Qualcomm Devices StateTransformer is a transformer-based model designed for trajectory prediction in self-driving scenarios. It integrates rasterized map data, agent context, and temporal dynamics to generate accurate future trajectories. This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/v0.59.0/src/qai_hub_models/models/statetransformer) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary). Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.com) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) to run these models on a hosted Qualcomm® device. ## Getting Started There are two ways to deploy this model on your device: ### Option 1: Download Pre-Exported Models Below are pre-exported model assets ready for deployment. | Runtime | Precision | Chipset | SDK Versions | Download | |---|---|---|---|---| | ONNX | float | Universal | QAIRT 2.45, ONNX Runtime 1.27.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/statetransformer/releases/v0.59.0/statetransformer-onnx-float.zip) | TFLITE | float | Universal | | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/statetransformer/releases/v0.59.0/statetransformer-tflite-float.zip) For more device-specific assets and performance metrics, visit **[StateTransformer on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/statetransformer)**. ### Option 2: Export with Custom Configurations Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/v0.59.0/src/qai_hub_models/models/statetransformer) Python library to compile and export the model with your own: - Custom weights (e.g., fine-tuned checkpoints) - Custom input shapes - Target device and runtime configurations This option is ideal if you need to customize the model beyond the default configuration provided here. See our repository for [StateTransformer on GitHub](https://github.com/qualcomm/ai-hub-models/blob/v0.59.0/src/qai_hub_models/models/statetransformer) for usage instructions. ## Model Details **Model Type:** Model_use_case.driver_assistance **Model Stats:** - Model checkpoint: pretrained-mixtral-small - Input resolution: 1x224x224x58, 1x224x224x58, 1x4x7 - Number of parameters: 90.7M - Model size (float): 348 MB ## Performance Summary | Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit |---|---|---|---|---|---|--- | StateTransformer | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 935.497 ms | 75 - 3989 MB | NPU | StateTransformer | ONNX | float | Qualcomm® Dragonwing™ Q-8750 | 771.09 ms | 89 - 3000 MB | NPU | StateTransformer | ONNX | float | Snapdragon® 8 Elite Mobile | 771.09 ms | 89 - 3000 MB | NPU | StateTransformer | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 493.767 ms | 308 - 326 MB | CPU | StateTransformer | TFLITE | float | Snapdragon® 8 Gen 1 Mobile | 713.923 ms | 265 - 284 MB | CPU | StateTransformer | TFLITE | float | Qualcomm® Dragonwing™ QCS8275 | 931.808 ms | 443 - 457 MB | CPU | StateTransformer | TFLITE | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 591.882 ms | 393 - 459 MB | CPU | StateTransformer | TFLITE | float | Qualcomm® SA8775P | 735.722 ms | 448 - 460 MB | CPU | StateTransformer | TFLITE | float | Qualcomm® SA8650P | 735.722 ms | 448 - 460 MB | CPU | StateTransformer | TFLITE | float | Qualcomm® SA8255P | 735.722 ms | 448 - 460 MB | CPU | StateTransformer | TFLITE | float | Qualcomm® QCS8450 | 713.923 ms | 265 - 284 MB | CPU | StateTransformer | TFLITE | float | Qualcomm® Dragonwing™ IQ-9075 | 758.484 ms | 238 - 659 MB | CPU | StateTransformer | TFLITE | float | Qualcomm® Dragonwing™ Q-8750 | 377.443 ms | 411 - 422 MB | CPU | StateTransformer | TFLITE | float | Qualcomm® SA7255P | 931.808 ms | 443 - 457 MB | CPU | StateTransformer | TFLITE | float | Qualcomm® SA8295P | 596.745 ms | 387 - 396 MB | CPU | StateTransformer | TFLITE | float | Snapdragon® 8 Elite Mobile | 377.443 ms | 411 - 422 MB | CPU | StateTransformer | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 252.872 ms | 192 - 204 MB | CPU ## License * The license for the original implementation of StateTransformer can be found [here](https://github.com/Tsinghua-MARS-Lab/StateTransformer/blob/main/setup.py). ## Community * Join [our AI Hub Slack community](https://aihub.qualcomm.com/community/slack) to collaborate, post questions and learn more about on-device AI. * For questions or feedback please [reach out to us](mailto:ai-hub-support@qti.qualcomm.com).