EfficientNet-V2-s / README.md

v0.46.0

64da2ba verified about 10 hours ago

8.3 kB

	---
	library_name: pytorch
	license: other
	tags:
	- backbone
	- bu_auto
	- android
	pipeline_tag: image-classification

	---

	![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/efficientnet_v2_s/web-assets/model_demo.png)

	# EfficientNet-V2-s: Optimized for Qualcomm Devices

	EfficientNetV2-s 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 EfficientNet-V2-s found [here](https://github.com/pytorch/vision/blob/main/torchvision/models/efficientnet.py).
	This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/efficientnet_v2_s) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).

	Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.com) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) 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 2.37, ONNX Runtime 1.23.0 \| [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/efficientnet_v2_s/releases/v0.46.0/efficientnet_v2_s-onnx-float.zip)
	\| ONNX \| w8a16 \| Universal \| QAIRT 2.37, ONNX Runtime 1.23.0 \| [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/efficientnet_v2_s/releases/v0.46.0/efficientnet_v2_s-onnx-w8a16.zip)
	\| QNN_DLC \| float \| Universal \| QAIRT 2.42 \| [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/efficientnet_v2_s/releases/v0.46.0/efficientnet_v2_s-qnn_dlc-float.zip)
	\| QNN_DLC \| w8a16 \| Universal \| QAIRT 2.42 \| [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/efficientnet_v2_s/releases/v0.46.0/efficientnet_v2_s-qnn_dlc-w8a16.zip)
	\| TFLITE \| float \| Universal \| QAIRT 2.42, TFLite 2.17.0 \| [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/efficientnet_v2_s/releases/v0.46.0/efficientnet_v2_s-tflite-float.zip)

	For more device-specific assets and performance metrics, visit [EfficientNet-V2-s on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/efficientnet_v2_s).


	### Option 2: Export with Custom Configurations

	Use the [Qualcomm® AI Hub Models](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/efficientnet_v2_s) 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 [EfficientNet-V2-s on GitHub](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/efficientnet_v2_s) for usage instructions.

	## Model Details

	Model Type: Model_use_case.image_classification

	Model Stats:
	- Model checkpoint: Imagenet
	- Input resolution: 384x384
	- Number of parameters: 21.4M
	- Model size (float): 81.7 MB
	- Model size (w8a16): 27.2 MB

	## Performance Summary
	\| Model \| Runtime \| Precision \| Chipset \| Inference Time (ms) \| Peak Memory Range (MB) \| Primary Compute Unit
	\|---\|---\|---\|---\|---\|---\|---
	\| EfficientNet-V2-s \| ONNX \| float \| Snapdragon® X Elite \| 2.668 ms \| 47 - 47 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| float \| Snapdragon® 8 Gen 3 Mobile \| 2.012 ms \| 0 - 211 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| float \| Qualcomm® QCS8550 (Proxy) \| 2.716 ms \| 0 - 134 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| float \| Qualcomm® QCS9075 \| 3.615 ms \| 0 - 4 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| float \| Snapdragon® 8 Elite For Galaxy Mobile \| 1.591 ms \| 0 - 134 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| float \| Snapdragon® 8 Elite Gen 5 Mobile \| 1.358 ms \| 0 - 134 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| w8a16 \| Snapdragon® X Elite \| 2.631 ms \| 24 - 24 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| w8a16 \| Snapdragon® 8 Gen 3 Mobile \| 1.826 ms \| 0 - 228 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| w8a16 \| Qualcomm® QCS6490 \| 311.643 ms \| 25 - 32 MB \| CPU
	\| EfficientNet-V2-s \| ONNX \| w8a16 \| Qualcomm® QCS8550 (Proxy) \| 2.624 ms \| 0 - 31 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| w8a16 \| Qualcomm® QCS9075 \| 2.989 ms \| 0 - 3 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| w8a16 \| Qualcomm® QCM6690 \| 142.621 ms \| 13 - 26 MB \| CPU
	\| EfficientNet-V2-s \| ONNX \| w8a16 \| Snapdragon® 8 Elite For Galaxy Mobile \| 1.335 ms \| 0 - 183 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| w8a16 \| Snapdragon® 7 Gen 4 Mobile \| 130.814 ms \| 15 - 28 MB \| CPU
	\| EfficientNet-V2-s \| ONNX \| w8a16 \| Snapdragon® 8 Elite Gen 5 Mobile \| 1.121 ms \| 0 - 183 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| float \| Snapdragon® X Elite \| 2.943 ms \| 1 - 1 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| float \| Snapdragon® 8 Gen 3 Mobile \| 1.959 ms \| 0 - 145 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| float \| Qualcomm® QCS8275 (Proxy) \| 10.921 ms \| 1 - 66 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| float \| Qualcomm® QCS8550 (Proxy) \| 2.671 ms \| 1 - 2 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| float \| Qualcomm® QCS9075 \| 3.659 ms \| 1 - 3 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| float \| Qualcomm® QCS8450 (Proxy) \| 5.634 ms \| 0 - 155 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| float \| Snapdragon® 8 Elite For Galaxy Mobile \| 1.549 ms \| 0 - 68 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| float \| Snapdragon® 8 Elite Gen 5 Mobile \| 1.218 ms \| 0 - 69 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Snapdragon® X Elite \| 2.929 ms \| 0 - 0 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Snapdragon® 8 Gen 3 Mobile \| 1.775 ms \| 0 - 147 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Qualcomm® QCS6490 \| 6.816 ms \| 0 - 2 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Qualcomm® QCS8275 (Proxy) \| 5.339 ms \| 0 - 105 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Qualcomm® QCS8550 (Proxy) \| 2.61 ms \| 0 - 2 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Qualcomm® QCS9075 \| 2.937 ms \| 0 - 2 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Qualcomm® QCM6690 \| 14.296 ms \| 0 - 226 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Qualcomm® QCS8450 (Proxy) \| 3.23 ms \| 0 - 153 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Snapdragon® 8 Elite For Galaxy Mobile \| 1.235 ms \| 0 - 108 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Snapdragon® 7 Gen 4 Mobile \| 2.919 ms \| 0 - 105 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Snapdragon® 8 Elite Gen 5 Mobile \| 1.001 ms \| 0 - 109 MB \| NPU
	\| EfficientNet-V2-s \| TFLITE \| float \| Snapdragon® 8 Gen 3 Mobile \| 1.952 ms \| 0 - 196 MB \| NPU
	\| EfficientNet-V2-s \| TFLITE \| float \| Qualcomm® QCS8275 (Proxy) \| 10.991 ms \| 0 - 113 MB \| NPU
	\| EfficientNet-V2-s \| TFLITE \| float \| Qualcomm® QCS8550 (Proxy) \| 2.632 ms \| 0 - 2 MB \| NPU
	\| EfficientNet-V2-s \| TFLITE \| float \| Qualcomm® QCS9075 \| 3.673 ms \| 0 - 50 MB \| NPU
	\| EfficientNet-V2-s \| TFLITE \| float \| Qualcomm® QCS8450 (Proxy) \| 5.714 ms \| 0 - 205 MB \| NPU
	\| EfficientNet-V2-s \| TFLITE \| float \| Snapdragon® 8 Elite For Galaxy Mobile \| 1.504 ms \| 0 - 120 MB \| NPU
	\| EfficientNet-V2-s \| TFLITE \| float \| Snapdragon® 8 Elite Gen 5 Mobile \| 1.226 ms \| 0 - 117 MB \| NPU

