Sequencer2D: Optimized for Qualcomm Devices

Sequencer2D is a vision transformer model that can classify images from the Imagenet dataset.

This is based on the implementation of Sequencer2D 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
TFLITE	w8a8	Universal	QAIRT 2.45	Download

For more device-specific assets and performance metrics, visit Sequencer2D 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 Sequencer2D on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.image_classification

Model Stats:

• Input resolution: 224x224
• Model checkpoint: sequencer2d_s
• Model size (float): 106 MB
• Model size (w8a8): 69.1 MB
• Number of parameters: 27.6M

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
Sequencer2D	ONNX	float	Snapdragon® X2 Elite	10.77 ms	2 - 2 MB	NPU
Sequencer2D	ONNX	float	Snapdragon® X Elite	17.025 ms	60 - 60 MB	NPU
Sequencer2D	ONNX	float	Snapdragon® 8 Gen 3 Mobile	10.845 ms	0 - 957 MB	NPU
Sequencer2D	ONNX	float	Snapdragon® 8 Gen 1 Mobile	22.488 ms	1 - 656 MB	NPU
Sequencer2D	ONNX	float	Qualcomm® Dragonwing™ IQ-8275	16.445 ms	0 - 4 MB	NPU
Sequencer2D	ONNX	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	16.068 ms	1 - 5 MB	NPU
Sequencer2D	ONNX	float	Qualcomm® QCS8450	22.488 ms	1 - 656 MB	NPU
Sequencer2D	ONNX	float	Qualcomm® Dragonwing™ IQ-9075	18.991 ms	1 - 4 MB	NPU
Sequencer2D	ONNX	float	Qualcomm® Dragonwing™ IQ-X7181	17.025 ms	60 - 60 MB	NPU
Sequencer2D	ONNX	float	Qualcomm® Dragonwing™ Q-8750	9.582 ms	1 - 737 MB	NPU
Sequencer2D	ONNX	float	Snapdragon® 8 Elite Mobile	9.582 ms	1 - 737 MB	NPU
Sequencer2D	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	8.06 ms	0 - 889 MB	NPU
Sequencer2D	QNN_DLC	float	Snapdragon® X2 Elite	11.853 ms	1 - 1 MB	NPU
Sequencer2D	QNN_DLC	float	Snapdragon® X Elite	21.576 ms	1 - 1 MB	NPU
Sequencer2D	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	14.197 ms	0 - 2197 MB	NPU
Sequencer2D	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	25.604 ms	0 - 818 MB	NPU
Sequencer2D	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-8275	21.03 ms	1 - 4 MB	NPU
Sequencer2D	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-8275	51.755 ms	1 - 981 MB	NPU
Sequencer2D	QNN_DLC	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	20.711 ms	1 - 3 MB	NPU
Sequencer2D	QNN_DLC	float	Qualcomm® SA8775P	24.047 ms	1 - 900 MB	NPU
Sequencer2D	QNN_DLC	float	Qualcomm® SA8650P	24.047 ms	1 - 900 MB	NPU
Sequencer2D	QNN_DLC	float	Qualcomm® SA8255P	24.047 ms	1 - 900 MB	NPU
Sequencer2D	QNN_DLC	float	Qualcomm® QCS8450	25.604 ms	0 - 818 MB	NPU
Sequencer2D	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-9075	23.887 ms	1 - 3 MB	NPU
Sequencer2D	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-X7181	21.576 ms	1 - 1 MB	NPU
Sequencer2D	QNN_DLC	float	Qualcomm® Dragonwing™ Q-8750	12.144 ms	1 - 1025 MB	NPU
