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 library to export with custom configurations. More details on model performance across various devices, can be found here.
Qualcomm AI Hub Models uses Qualcomm AI Hub Workbench to compile, profile, and evaluate this model. Sign up 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 2.50, ONNX Runtime 1.27.1 | Download |
| TFLITE | float | Universal | Download |
For more device-specific assets and performance metrics, visit StateTransformer on Qualcomm® AI Hub.
Option 2: Export with Custom Configurations
Use the Qualcomm® AI Hub Models 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 for usage instructions.
Model Details
Model Type: Model_use_case.driver_assistance
Model Stats:
- Input resolution: 1x224x224x58, 1x224x224x58, 1x4x7
- Model checkpoint: pretrained-mixtral-small
- Model size (float): 348 MB
- Number of parameters: 90.7M
Performance Summary
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit |
|---|---|---|---|---|---|---|
| StateTransformer | ONNX | float | Snapdragon® 8 Elite Gen 5 For Galaxy Mobile | 192.571 ms | 173 - 4349 MB | NPU |
| StateTransformer | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 210.813 ms | 177 - 4275 MB | NPU |
| StateTransformer | ONNX | float | Snapdragon® X2 Elite | 177.638 ms | 279 - 279 MB | NPU |
| StateTransformer | ONNX | float | Snapdragon® X Elite | 434.924 ms | 279 - 279 MB | NPU |
| StateTransformer | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 253.642 ms | 188 - 4421 MB | NPU |
| StateTransformer | ONNX | float | Qualcomm® Dragonwing™ IQ-8275 | 323.989 ms | 100 - 142 MB | NPU |
| StateTransformer | ONNX | float | Qualcomm® Dragonwing™ IQ-9075 | 279.812 ms | 105 - 146 MB | NPU |
| StateTransformer | ONNX | float | Qualcomm® Dragonwing™ IQ-X7181 | 434.924 ms | 279 - 279 MB | NPU |
| StateTransformer | ONNX | float | Qualcomm® Dragonwing™ Q-8750 | 210.813 ms | 177 - 4275 MB | NPU |
| StateTransformer | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 661.397 ms | 292 - 3228 MB | CPU |
| StateTransformer | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 898.831 ms | 292 - 4365 MB | CPU |
| StateTransformer | TFLITE | float | Qualcomm® Dragonwing™ IQ-8275 | 1520.712 ms | 294 - 813 MB | CPU |
| StateTransformer | TFLITE | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1114.129 ms | 138 - 401 MB | CPU |
| StateTransformer | TFLITE | float | Qualcomm® SA8775P | 1270.002 ms | 294 - 4058 MB | CPU |
| StateTransformer | TFLITE | float | Qualcomm® SA8650P | 1270.002 ms | 294 - 4058 MB | CPU |
| StateTransformer | TFLITE | float | Qualcomm® SA8255P | 1270.002 ms | 294 - 4058 MB | CPU |
| StateTransformer | TFLITE | float | Qualcomm® QCS8450 | 808.833 ms | 450 - 470 MB | CPU |
| StateTransformer | TFLITE | float | Qualcomm® Dragonwing™ IQ-9075 | 1311.722 ms | 294 - 813 MB | CPU |
| StateTransformer | TFLITE | float | Qualcomm® Dragonwing™ Q-8750 | 661.397 ms | 292 - 3228 MB | CPU |
| StateTransformer | TFLITE | float | Qualcomm® SA8295P | 954.001 ms | 290 - 3938 MB | CPU |
License
- The license for the original implementation of StateTransformer can be found here.
Community
- Join our AI Hub Slack community to collaborate, post questions and learn more about on-device AI.
- For questions or feedback please reach out to us.