Swin-Base: Optimized for Qualcomm Devices

SwinBase is a machine learning model that can classify images from the Imagenet dataset. It can also be used as a backbone in building more complex models for specific use cases.

This is based on the implementation of Swin-Base 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
ONNX	w8a16	Universal	QAIRT 2.45, ONNX Runtime 1.27.1	Download
QNN_DLC	float	Universal	QAIRT 2.45	Download
QNN_DLC	w8a16	Universal	QAIRT 2.45	Download
TFLITE	float	Universal	QAIRT 2.45	Download

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

Model Details

Model Type: Model_use_case.image_classification

Model Stats:

• Input resolution: 224x224
• Model checkpoint: Imagenet
• Model size (float): 339 MB
• Model size (w8a16): 90.2 MB
• Number of parameters: 88.8M

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
Swin-Base	ONNX	float	Snapdragon® X2 Elite	15.158 ms	2 - 2 MB	NPU
Swin-Base	ONNX	float	Snapdragon® X Elite	20.012 ms	175 - 175 MB	NPU
Swin-Base	ONNX	float	Snapdragon® 8 Gen 3 Mobile	13.247 ms	0 - 548 MB	NPU
Swin-Base	ONNX	float	Snapdragon® 8 Gen 1 Mobile	29.207 ms	1 - 1065 MB	NPU
Swin-Base	ONNX	float	Qualcomm® Dragonwing™ IQ-8275	21.316 ms	0 - 4 MB	NPU
Swin-Base	ONNX	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	19.223 ms	0 - 194 MB	NPU
Swin-Base	ONNX	float	Qualcomm® QCS8450	29.207 ms	1 - 1065 MB	NPU
Swin-Base	ONNX	float	Qualcomm® Dragonwing™ IQ-9075	22.522 ms	1 - 4 MB	NPU
Swin-Base	ONNX	float	Qualcomm® Dragonwing™ IQ-X7181	20.012 ms	175 - 175 MB	NPU
Swin-Base	ONNX	float	Qualcomm® Dragonwing™ Q-8750	9.865 ms	0 - 400 MB	NPU
Swin-Base	ONNX	float	Snapdragon® 8 Elite Mobile	9.865 ms	0 - 400 MB	NPU
Swin-Base	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	7.977 ms	0 - 420 MB	NPU
Swin-Base	ONNX	w8a16	Snapdragon® X2 Elite	6.983 ms	1 - 1 MB	NPU
Swin-Base	ONNX	w8a16	Snapdragon® X Elite	16.232 ms	92 - 92 MB	NPU
Swin-Base	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	10.52 ms	0 - 535 MB	NPU
Swin-Base	ONNX	w8a16	Snapdragon® 8 Gen 1 Mobile	20.712 ms	1 - 539 MB	NPU
Swin-Base	ONNX	w8a16	Qualcomm® Dragonwing™ QCS6490	58.229 ms	0 - 3 MB	NPU
Swin-Base	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-8275	14.217 ms	0 - 4 MB	NPU
Swin-Base	ONNX	w8a16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	15.45 ms	0 - 99 MB	NPU
Swin-Base	ONNX	w8a16	Qualcomm® QCS8450	20.712 ms	1 - 539 MB	NPU
Swin-Base	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-9075	16.302 ms	0 - 3 MB	NPU
Swin-Base	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-X7181	16.232 ms	92 - 92 MB	NPU
Swin-Base	ONNX	w8a16	Qualcomm® Dragonwing™ Q-6690	98.428 ms	0 - 679 MB	NPU
Swin-Base	ONNX	w8a16	Qualcomm® Dragonwing™ Q-7790	18.059 ms	0 - 625 MB	NPU
Swin-Base	ONNX	w8a16	Qualcomm® Dragonwing™ Q-8750	8.213 ms	0 - 417 MB	NPU
Swin-Base	ONNX	w8a16	Snapdragon® 8 Elite Mobile	8.213 ms	0 - 417 MB	NPU
Swin-Base	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	6.538 ms	0 - 429 MB	NPU
Swin-Base	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	18.059 ms	0 - 625 MB	NPU
Swin-Base	QNN_DLC	float	Snapdragon® X2 Elite	8.451 ms	1 - 1 MB	NPU
Swin-Base	QNN_DLC	float	Snapdragon® X Elite	19.327 ms	1 - 1 MB	NPU
Swin-Base	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	12.488 ms	0 - 516 MB	NPU
Swin-Base	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	29.023 ms	1 - 502 MB	NPU
Swin-Base	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-8275	20.55 ms	1 - 4 MB	NPU
Swin-Base	QNN_DLC	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	18.375 ms	1 - 3 MB	NPU
