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 2.42, ONNX Runtime 1.24.1	Download
ONNX	w8a8	Universal	QAIRT 2.42, ONNX Runtime 1.24.1	Download
QNN_DLC	float	Universal	QAIRT 2.43	Download
QNN_DLC	w8a8	Universal	QAIRT 2.43	Download
TFLITE	float	Universal	QAIRT 2.43, TFLite 2.17.0	Download
TFLITE	w8a8	Universal	QAIRT 2.43, TFLite 2.17.0	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	2.928 ms	132 - 132 MB	NPU
WideResNet50	ONNX	float	Snapdragon® X Elite	4.443 ms	132 - 132 MB	NPU
WideResNet50	ONNX	float	Snapdragon® 8 Gen 3 Mobile	3.487 ms	0 - 227 MB	NPU
WideResNet50	ONNX	float	Qualcomm® QCS8550 (Proxy)	4.573 ms	1 - 3 MB	NPU
WideResNet50	ONNX	float	Qualcomm® QCS9075	6.779 ms	0 - 4 MB	NPU
WideResNet50	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	2.818 ms	0 - 168 MB	NPU
WideResNet50	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	2.634 ms	1 - 189 MB	NPU
WideResNet50	ONNX	w8a8	Snapdragon® X2 Elite	0.811 ms	66 - 66 MB	NPU
WideResNet50	ONNX	w8a8	Snapdragon® X Elite	1.798 ms	66 - 66 MB	NPU
WideResNet50	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	1.415 ms	0 - 131 MB	NPU
WideResNet50	ONNX	w8a8	Qualcomm® QCS6490	76.112 ms	10 - 106 MB	CPU
WideResNet50	ONNX	w8a8	Qualcomm® QCS8550 (Proxy)	1.853 ms	0 - 42 MB	NPU
WideResNet50	ONNX	w8a8	Qualcomm® QCS9075	1.919 ms	0 - 3 MB	NPU
WideResNet50	ONNX	w8a8	Qualcomm® QCM6690	61.279 ms	10 - 19 MB	CPU
WideResNet50	ONNX	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	1.223 ms	0 - 52 MB	NPU
WideResNet50	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	58.096 ms	8 - 17 MB	CPU
WideResNet50	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	1.156 ms	0 - 56 MB	NPU
WideResNet50	QNN_DLC	float	Snapdragon® X2 Elite	2.523 ms	1 - 1 MB	NPU
WideResNet50	QNN_DLC	float	Snapdragon® X Elite	4.756 ms	1 - 1 MB	NPU
WideResNet50	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	3.647 ms	0 - 227 MB	NPU
WideResNet50	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	24.586 ms	1 - 175 MB	NPU
WideResNet50	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	4.864 ms	1 - 2 MB	NPU
WideResNet50	QNN_DLC	float	Qualcomm® SA8775P	7.16 ms	1 - 170 MB	NPU
WideResNet50	QNN_DLC	float	Qualcomm® QCS9075	7.076 ms	1 - 3 MB	NPU
WideResNet50	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	9.2 ms	0 - 197 MB	NPU
WideResNet50	QNN_DLC	float	Qualcomm® SA7255P	24.586 ms	1 - 175 MB	NPU
WideResNet50	QNN_DLC	float	Qualcomm® SA8295P	7.903 ms	1 - 153 MB	NPU
WideResNet50	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	2.926 ms	1 - 177 MB	NPU
WideResNet50	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	2.544 ms	1 - 172 MB	NPU
WideResNet50	QNN_DLC	w8a8	Snapdragon® X2 Elite	0.871 ms	0 - 0 MB	NPU
WideResNet50	QNN_DLC	w8a8	Snapdragon® X Elite	1.801 ms	0 - 0 MB	NPU
WideResNet50	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	1.42 ms	0 - 127 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® QCS6490	7.823 ms	0 - 2 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® QCS8275 (Proxy)	3.953 ms	0 - 52 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	1.861 ms	0 - 2 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® SA8775P	2.008 ms	0 - 52 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® QCS9075	1.907 ms	0 - 2 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® QCM6690	17.133 ms	0 - 212 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® QCS8450 (Proxy)	2.542 ms	0 - 128 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® SA7255P	3.953 ms	0 - 52 MB	NPU
WideResNet50	QNN_DLC	w8a8	Qualcomm® SA8295P	2.743 ms	0 - 49 MB	NPU
WideResNet50	QNN_DLC	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	1.145 ms	0 - 53 MB	NPU
WideResNet50	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	2.864 ms	0 - 210 MB	NPU
WideResNet50	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	1.099 ms	0 - 53 MB	NPU
WideResNet50	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	3.566 ms	0 - 253 MB	NPU
WideResNet50	TFLITE	float	Qualcomm® QCS8275 (Proxy)	23.673 ms	0 - 115 MB	NPU
WideResNet50	TFLITE	float	Qualcomm® QCS8550 (Proxy)	4.771 ms	0 - 2 MB	NPU
WideResNet50	TFLITE	float	Qualcomm® SA8775P	7.037 ms	0 - 115 MB	NPU
WideResNet50	TFLITE	float	Qualcomm® QCS9075	7.064 ms	0 - 134 MB	NPU
WideResNet50	TFLITE	float	Qualcomm® QCS8450 (Proxy)	8.178 ms	0 - 230 MB	NPU
WideResNet50	TFLITE	float	Qualcomm® SA7255P	23.673 ms	0 - 115 MB	NPU
WideResNet50	TFLITE	float	Qualcomm® SA8295P	7.682 ms	0 - 101 MB	NPU
WideResNet50	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	2.881 ms	0 - 120 MB	NPU
WideResNet50	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	2.5 ms	0 - 118 MB	NPU
WideResNet50	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	1.319 ms	0 - 128 MB	NPU
WideResNet50	TFLITE	w8a8	Qualcomm® QCS6490	6.808 ms	0 - 68 MB	NPU
WideResNet50	TFLITE	w8a8	Qualcomm® QCS8275 (Proxy)	3.776 ms	0 - 50 MB	NPU
WideResNet50	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	1.748 ms	0 - 2 MB	NPU
WideResNet50	TFLITE	w8a8	Qualcomm® SA8775P	1.902 ms	0 - 51 MB	NPU
WideResNet50	TFLITE	w8a8	Qualcomm® QCS9075	1.744 ms	0 - 68 MB	NPU
WideResNet50	TFLITE	w8a8	Qualcomm® QCM6690	16.895 ms	0 - 211 MB	NPU
WideResNet50	TFLITE	w8a8	Qualcomm® QCS8450 (Proxy)	2.401 ms	0 - 129 MB	NPU
WideResNet50	TFLITE	w8a8	Qualcomm® SA7255P	3.776 ms	0 - 50 MB	NPU
WideResNet50	TFLITE	w8a8	Qualcomm® SA8295P	2.574 ms	0 - 54 MB	NPU
WideResNet50	TFLITE	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	1.112 ms	0 - 52 MB	NPU
WideResNet50	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	2.747 ms	0 - 205 MB	NPU
WideResNet50	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	1.05 ms	0 - 53 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.