Sequencer2D	QNN_DLC	float	Qualcomm® SA7255P	51.755 ms	1 - 981 MB	NPU
Sequencer2D	QNN_DLC	float	Qualcomm® SA8295P	25.937 ms	1 - 656 MB	NPU
Sequencer2D	QNN_DLC	float	Snapdragon® 8 Elite Mobile	12.144 ms	1 - 1025 MB	NPU
Sequencer2D	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	9.558 ms	0 - 1188 MB	NPU
Sequencer2D	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	11.91 ms	0 - 979 MB	NPU
Sequencer2D	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	21.277 ms	0 - 732 MB	NPU
Sequencer2D	TFLITE	float	Qualcomm® Dragonwing™ IQ-8275	17.148 ms	0 - 76 MB	NPU
Sequencer2D	TFLITE	float	Qualcomm® Dragonwing™ IQ-8275	37.304 ms	0 - 796 MB	NPU
Sequencer2D	TFLITE	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	16.892 ms	0 - 3 MB	NPU
Sequencer2D	TFLITE	float	Qualcomm® SA8775P	19.072 ms	0 - 795 MB	NPU
Sequencer2D	TFLITE	float	Qualcomm® SA8650P	19.072 ms	0 - 795 MB	NPU
Sequencer2D	TFLITE	float	Qualcomm® SA8255P	19.072 ms	0 - 795 MB	NPU
Sequencer2D	TFLITE	float	Qualcomm® QCS8450	21.277 ms	0 - 732 MB	NPU
Sequencer2D	TFLITE	float	Qualcomm® Dragonwing™ IQ-9075	18.512 ms	0 - 75 MB	NPU
Sequencer2D	TFLITE	float	Qualcomm® Dragonwing™ Q-8750	9.377 ms	0 - 795 MB	NPU
Sequencer2D	TFLITE	float	Qualcomm® SA7255P	37.304 ms	0 - 796 MB	NPU
Sequencer2D	TFLITE	float	Qualcomm® SA8295P	20.518 ms	0 - 485 MB	NPU
Sequencer2D	TFLITE	float	Snapdragon® 8 Elite Mobile	9.377 ms	0 - 795 MB	NPU
Sequencer2D	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	7.246 ms	0 - 761 MB	NPU
Sequencer2D	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	13.514 ms	0 - 947 MB	NPU
Sequencer2D	TFLITE	w8a8	Snapdragon® 8 Gen 1 Mobile	21.874 ms	0 - 811 MB	NPU
Sequencer2D	TFLITE	w8a8	Qualcomm® Dragonwing™ IQ-8275	17.839 ms	0 - 64 MB	NPU
Sequencer2D	TFLITE	w8a8	Qualcomm® Dragonwing™ IQ-8275	36.307 ms	0 - 862 MB	NPU
Sequencer2D	TFLITE	w8a8	Qualcomm® Dragonwing™ QCS8550 (Proxy)	19.235 ms	0 - 10 MB	NPU
Sequencer2D	TFLITE	w8a8	Qualcomm® SA8775P	20.39 ms	0 - 862 MB	NPU
Sequencer2D	TFLITE	w8a8	Qualcomm® SA8650P	20.39 ms	0 - 862 MB	NPU
Sequencer2D	TFLITE	w8a8	Qualcomm® SA8255P	20.39 ms	0 - 862 MB	NPU
Sequencer2D	TFLITE	w8a8	Qualcomm® QCS8450	21.874 ms	0 - 811 MB	NPU
Sequencer2D	TFLITE	w8a8	Qualcomm® Dragonwing™ IQ-9075	19.715 ms	0 - 64 MB	NPU
Sequencer2D	TFLITE	w8a8	Qualcomm® Dragonwing™ Q-8750	10.992 ms	0 - 791 MB	NPU
Sequencer2D	TFLITE	w8a8	Qualcomm® SA7255P	36.307 ms	0 - 862 MB	NPU
Sequencer2D	TFLITE	w8a8	Qualcomm® SA8295P	22.924 ms	0 - 616 MB	NPU
Sequencer2D	TFLITE	w8a8	Snapdragon® 8 Elite Mobile	10.992 ms	0 - 791 MB	NPU
Sequencer2D	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	9.068 ms	0 - 823 MB	NPU

License

• The license for the original implementation of Sequencer2D can be found here.

References

• Sequencer: Deep LSTM for Image Classification
• Source Model Implementation

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.