Swin-Tiny: Optimized for Qualcomm Devices

SwinTiny 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-Tiny 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 2.45	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-Tiny 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-Tiny on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.image_classification

Model Stats:

• Model checkpoint: Imagenet
• Input resolution: 224x224
• Number of parameters: 28.8M
• Model size (float): 110 MB
• Model size (w8a16): 29.9 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
Swin-Tiny	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	3.614 ms	1 - 431 MB	NPU
Swin-Tiny	ONNX	float	Snapdragon® 8 Elite Mobile	4.361 ms	0 - 432 MB	NPU
Swin-Tiny	ONNX	float	Snapdragon® X2 Elite	3.826 ms	57 - 57 MB	NPU
Swin-Tiny	ONNX	float	Snapdragon® X Elite	9.151 ms	57 - 57 MB	NPU
Swin-Tiny	ONNX	float	Snapdragon® X Elite	9.151 ms	57 - 57 MB	NPU
Swin-Tiny	ONNX	float	Snapdragon® 8 Gen 3 Mobile	5.747 ms	0 - 303 MB	NPU
Swin-Tiny	ONNX	float	Qualcomm® QCS8550 (Proxy)	8.646 ms	0 - 67 MB	NPU
Swin-Tiny	ONNX	float	Qualcomm® QCS9075	11.243 ms	0 - 4 MB	NPU
Swin-Tiny	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	4.361 ms	0 - 432 MB	NPU
Swin-Tiny	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	3.43 ms	0 - 240 MB	NPU
Swin-Tiny	ONNX	w8a16	Snapdragon® 8 Elite Mobile	4.257 ms	0 - 242 MB	NPU
Swin-Tiny	ONNX	w8a16	Snapdragon® X2 Elite	3.683 ms	31 - 31 MB	NPU
Swin-Tiny	ONNX	w8a16	Snapdragon® X Elite	8.873 ms	30 - 30 MB	NPU
Swin-Tiny	ONNX	w8a16	Snapdragon® X Elite	8.873 ms	30 - 30 MB	NPU
Swin-Tiny	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	5.454 ms	0 - 325 MB	NPU
Swin-Tiny	ONNX	w8a16	Qualcomm® QCS6490	386.681 ms	94 - 109 MB	CPU
Swin-Tiny	ONNX	w8a16	Qualcomm® QCS8550 (Proxy)	8.355 ms	0 - 42 MB	NPU
Swin-Tiny	ONNX	w8a16	Qualcomm® QCS9075	9.877 ms	0 - 3 MB	NPU
Swin-Tiny	ONNX	w8a16	Qualcomm® QCM6690	211.749 ms	102 - 119 MB	CPU
Swin-Tiny	ONNX	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	4.257 ms	0 - 242 MB	NPU
Swin-Tiny	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	199.0 ms	103 - 120 MB	CPU
Swin-Tiny	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	199.0 ms	103 - 120 MB	CPU
Swin-Tiny	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	3.745 ms	1 - 410 MB	NPU
Swin-Tiny	QNN_DLC	float	Snapdragon® 8 Elite Mobile	4.541 ms	0 - 414 MB	NPU
Swin-Tiny	QNN_DLC	float	Snapdragon® X2 Elite	4.248 ms	1 - 1 MB	NPU
Swin-Tiny	QNN_DLC	float	Snapdragon® X Elite	9.873 ms	1 - 1 MB	NPU
Swin-Tiny	QNN_DLC	float	Snapdragon® X Elite	9.873 ms	1 - 1 MB	NPU
Swin-Tiny	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	5.99 ms	0 - 246 MB	NPU
Swin-Tiny	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	21.285 ms	1 - 175 MB	NPU
Swin-Tiny	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	9.201 ms	1 - 188 MB	NPU
