WideResNet50: Optimized for Qualcomm Devices

WideResNet50 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 WideResNet50 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 1.25.0	Download
ONNX	w8a8	Universal	QAIRT 2.45, ONNX Runtime 1.25.0	Download
QNN_DLC	float	Universal	QAIRT 2.45	Download
QNN_DLC	w8a8	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 WideResNet50 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 WideResNet50 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: 68.9M
• Model size (float): 263 MB
• Model size (w8a8): 66.6 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
WideResNet50	ONNX	float	Snapdragon® X2 Elite	2.867 ms	180 - 180 MB	NPU
WideResNet50	ONNX	float	Snapdragon® X Elite	4.219 ms	148 - 148 MB	NPU
WideResNet50	ONNX	float	Snapdragon® 8 Gen 3 Mobile	3.529 ms	0 - 209 MB	NPU
WideResNet50	ONNX	float	Snapdragon® 8 Gen 1 Mobile	8.812 ms	1 - 189 MB	NPU
WideResNet50	ONNX	float	Qualcomm® QCS8550 (Proxy)	4.542 ms	1 - 4 MB	NPU
WideResNet50	ONNX	float	Qualcomm® QCS8450	8.812 ms	1 - 189 MB	NPU
WideResNet50	ONNX	float	Snapdragon® 8 Elite Mobile	2.912 ms	0 - 172 MB	NPU
WideResNet50	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	2.411 ms	1 - 169 MB	NPU
WideResNet50	ONNX	float	Qualcomm® QCS9075	6.764 ms	1 - 46 MB	NPU
WideResNet50	ONNX	float	Qualcomm® QCS8750	2.912 ms	0 - 172 MB	NPU
WideResNet50	ONNX	float	Qualcomm® QCS7181	4.219 ms	148 - 148 MB	NPU
WideResNet50	ONNX	w8a8	Snapdragon® X2 Elite	0.81 ms	181 - 181 MB	NPU
WideResNet50	ONNX	w8a8	Snapdragon® X Elite	1.683 ms	149 - 149 MB	NPU
WideResNet50	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	1.395 ms	0 - 123 MB	NPU
WideResNet50	ONNX	w8a8	Snapdragon® 8 Gen 1 Mobile	2.587 ms	0 - 125 MB	NPU
WideResNet50	ONNX	w8a8	Qualcomm® QCS6490	7.25 ms	0 - 45 MB	NPU
WideResNet50	ONNX	w8a8	Qualcomm® QCS8550 (Proxy)	1.861 ms	0 - 4 MB	NPU
WideResNet50	ONNX	w8a8	Qualcomm® QCS8450	2.587 ms	0 - 125 MB	NPU
WideResNet50	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	1.159 ms	0 - 55 MB	NPU
WideResNet50	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	2.809 ms	0 - 211 MB	NPU
WideResNet50	ONNX	w8a8	Qualcomm® QCM6690	20.597 ms	0 - 212 MB	NPU
WideResNet50	ONNX	w8a8	Qualcomm® QCS9075	1.93 ms	0 - 45 MB	NPU
WideResNet50	ONNX	w8a8	Snapdragon® 8 Elite Mobile	1.224 ms	0 - 57 MB	NPU
WideResNet50	ONNX	w8a8	Qualcomm® QCS7790	2.809 ms	0 - 211 MB	NPU
WideResNet50	ONNX	w8a8	Qualcomm® QCS8750	1.224 ms	0 - 57 MB	NPU
WideResNet50	ONNX	w8a8	Qualcomm® QCS7181	1.683 ms	149 - 149 MB	NPU
WideResNet50	QNN_DLC	float	Snapdragon® X2 Elite	2.532 ms	1 - 1 MB	NPU
WideResNet50	QNN_DLC	float	Snapdragon® X Elite	4.672 ms	1 - 1 MB	NPU
WideResNet50	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	3.634 ms	0 - 224 MB	NPU
WideResNet50	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	9.221 ms	0 - 197 MB	NPU
WideResNet50	QNN_DLC	float	Qualcomm® QCS8275	24.489 ms	1 - 174 MB	NPU
WideResNet50	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	4.803 ms	1 - 361 MB	NPU
WideResNet50	QNN_DLC	float	Qualcomm® QCS8450	9.221 ms	0 - 197 MB	NPU
WideResNet50	QNN_DLC	float	Snapdragon® 8 Elite Mobile	2.96 ms	1 - 177 MB	NPU
WideResNet50	QNN_DLC	float	Qualcomm® SA8295P	7.837 ms	1 - 152 MB	NPU
WideResNet50	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	2.557 ms	1 - 173 MB	NPU
WideResNet50	QNN_DLC	float	Qualcomm® SA7255P	24.489 ms	1 - 174 MB	NPU
WideResNet50	QNN_DLC	float	Qualcomm® QCS9075	7.065 ms	1 - 3 MB	NPU
WideResNet50	QNN_DLC	float	Qualcomm® QCS8750	2.96 ms	1 - 177 MB	NPU
WideResNet50	QNN_DLC	float	Qualcomm® QCS7181	4.672 ms	1 - 1 MB	NPU
WideResNet50	QNN_DLC	w8a8	Snapdragon® X2 Elite	0.884 ms	0 - 0 MB	NPU
WideResNet50	QNN_DLC	w8a8	Snapdragon® X Elite	1.773 ms	1 - 1 MB	NPU
WideResNet50	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	1.398 ms	0 - 131 MB	NPU