Swin-Base	QNN_DLC	float	Qualcomm® SA8775P	21.367 ms	1 - 362 MB	NPU
Swin-Base	QNN_DLC	float	Qualcomm® SA8650P	21.367 ms	1 - 362 MB	NPU
Swin-Base	QNN_DLC	float	Qualcomm® SA8255P	21.367 ms	1 - 362 MB	NPU
Swin-Base	QNN_DLC	float	Qualcomm® QCS8450	29.023 ms	1 - 502 MB	NPU
Swin-Base	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-9075	21.542 ms	1 - 3 MB	NPU
Swin-Base	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-X7181	19.327 ms	1 - 1 MB	NPU
Swin-Base	QNN_DLC	float	Qualcomm® Dragonwing™ Q-8750	9.337 ms	1 - 363 MB	NPU
Swin-Base	QNN_DLC	float	Qualcomm® SA7255P	53.601 ms	1 - 363 MB	NPU
Swin-Base	QNN_DLC	float	Qualcomm® SA8295P	27.618 ms	1 - 353 MB	NPU
Swin-Base	QNN_DLC	float	Snapdragon® 8 Elite Mobile	9.337 ms	1 - 363 MB	NPU
Swin-Base	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	7.538 ms	0 - 383 MB	NPU
Swin-Base	QNN_DLC	w8a16	Snapdragon® X2 Elite	8.522 ms	0 - 0 MB	NPU
Swin-Base	QNN_DLC	w8a16	Snapdragon® X Elite	20.587 ms	0 - 0 MB	NPU
Swin-Base	QNN_DLC	w8a16	Snapdragon® 8 Gen 3 Mobile	12.819 ms	0 - 510 MB	NPU
Swin-Base	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-8275	17.252 ms	0 - 3 MB	NPU
Swin-Base	QNN_DLC	w8a16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	19.196 ms	0 - 3 MB	NPU
Swin-Base	QNN_DLC	w8a16	Qualcomm® SA8775P	19.574 ms	0 - 411 MB	NPU
Swin-Base	QNN_DLC	w8a16	Qualcomm® SA8650P	19.574 ms	0 - 411 MB	NPU
Swin-Base	QNN_DLC	w8a16	Qualcomm® SA8255P	19.574 ms	0 - 411 MB	NPU
Swin-Base	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-9075	19.933 ms	0 - 2 MB	NPU
Swin-Base	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-X7181	20.587 ms	0 - 0 MB	NPU
Swin-Base	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-6690	123.964 ms	0 - 929 MB	NPU
Swin-Base	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-7790	21.699 ms	0 - 604 MB	NPU
Swin-Base	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-8750	9.508 ms	0 - 406 MB	NPU
Swin-Base	QNN_DLC	w8a16	Qualcomm® SA7255P	35.433 ms	0 - 412 MB	NPU
Swin-Base	QNN_DLC	w8a16	Snapdragon® 8 Elite Mobile	9.508 ms	0 - 406 MB	NPU
Swin-Base	QNN_DLC	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	7.588 ms	0 - 421 MB	NPU
Swin-Base	QNN_DLC	w8a16	Snapdragon® 7 Gen 4 Mobile	21.699 ms	0 - 604 MB	NPU
Swin-Base	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	12.828 ms	0 - 1058 MB	NPU
Swin-Base	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	29.133 ms	0 - 506 MB	NPU
Swin-Base	TFLITE	float	Qualcomm® Dragonwing™ IQ-8275	20.726 ms	0 - 179 MB	NPU
Swin-Base	TFLITE	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	19.13 ms	0 - 4 MB	NPU
Swin-Base	TFLITE	float	Qualcomm® SA8775P	21.905 ms	0 - 375 MB	NPU
Swin-Base	TFLITE	float	Qualcomm® SA8650P	21.905 ms	0 - 375 MB	NPU
Swin-Base	TFLITE	float	Qualcomm® SA8255P	21.905 ms	0 - 375 MB	NPU
Swin-Base	TFLITE	float	Qualcomm® QCS8450	29.133 ms	0 - 506 MB	NPU
Swin-Base	TFLITE	float	Qualcomm® Dragonwing™ IQ-9075	21.95 ms	0 - 178 MB	NPU
Swin-Base	TFLITE	float	Qualcomm® Dragonwing™ Q-8750	9.671 ms	0 - 386 MB	NPU
Swin-Base	TFLITE	float	Qualcomm® SA7255P	53.445 ms	0 - 710 MB	NPU
Swin-Base	TFLITE	float	Qualcomm® SA8295P	28.296 ms	0 - 371 MB	NPU
Swin-Base	TFLITE	float	Snapdragon® 8 Elite Mobile	9.671 ms	0 - 386 MB	NPU
Swin-Base	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	7.944 ms	0 - 396 MB	NPU

License

• The license for the original implementation of Swin-Base can be found here.

References

• Swin Transformer: Hierarchical Vision Transformer using Shifted Windows
• 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.