EfficientFormer: Optimized for Qualcomm Devices

EfficientFormer is a vision transformer model that can classify images from the Imagenet dataset.

This is based on the implementation of EfficientFormer 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 EfficientFormer 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 EfficientFormer on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.image_classification

Model Stats:

• Model checkpoint: efficientformer_l1_300d
• Input resolution: 224x224
• Number of parameters: 12.3M
• Model size (float): 46.9 MB
• Model size (w8a16): 12.2 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
EfficientFormer	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	0.607 ms	0 - 52 MB	NPU
EfficientFormer	ONNX	float	Snapdragon® 8 Elite Mobile	0.696 ms	0 - 42 MB	NPU
EfficientFormer	ONNX	float	Snapdragon® X2 Elite	0.647 ms	25 - 25 MB	NPU
EfficientFormer	ONNX	float	Snapdragon® X Elite	1.557 ms	24 - 24 MB	NPU
EfficientFormer	ONNX	float	Snapdragon® X Elite	1.557 ms	24 - 24 MB	NPU
EfficientFormer	ONNX	float	Snapdragon® 8 Gen 3 Mobile	0.953 ms	0 - 89 MB	NPU
EfficientFormer	ONNX	float	Qualcomm® QCS8550 (Proxy)	1.308 ms	0 - 30 MB	NPU
EfficientFormer	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	0.696 ms	0 - 42 MB	NPU
EfficientFormer	ONNX	float	Qualcomm® QCS9075	1.872 ms	1 - 3 MB	NPU
EfficientFormer	ONNX	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	0.53 ms	0 - 67 MB	NPU
EfficientFormer	ONNX	w8a16	Snapdragon® 8 Elite Mobile	0.648 ms	0 - 57 MB	NPU
EfficientFormer	ONNX	w8a16	Snapdragon® X2 Elite	0.569 ms	12 - 12 MB	NPU
EfficientFormer	ONNX	w8a16	Snapdragon® X Elite	1.666 ms	12 - 12 MB	NPU
EfficientFormer	ONNX	w8a16	Snapdragon® X Elite	1.666 ms	12 - 12 MB	NPU
EfficientFormer	ONNX	w8a16	Snapdragon® 8 Gen 3 Mobile	0.972 ms	0 - 92 MB	NPU
EfficientFormer	ONNX	w8a16	Qualcomm® QCS6490	139.909 ms	20 - 25 MB	CPU
EfficientFormer	ONNX	w8a16	Qualcomm® QCS8550 (Proxy)	1.417 ms	0 - 19 MB	NPU
EfficientFormer	ONNX	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	0.648 ms	0 - 57 MB	NPU
EfficientFormer	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	62.902 ms	23 - 32 MB	CPU
EfficientFormer	ONNX	w8a16	Qualcomm® QCM6690	66.355 ms	11 - 20 MB	CPU
EfficientFormer	ONNX	w8a16	Qualcomm® QCS9075	1.642 ms	0 - 3 MB	NPU
EfficientFormer	ONNX	w8a16	Snapdragon® 7 Gen 4 Mobile	62.902 ms	23 - 32 MB	CPU
EfficientFormer	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	0.648 ms	1 - 44 MB	NPU
EfficientFormer	QNN_DLC	float	Snapdragon® 8 Elite Mobile	0.778 ms	1 - 44 MB	NPU
EfficientFormer	QNN_DLC	float	Snapdragon® X2 Elite	0.873 ms	1 - 1 MB	NPU
EfficientFormer	QNN_DLC	float	Snapdragon® X Elite	1.702 ms	1 - 1 MB	NPU
EfficientFormer	QNN_DLC	float	Snapdragon® X Elite	1.702 ms	1 - 1 MB	NPU
EfficientFormer	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	1.034 ms	0 - 75 MB	NPU
EfficientFormer	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	1.487 ms	1 - 2 MB	NPU
