EfficientViT-b2-cls: Optimized for Qualcomm Devices

EfficientViT 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 EfficientViT-b2-cls 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 EfficientViT-b2-cls 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 EfficientViT-b2-cls 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): 92.9 MB
Number of parameters: 24.3M

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
EfficientViT-b2-cls	ONNX	float	Snapdragon® X2 Elite	2.526 ms	2 - 2 MB	NPU
EfficientViT-b2-cls	ONNX	float	Snapdragon® X Elite	5.063 ms	49 - 49 MB	NPU
EfficientViT-b2-cls	ONNX	float	Snapdragon® 8 Gen 3 Mobile	3.229 ms	0 - 140 MB	NPU
EfficientViT-b2-cls	ONNX	float	Snapdragon® 8 Gen 1 Mobile	6.33 ms	1 - 143 MB	NPU
EfficientViT-b2-cls	ONNX	float	Qualcomm® Dragonwing™ IQ-8275	5.062 ms	0 - 4 MB	NPU
EfficientViT-b2-cls	ONNX	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	4.894 ms	0 - 58 MB	NPU
EfficientViT-b2-cls	ONNX	float	Qualcomm® QCS8450	6.33 ms	1 - 143 MB	NPU
EfficientViT-b2-cls	ONNX	float	Qualcomm® Dragonwing™ IQ-9075	5.518 ms	1 - 4 MB	NPU
EfficientViT-b2-cls	ONNX	float	Qualcomm® Dragonwing™ IQ-X7181	5.063 ms	49 - 49 MB	NPU
EfficientViT-b2-cls	ONNX	float	Qualcomm® Dragonwing™ Q-8750	2.687 ms	0 - 72 MB	NPU
EfficientViT-b2-cls	ONNX	float	Snapdragon® 8 Elite Mobile	2.687 ms	0 - 72 MB	NPU
EfficientViT-b2-cls	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	2.443 ms	0 - 71 MB	NPU
EfficientViT-b2-cls	ONNX	w8a16	Snapdragon® X2 Elite	2.345 ms	1 - 1 MB	NPU
EfficientViT-b2-cls	ONNX	w8a16	Snapdragon® X Elite	4.847 ms	27 - 27 MB	NPU
EfficientViT-b2-cls	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	3.085 ms	0 - 159 MB	NPU
EfficientViT-b2-cls	ONNX	w8a16	Snapdragon® 8 Gen 1 Mobile	5.774 ms	0 - 169 MB	NPU
EfficientViT-b2-cls	ONNX	w8a16	Qualcomm® Dragonwing™ QCS6490	15.328 ms	0 - 3 MB	NPU
EfficientViT-b2-cls	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-8275	4.218 ms	0 - 4 MB	NPU
EfficientViT-b2-cls	ONNX	w8a16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	4.636 ms	0 - 47 MB	NPU
EfficientViT-b2-cls	ONNX	w8a16	Qualcomm® QCS8450	5.774 ms	0 - 169 MB	NPU
EfficientViT-b2-cls	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-9075	4.942 ms	0 - 3 MB	NPU
EfficientViT-b2-cls	ONNX	w8a16	Qualcomm® Dragonwing™ IQ-X7181	4.847 ms	27 - 27 MB	NPU
EfficientViT-b2-cls	ONNX	w8a16	Qualcomm® Dragonwing™ Q-6690	19.303 ms	0 - 256 MB	NPU
EfficientViT-b2-cls	ONNX	w8a16	Qualcomm® Dragonwing™ Q-7790	4.988 ms	0 - 116 MB	NPU
EfficientViT-b2-cls	ONNX	w8a16	Qualcomm® Dragonwing™ Q-8750	2.531 ms	0 - 113 MB	NPU
EfficientViT-b2-cls	ONNX	w8a16	Snapdragon® 8 Elite Mobile	2.531 ms	0 - 113 MB	NPU
EfficientViT-b2-cls	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	2.171 ms	0 - 116 MB	NPU
EfficientViT-b2-cls	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	4.988 ms	0 - 116 MB	NPU
EfficientViT-b2-cls	QNN_DLC	float	Snapdragon® X2 Elite	2.954 ms	1 - 1 MB	NPU
EfficientViT-b2-cls	QNN_DLC	float	Snapdragon® X Elite	6.331 ms	1 - 1 MB	NPU
EfficientViT-b2-cls	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	3.746 ms	0 - 140 MB	NPU
EfficientViT-b2-cls	QNN_DLC	float	Snapdragon® 8 Gen 1 Mobile	7.153 ms	0 - 144 MB	NPU
EfficientViT-b2-cls	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-8275	6.191 ms	1 - 4 MB	NPU
EfficientViT-b2-cls	QNN_DLC	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	5.387 ms	1 - 2 MB	NPU
EfficientViT-b2-cls	QNN_DLC	float	Qualcomm® SA8775P	6.803 ms	1 - 69 MB	NPU
EfficientViT-b2-cls	QNN_DLC	float	Qualcomm® SA8650P	6.803 ms	1 - 69 MB	NPU
EfficientViT-b2-cls	QNN_DLC	float	Qualcomm® SA8255P	6.803 ms	1 - 69 MB	NPU
EfficientViT-b2-cls	QNN_DLC	float	Qualcomm® QCS8450	7.153 ms	0 - 144 MB	NPU
