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 2.42, ONNX Runtime 1.24.1	Download
ONNX	w8a16	Universal	QAIRT 2.42, ONNX Runtime 1.24.1	Download
QNN_DLC	float	Universal	QAIRT 2.43	Download
TFLITE	float	Universal	QAIRT 2.43, TFLite 2.17.0	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:

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

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
Sequencer2D	ONNX	float	Snapdragon® X2 Elite	9.798 ms	61 - 61 MB	NPU
Sequencer2D	ONNX	float	Snapdragon® X Elite	16.992 ms	59 - 59 MB	NPU
Sequencer2D	ONNX	float	Snapdragon® 8 Gen 3 Mobile	11.177 ms	1 - 2287 MB	NPU
Sequencer2D	ONNX	float	Qualcomm® QCS8550 (Proxy)	16.359 ms	0 - 85 MB	NPU
Sequencer2D	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	9.796 ms	1 - 1045 MB	NPU
Sequencer2D	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	7.885 ms	0 - 1366 MB	NPU
Sequencer2D	ONNX	w8a16	Snapdragon® X2 Elite	10.38 ms	45 - 45 MB	NPU
Sequencer2D	ONNX	w8a16	Snapdragon® X Elite	18.129 ms	43 - 43 MB	NPU
Sequencer2D	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	11.455 ms	0 - 882 MB	NPU
Sequencer2D	ONNX	w8a16	Qualcomm® QCS6490	642.124 ms	46 - 53 MB	CPU
Sequencer2D	ONNX	w8a16	Qualcomm® QCS8550 (Proxy)	16.97 ms	0 - 61 MB	NPU
Sequencer2D	ONNX	w8a16	Qualcomm® QCM6690	308.884 ms	49 - 66 MB	CPU
Sequencer2D	ONNX	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	10.101 ms	0 - 708 MB	NPU
Sequencer2D	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	299.653 ms	26 - 39 MB	CPU
Sequencer2D	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	8.069 ms	0 - 728 MB	NPU
Sequencer2D	QNN_DLC	float	Snapdragon® X2 Elite	12.547 ms	1 - 1 MB	NPU
Sequencer2D	QNN_DLC	float	Snapdragon® X Elite	21.837 ms	1 - 1 MB	NPU
Sequencer2D	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	14.715 ms	0 - 2256 MB	NPU
Sequencer2D	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	53.715 ms	1 - 976 MB	NPU
Sequencer2D	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	20.713 ms	1 - 4 MB	NPU
Sequencer2D	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	26.646 ms	0 - 846 MB	NPU
Sequencer2D	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	12.59 ms	1 - 1044 MB	NPU
Sequencer2D	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	9.927 ms	0 - 1161 MB	NPU
Sequencer2D	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	12.14 ms	0 - 1030 MB	NPU
Sequencer2D	TFLITE	float	Qualcomm® QCS8275 (Proxy)	37.833 ms	0 - 794 MB	NPU
Sequencer2D	TFLITE	float	Qualcomm® QCS8550 (Proxy)	17.121 ms	0 - 8 MB	NPU
Sequencer2D	TFLITE	float	Qualcomm® QCS8450 (Proxy)	21.715 ms	0 - 722 MB	NPU
Sequencer2D	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	9.392 ms	0 - 791 MB	NPU
Sequencer2D	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	7.136 ms	0 - 768 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.