EfficientFormer	QNN_DLC	float	Qualcomm® SA8775P	2.044 ms	1 - 43 MB	NPU
EfficientFormer	QNN_DLC	float	Qualcomm® SA8775P	2.044 ms	1 - 43 MB	NPU
EfficientFormer	QNN_DLC	float	Qualcomm® SA8775P	2.044 ms	1 - 43 MB	NPU
EfficientFormer	QNN_DLC	float	Qualcomm® SA7255P	4.87 ms	1 - 40 MB	NPU
EfficientFormer	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	5.585 ms	0 - 78 MB	NPU
EfficientFormer	QNN_DLC	float	Qualcomm® SA8295P	4.019 ms	0 - 39 MB	NPU
EfficientFormer	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	0.778 ms	1 - 44 MB	NPU
EfficientFormer	QNN_DLC	float	Qualcomm® QCS9075	1.958 ms	1 - 3 MB	NPU
EfficientFormer	QNN_DLC	w8a16	Snapdragon® 8 Elite Gen 5 Mobile	0.594 ms	0 - 64 MB	NPU
EfficientFormer	QNN_DLC	w8a16	Snapdragon® 8 Elite Mobile	0.709 ms	0 - 56 MB	NPU
EfficientFormer	QNN_DLC	w8a16	Snapdragon® X2 Elite	0.852 ms	0 - 0 MB	NPU
EfficientFormer	QNN_DLC	w8a16	Snapdragon® X Elite	1.793 ms	0 - 0 MB	NPU
EfficientFormer	QNN_DLC	w8a16	Snapdragon® X Elite	1.793 ms	0 - 0 MB	NPU
EfficientFormer	QNN_DLC	w8a16	Snapdragon® 8 Gen 3 Mobile	1.075 ms	0 - 79 MB	NPU
EfficientFormer	QNN_DLC	w8a16	Qualcomm® QCS8550 (Proxy)	1.558 ms	0 - 51 MB	NPU
EfficientFormer	QNN_DLC	w8a16	Qualcomm® SA8775P	1.896 ms	0 - 59 MB	NPU
EfficientFormer	QNN_DLC	w8a16	Qualcomm® SA8775P	1.896 ms	0 - 59 MB	NPU
EfficientFormer	QNN_DLC	w8a16	Qualcomm® SA8775P	1.896 ms	0 - 59 MB	NPU
EfficientFormer	QNN_DLC	w8a16	Qualcomm® SA7255P	3.216 ms	0 - 57 MB	NPU
EfficientFormer	QNN_DLC	w8a16	Snapdragon® 8 Elite For Galaxy Mobile	0.709 ms	0 - 56 MB	NPU
EfficientFormer	QNN_DLC	w8a16	Snapdragon® 7 Gen 4 Mobile	1.698 ms	0 - 62 MB	NPU
EfficientFormer	QNN_DLC	w8a16	Qualcomm® QCM6690	7.005 ms	0 - 175 MB	NPU
EfficientFormer	QNN_DLC	w8a16	Qualcomm® QCS9075	1.763 ms	0 - 2 MB	NPU
EfficientFormer	QNN_DLC	w8a16	Snapdragon® 7 Gen 4 Mobile	1.698 ms	0 - 62 MB	NPU
EfficientFormer	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	0.649 ms	0 - 55 MB	NPU
EfficientFormer	TFLITE	float	Snapdragon® 8 Elite Mobile	0.781 ms	0 - 57 MB	NPU
EfficientFormer	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	1.044 ms	0 - 97 MB	NPU
EfficientFormer	TFLITE	float	Qualcomm® QCS8550 (Proxy)	1.484 ms	0 - 3 MB	NPU
EfficientFormer	TFLITE	float	Qualcomm® SA8775P	2.074 ms	0 - 54 MB	NPU
EfficientFormer	TFLITE	float	Qualcomm® SA8775P	2.074 ms	0 - 54 MB	NPU
EfficientFormer	TFLITE	float	Qualcomm® SA8775P	2.074 ms	0 - 54 MB	NPU
EfficientFormer	TFLITE	float	Qualcomm® SA7255P	4.863 ms	0 - 51 MB	NPU
EfficientFormer	TFLITE	float	Qualcomm® QCS8450 (Proxy)	5.566 ms	0 - 95 MB	NPU
EfficientFormer	TFLITE	float	Qualcomm® SA8295P	3.992 ms	0 - 48 MB	NPU
EfficientFormer	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	0.781 ms	0 - 57 MB	NPU
EfficientFormer	TFLITE	float	Qualcomm® QCS9075	1.949 ms	0 - 27 MB	NPU

License

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

References

• Rethinking Vision Transformers for MobileNet Size and Speed
• 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.