EfficientViT-b2-cls	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-9075	6.542 ms	3 - 5 MB	NPU
EfficientViT-b2-cls	QNN_DLC	float	Qualcomm® Dragonwing™ IQ-X7181	6.331 ms	1 - 1 MB	NPU
EfficientViT-b2-cls	QNN_DLC	float	Qualcomm® Dragonwing™ Q-8750	2.763 ms	1 - 71 MB	NPU
EfficientViT-b2-cls	QNN_DLC	float	Qualcomm® SA7255P	12.841 ms	1 - 67 MB	NPU
EfficientViT-b2-cls	QNN_DLC	float	Qualcomm® SA8295P	7.394 ms	1 - 73 MB	NPU
EfficientViT-b2-cls	QNN_DLC	float	Snapdragon® 8 Elite Mobile	2.763 ms	1 - 71 MB	NPU
EfficientViT-b2-cls	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	2.336 ms	1 - 72 MB	NPU
EfficientViT-b2-cls	QNN_DLC	w8a16	Snapdragon® X2 Elite	2.711 ms	0 - 0 MB	NPU
EfficientViT-b2-cls	QNN_DLC	w8a16	Snapdragon® X Elite	5.444 ms	0 - 0 MB	NPU
EfficientViT-b2-cls	QNN_DLC	w8a16	Snapdragon® 8 Gen 3 Mobile	3.296 ms	0 - 144 MB	NPU
EfficientViT-b2-cls	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-8275	4.425 ms	0 - 3 MB	NPU
EfficientViT-b2-cls	QNN_DLC	w8a16	Qualcomm® Dragonwing™ QCS8550 (Proxy)	4.971 ms	0 - 92 MB	NPU
EfficientViT-b2-cls	QNN_DLC	w8a16	Qualcomm® SA8775P	5.456 ms	0 - 106 MB	NPU
EfficientViT-b2-cls	QNN_DLC	w8a16	Qualcomm® SA8650P	5.456 ms	0 - 106 MB	NPU
EfficientViT-b2-cls	QNN_DLC	w8a16	Qualcomm® SA8255P	5.456 ms	0 - 106 MB	NPU
EfficientViT-b2-cls	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-9075	5.287 ms	0 - 2 MB	NPU
EfficientViT-b2-cls	QNN_DLC	w8a16	Qualcomm® Dragonwing™ IQ-X7181	5.444 ms	0 - 0 MB	NPU
EfficientViT-b2-cls	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-6690	23.752 ms	0 - 229 MB	NPU
EfficientViT-b2-cls	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-7790	5.179 ms	0 - 100 MB	NPU
EfficientViT-b2-cls	QNN_DLC	w8a16	Qualcomm® Dragonwing™ Q-8750	2.52 ms	0 - 107 MB	NPU
EfficientViT-b2-cls	QNN_DLC	w8a16	Qualcomm® SA7255P	9.255 ms	0 - 104 MB	NPU
EfficientViT-b2-cls	QNN_DLC	w8a16	Snapdragon® 8 Elite Mobile	2.52 ms	0 - 107 MB	NPU
EfficientViT-b2-cls	QNN_DLC	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	2.117 ms	0 - 106 MB	NPU
EfficientViT-b2-cls	QNN_DLC	w8a16	Snapdragon® 7 Gen 4 Mobile	5.179 ms	0 - 100 MB	NPU
EfficientViT-b2-cls	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	3.722 ms	0 - 177 MB	NPU
EfficientViT-b2-cls	TFLITE	float	Snapdragon® 8 Gen 1 Mobile	7.128 ms	0 - 187 MB	NPU
EfficientViT-b2-cls	TFLITE	float	Qualcomm® Dragonwing™ IQ-8275	6.168 ms	0 - 52 MB	NPU
EfficientViT-b2-cls	TFLITE	float	Qualcomm® Dragonwing™ QCS8550 (Proxy)	5.477 ms	0 - 3 MB	NPU
EfficientViT-b2-cls	TFLITE	float	Qualcomm® SA8775P	6.819 ms	0 - 111 MB	NPU
EfficientViT-b2-cls	TFLITE	float	Qualcomm® SA8650P	6.819 ms	0 - 111 MB	NPU
EfficientViT-b2-cls	TFLITE	float	Qualcomm® SA8255P	6.819 ms	0 - 111 MB	NPU
EfficientViT-b2-cls	TFLITE	float	Qualcomm® QCS8450	7.128 ms	0 - 187 MB	NPU
EfficientViT-b2-cls	TFLITE	float	Qualcomm® Dragonwing™ IQ-9075	6.518 ms	0 - 52 MB	NPU
EfficientViT-b2-cls	TFLITE	float	Qualcomm® Dragonwing™ Q-8750	2.767 ms	0 - 115 MB	NPU
EfficientViT-b2-cls	TFLITE	float	Qualcomm® SA7255P	12.878 ms	0 - 110 MB	NPU
EfficientViT-b2-cls	TFLITE	float	Qualcomm® SA8295P	7.46 ms	0 - 115 MB	NPU
EfficientViT-b2-cls	TFLITE	float	Snapdragon® 8 Elite Mobile	2.767 ms	0 - 115 MB	NPU
EfficientViT-b2-cls	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	2.358 ms	0 - 118 MB	NPU

License

The license for the original implementation of EfficientViT-b2-cls can be found here.

References

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.

Downloads last month: -; Downloads are not tracked for this model. How to track

Paper for qualcomm/EfficientViT-b2-cls

EfficientViT: Lightweight Multi-Scale Attention for On-Device Semantic Segmentation

Paper • 2205.14756 • Published May 29, 2022 • 1