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 1.24.3	Download
ONNX	w8a16	Universal	QAIRT 2.42, ONNX Runtime 1.24.3	Download
QNN_DLC	float	Universal	QAIRT 2.43	Download
TFLITE	float	Universal	QAIRT 2.43, TFLite 2.19.1	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® 8 Elite Gen 5 Mobile	7.9 ms	0 - 1367 MB	NPU
Sequencer2D	ONNX	float	Snapdragon® X2 Elite	9.785 ms	61 - 61 MB	NPU
Sequencer2D	ONNX	float	Snapdragon® X Elite	16.994 ms	59 - 59 MB	NPU
Sequencer2D	ONNX	float	Snapdragon® 8 Gen 3 Mobile	11.297 ms	0 - 2277 MB	NPU
Sequencer2D	ONNX	float	Qualcomm® QCS8550 (Proxy)	15.971 ms	0 - 85 MB	NPU
Sequencer2D	ONNX	float	Qualcomm® QCS9075	19.511 ms	0 - 4 MB	NPU
Sequencer2D	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	9.859 ms	1 - 1045 MB	NPU
Sequencer2D	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	8.082 ms	0 - 729 MB	NPU
Sequencer2D	ONNX	w8a16	Snapdragon® X2 Elite	9.941 ms	45 - 45 MB	NPU
Sequencer2D	ONNX	w8a16	Snapdragon® X Elite	18.122 ms	43 - 43 MB	NPU
Sequencer2D	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	11.455 ms	0 - 881 MB	NPU
Sequencer2D	ONNX	w8a16	Qualcomm® QCS6490	622.275 ms	51 - 56 MB	CPU
Sequencer2D	ONNX	w8a16	Qualcomm® QCS8550 (Proxy)	17.426 ms	0 - 72 MB	NPU
Sequencer2D	ONNX	w8a16	Qualcomm® QCS9075	20.243 ms	0 - 3 MB	NPU
Sequencer2D	ONNX	w8a16	Qualcomm® QCM6690	307.245 ms	49 - 66 MB	CPU
Sequencer2D	ONNX	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	10.202 ms	0 - 704 MB	NPU
Sequencer2D	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	300.329 ms	49 - 63 MB	CPU
Sequencer2D	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	9.531 ms	0 - 876 MB	NPU
Sequencer2D	QNN_DLC	float	Snapdragon® X2 Elite	12.526 ms	1 - 1 MB	NPU
Sequencer2D	QNN_DLC	float	Snapdragon® X Elite	21.88 ms	1 - 1 MB	NPU
Sequencer2D	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	14.974 ms	0 - 2258 MB	NPU
Sequencer2D	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	53.777 ms	1 - 976 MB	NPU
Sequencer2D	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	20.811 ms	1 - 3 MB	NPU
Sequencer2D	QNN_DLC	float	Qualcomm® QCS9075	23.092 ms	1 - 3 MB	NPU
Sequencer2D	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	26.443 ms	0 - 867 MB	NPU
Sequencer2D	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	12.671 ms	0 - 1050 MB	NPU
Sequencer2D	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	7.129 ms	0 - 768 MB	NPU
Sequencer2D	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	12.035 ms	0 - 1030 MB	NPU
Sequencer2D	TFLITE	float	Qualcomm® QCS8275 (Proxy)	37.877 ms	0 - 794 MB	NPU
Sequencer2D	TFLITE	float	Qualcomm® QCS8550 (Proxy)	17.075 ms	0 - 12 MB	NPU
Sequencer2D	TFLITE	float	Qualcomm® QCS9075	20.27 ms	0 - 75 MB	NPU
Sequencer2D	TFLITE	float	Qualcomm® QCS8450 (Proxy)	21.422 ms	0 - 720 MB	NPU
Sequencer2D	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	9.293 ms	0 - 790 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.