Swin-Tiny	QNN_DLC	float	Qualcomm® SA8775P	10.458 ms	1 - 438 MB	NPU
Swin-Tiny	QNN_DLC	float	Qualcomm® SA8775P	10.458 ms	1 - 438 MB	NPU
Swin-Tiny	QNN_DLC	float	Qualcomm® SA8775P	10.458 ms	1 - 438 MB	NPU
Swin-Tiny	QNN_DLC	float	Qualcomm® QCS9075	11.104 ms	3 - 5 MB	NPU
Swin-Tiny	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	14.525 ms	0 - 243 MB	NPU
Swin-Tiny	QNN_DLC	float	Qualcomm® SA7255P	21.285 ms	1 - 175 MB	NPU
Swin-Tiny	QNN_DLC	float	Qualcomm® SA8295P	13.811 ms	1 - 433 MB	NPU
Swin-Tiny	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	4.541 ms	0 - 414 MB	NPU
Swin-Tiny	QNN_DLC	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	3.829 ms	0 - 435 MB	NPU
Swin-Tiny	QNN_DLC	w8a16	Snapdragon® 8 Elite Mobile	4.814 ms	0 - 430 MB	NPU
Swin-Tiny	QNN_DLC	w8a16	Snapdragon® X2 Elite	4.391 ms	0 - 0 MB	NPU
Swin-Tiny	QNN_DLC	w8a16	Snapdragon® X Elite	10.854 ms	0 - 0 MB	NPU
Swin-Tiny	QNN_DLC	w8a16	Snapdragon® X Elite	10.854 ms	0 - 0 MB	NPU
Swin-Tiny	QNN_DLC	w8a16	Snapdragon® 8 Gen 3 Mobile	6.516 ms	0 - 509 MB	NPU
Swin-Tiny	QNN_DLC	w8a16	Qualcomm® QCS8275 (Proxy)	17.653 ms	0 - 434 MB	NPU
Swin-Tiny	QNN_DLC	w8a16	Qualcomm® QCS8550 (Proxy)	10.174 ms	0 - 9 MB	NPU
Swin-Tiny	QNN_DLC	w8a16	Qualcomm® SA8775P	10.617 ms	0 - 440 MB	NPU
Swin-Tiny	QNN_DLC	w8a16	Qualcomm® SA8775P	10.617 ms	0 - 440 MB	NPU
Swin-Tiny	QNN_DLC	w8a16	Qualcomm® SA8775P	10.617 ms	0 - 440 MB	NPU
Swin-Tiny	QNN_DLC	w8a16	Qualcomm® QCS9075	11.772 ms	0 - 2 MB	NPU
Swin-Tiny	QNN_DLC	w8a16	Qualcomm® QCM6690	44.47 ms	0 - 429 MB	NPU
Swin-Tiny	QNN_DLC	w8a16	Qualcomm® SA7255P	17.653 ms	0 - 434 MB	NPU
Swin-Tiny	QNN_DLC	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	4.814 ms	0 - 430 MB	NPU
Swin-Tiny	QNN_DLC	w8a16	Snapdragon® 7 Gen 4 Mobile	10.555 ms	0 - 288 MB	NPU
Swin-Tiny	QNN_DLC	w8a16	Snapdragon® 7 Gen 4 Mobile	10.555 ms	0 - 288 MB	NPU
Swin-Tiny	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	3.805 ms	0 - 200 MB	NPU
Swin-Tiny	TFLITE	float	Snapdragon® 8 Elite Mobile	4.622 ms	0 - 200 MB	NPU
Swin-Tiny	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	6.043 ms	0 - 253 MB	NPU
Swin-Tiny	TFLITE	float	Qualcomm® QCS8275 (Proxy)	21.205 ms	0 - 200 MB	NPU
Swin-Tiny	TFLITE	float	Qualcomm® QCS8550 (Proxy)	9.285 ms	0 - 3 MB	NPU
Swin-Tiny	TFLITE	float	Qualcomm® SA8775P	10.405 ms	0 - 201 MB	NPU
Swin-Tiny	TFLITE	float	Qualcomm® SA8775P	10.405 ms	0 - 201 MB	NPU
Swin-Tiny	TFLITE	float	Qualcomm® SA8775P	10.405 ms	0 - 201 MB	NPU
Swin-Tiny	TFLITE	float	Qualcomm® QCS9075	11.419 ms	0 - 60 MB	NPU
Swin-Tiny	TFLITE	float	Qualcomm® QCS8450 (Proxy)	14.359 ms	0 - 258 MB	NPU
Swin-Tiny	TFLITE	float	Qualcomm® SA7255P	21.205 ms	0 - 200 MB	NPU
Swin-Tiny	TFLITE	float	Qualcomm® SA8295P	13.835 ms	0 - 199 MB	NPU
Swin-Tiny	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	4.622 ms	0 - 200 MB	NPU

License

• The license for the original implementation of Swin-Tiny 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.