You can deploy this model and start earning money today!

qualcomm/
ACT

Views

No views yet

ACT: Optimized for Qualcomm Devices

ACT (Action Chunking with Transformers) is a robotic policy model that is trained to predict the next chunk of actions that the robotic hand is expected to perform.

This is based on the implementation of ACT found here. 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 2.45, ONNX Runtime 1.27.1	Download
QNN_DLC	float	Universal	QAIRT 2.45	Download
TFLITE	float	Universal	QAIRT 2.45	Download

For more device-specific assets and performance metrics, visit ACT 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 ACT on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.robotics

Model Stats:

Input resolution: 480x640
Model checkpoint: act
Model size (float): 255M
Number of parameters: 83.92M

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
ACT	ONNX	float	Snapdragon® X2 Elite	6.754 ms	4 - 4 MB	NPU
ACT	ONNX	float	Snapdragon® X Elite	12.829 ms	64 - 64 MB	NPU
ACT	ONNX	float	Snapdragon® 8 Gen 3 Mobile	8.575 ms	3 - 510 MB	NPU
ACT	ONNX	float	Snapdragon® 8 Gen 1 Mobile	19.484 ms	1 - 316 MB	NPU
ACT	ONNX	float	Qualcomm® Dragonwing™ IQ-8275	17.437 ms	3 - 10 MB	NPU
ACT	ONNX	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	12.221 ms	4 - 7 MB	NPU
ACT	ONNX	float	Qualcomm® QCS8450	19.484 ms	1 - 316 MB	NPU
ACT	ONNX	float	Qualcomm® Dragonwing™ IQ-9075	17.784 ms	4 - 10 MB	NPU
ACT	ONNX	float	Qualcomm® Dragonwing™ IQ-X7181	12.829 ms	64 - 64 MB	NPU
ACT	ONNX	float	Qualcomm® Dragonwing™ Q-8750	6.844 ms	1 - 414 MB	NPU
ACT	ONNX	float	Snapdragon® 8 Elite Mobile	6.844 ms	1 - 414 MB	NPU
ACT	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	5.973 ms	0 - 444 MB	NPU
ACT	QNN_DLC	float	Snapdragon® X2 Elite	5.295 ms	4 - 4 MB	NPU
ACT	QNN_DLC	float	Snapdragon® X Elite	11.001 ms	4 - 4 MB	NPU
ACT	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	6.945 ms	4 - 338 MB	NPU
ACT	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	20.764 ms	3 - 261 MB	NPU
ACT	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-8275	14.505 ms	4 - 9 MB	NPU
ACT	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-8275	44.849 ms	1 - 293 MB	NPU
ACT	QNN_DLC	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	10.272 ms	4 - 6 MB	NPU
ACT	QNN_DLC	float	Qualcomm® SA8775P	14.479 ms	1 - 270 MB	NPU
ACT	QNN_DLC	float	Qualcomm® SA8650P	14.479 ms	1 - 270 MB	NPU
ACT	QNN_DLC	float	Qualcomm® SA8255P	14.479 ms	1 - 270 MB	NPU
ACT	QNN_DLC	float	Qualcomm® QCS8450	20.764 ms	3 - 261 MB	NPU
ACT	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-9075	14.25 ms	4 - 9 MB	NPU
ACT	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-X7181	11.001 ms	4 - 4 MB	NPU
ACT	QNN_DLC	float	Qualcomm® Dragonwing™ Q-8750	5.442 ms	0 - 293 MB	NPU
ACT	QNN_DLC	float	Qualcomm® SA7255P	44.849 ms	1 - 293 MB	NPU
ACT	QNN_DLC	float	Qualcomm® SA8295P	16.537 ms	0 - 210 MB	NPU
ACT	QNN_DLC	float	Snapdragon® 8 Elite Mobile	5.442 ms	0 - 293 MB	NPU
ACT	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	4.382 ms	4 - 315 MB	NPU
ACT	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	6.115 ms	0 - 413 MB	NPU
ACT	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	16.427 ms	0 - 272 MB	NPU
ACT	TFLITE	float	Qualcomm® Dragonwing™ IQ-8275	14.235 ms	0 - 71 MB	NPU
ACT	TFLITE	float	Qualcomm® Dragonwing™ IQ-8275	44.749 ms	0 - 330 MB	NPU
ACT	TFLITE	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	8.442 ms	0 - 3 MB	NPU
ACT	TFLITE	float	Qualcomm® SA8775P	13.841 ms	0 - 281 MB	NPU
ACT	TFLITE	float	Qualcomm® SA8650P	13.841 ms	0 - 281 MB	NPU
ACT	TFLITE	float	Qualcomm® SA8255P	13.841 ms	0 - 281 MB	NPU
ACT	TFLITE	float	Qualcomm® QCS8450	16.427 ms	0 - 272 MB	NPU
ACT	TFLITE	float	Qualcomm® Dragonwing™ IQ-9075	13.852 ms	0 - 71 MB	NPU
ACT	TFLITE	float	Qualcomm® Dragonwing™ Q-8750	4.74 ms	0 - 283 MB	NPU
ACT	TFLITE	float	Qualcomm® SA7255P	44.749 ms	0 - 330 MB	NPU
ACT	TFLITE	float	Qualcomm® SA8295P	15.693 ms	0 - 208 MB	NPU
ACT	TFLITE	float	Snapdragon® 8 Elite Mobile	4.74 ms	0 - 283 MB	NPU
ACT	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	3.792 ms	0 - 324 MB	NPU

License

The license for the original implementation of ACT can be found here.

References

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.

qualcomm/ACT