WideResNet50	QNN_DLC	w8a8	Snapdragon® 8 Gen 1 Mobile	2.515 ms	0 - 132 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® QCS6490	7.762 ms	0 - 2 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® QCS8275	4.034 ms	0 - 55 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	1.851 ms	0 - 2 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® QCS8450	2.515 ms	0 - 132 MB	NPU
WideResNet50	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	1.094 ms	0 - 59 MB	NPU
WideResNet50	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	2.858 ms	0 - 211 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® QCM6690	18.132 ms	0 - 217 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® QCS9075	1.906 ms	2 - 4 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® SA7255P	4.034 ms	0 - 55 MB	NPU
WideResNet50	QNN_DLC	w8a8	Snapdragon® 8 Elite Mobile	1.166 ms	0 - 56 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® SA8295P	2.725 ms	0 - 56 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® QCS7790	2.858 ms	0 - 211 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® QCS8750	1.166 ms	0 - 56 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® QCS7181	1.773 ms	1 - 1 MB	NPU
WideResNet50	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	3.553 ms	0 - 214 MB	NPU
WideResNet50	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	8.171 ms	0 - 192 MB	NPU
WideResNet50	TFLITE	float	Qualcomm® QCS8275	23.802 ms	0 - 100 MB	NPU
WideResNet50	TFLITE	float	Qualcomm® QCS8550 (Proxy)	4.775 ms	0 - 2 MB	NPU
WideResNet50	TFLITE	float	Qualcomm® SA8775P	50.623 ms	0 - 17 MB	GPU
WideResNet50	TFLITE	float	Qualcomm® SA8650P	50.623 ms	0 - 17 MB	GPU
WideResNet50	TFLITE	float	Qualcomm® SA8255P	50.623 ms	0 - 17 MB	GPU
WideResNet50	TFLITE	float	Qualcomm® QCS8450	8.171 ms	0 - 192 MB	NPU
WideResNet50	TFLITE	float	Snapdragon® 8 Elite Mobile	2.845 ms	0 - 107 MB	NPU
WideResNet50	TFLITE	float	Qualcomm® SA8295P	7.757 ms	0 - 88 MB	NPU
WideResNet50	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	2.523 ms	0 - 98 MB	NPU
WideResNet50	TFLITE	float	Qualcomm® SA7255P	23.802 ms	0 - 100 MB	NPU
WideResNet50	TFLITE	float	Qualcomm® QCS9075	7.054 ms	0 - 134 MB	NPU
WideResNet50	TFLITE	float	Qualcomm® QCS8750	2.845 ms	0 - 107 MB	NPU
WideResNet50	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	1.333 ms	0 - 130 MB	NPU
WideResNet50	TFLITE	w8a8	Snapdragon® 8 Gen 1 Mobile	2.388 ms	0 - 129 MB	NPU
WideResNet50	TFLITE	w8a8	Qualcomm® QCS6490	6.931 ms	0 - 68 MB	NPU
WideResNet50	TFLITE	w8a8	Qualcomm® QCS8275	3.728 ms	0 - 52 MB	NPU
WideResNet50	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	1.731 ms	0 - 2 MB	NPU
WideResNet50	TFLITE	w8a8	Qualcomm® SA8775P	50.869 ms	1 - 22 MB	GPU
WideResNet50	TFLITE	w8a8	Qualcomm® SA8650P	50.869 ms	1 - 22 MB	GPU
WideResNet50	TFLITE	w8a8	Qualcomm® SA8255P	50.869 ms	1 - 22 MB	GPU
WideResNet50	TFLITE	w8a8	Qualcomm® QCS8450	2.388 ms	0 - 129 MB	NPU
WideResNet50	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	1.045 ms	0 - 55 MB	NPU
WideResNet50	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	2.72 ms	0 - 207 MB	NPU
WideResNet50	TFLITE	w8a8	Qualcomm® QCM6690	16.724 ms	0 - 213 MB	NPU
WideResNet50	TFLITE	w8a8	Qualcomm® QCS9075	1.739 ms	0 - 68 MB	NPU
WideResNet50	TFLITE	w8a8	Qualcomm® SA7255P	3.728 ms	0 - 52 MB	NPU
WideResNet50	TFLITE	w8a8	Snapdragon® 8 Elite Mobile	1.123 ms	0 - 55 MB	NPU
WideResNet50	TFLITE	w8a8	Qualcomm® SA8295P	2.582 ms	0 - 50 MB	NPU
WideResNet50	TFLITE	w8a8	Qualcomm® QCS7790	2.72 ms	0 - 207 MB	NPU
WideResNet50	TFLITE	w8a8	Qualcomm® QCS8750	1.123 ms	0 - 55 MB	NPU

License

• The license for the original implementation of WideResNet50 can be found here.

References

• Wide Residual Networks
• 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.