	## License
	* The license for the original implementation of EfficientNet-V2-s can be found
	[here](https://github.com/pytorch/vision/blob/main/LICENSE).

	## References
	* [EfficientNetV2: Smaller Models and Faster Training](https://arxiv.org/abs/2104.00298)
	* [Source Model Implementation](https://github.com/pytorch/vision/blob/main/torchvision/models/efficientnet.py)

	## Community
	* Join [our AI Hub Slack community](https://aihub.qualcomm.com/community/slack) to collaborate, post questions and learn more about on-device AI.
	* For questions or feedback please [reach out to us](mailto:ai-hub-support@qti.qualcomm.com).

	---
	library_name: pytorch
	license: other
	tags:
	- backbone
	- bu_auto
	- android
	pipeline_tag: image-classification

	---

	![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/efficientnet_v2_s/web-assets/model_demo.png)

	# EfficientNet-V2-s: Optimized for Qualcomm Devices

	EfficientNetV2-s 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 EfficientNet-V2-s found [here](https://github.com/pytorch/vision/blob/main/torchvision/models/efficientnet.py).
	This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/efficientnet_v2_s) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).

	Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.com) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) 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 2.37, ONNX Runtime 1.23.0 \| [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/efficientnet_v2_s/releases/v0.46.0/efficientnet_v2_s-onnx-float.zip)
	\| ONNX \| w8a16 \| Universal \| QAIRT 2.37, ONNX Runtime 1.23.0 \| [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/efficientnet_v2_s/releases/v0.46.0/efficientnet_v2_s-onnx-w8a16.zip)
	\| QNN_DLC \| float \| Universal \| QAIRT 2.42 \| [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/efficientnet_v2_s/releases/v0.46.0/efficientnet_v2_s-qnn_dlc-float.zip)
	\| QNN_DLC \| w8a16 \| Universal \| QAIRT 2.42 \| [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/efficientnet_v2_s/releases/v0.46.0/efficientnet_v2_s-qnn_dlc-w8a16.zip)
	\| TFLITE \| float \| Universal \| QAIRT 2.42, TFLite 2.17.0 \| [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/efficientnet_v2_s/releases/v0.46.0/efficientnet_v2_s-tflite-float.zip)

	For more device-specific assets and performance metrics, visit [EfficientNet-V2-s on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/efficientnet_v2_s).


	### Option 2: Export with Custom Configurations

	Use the [Qualcomm® AI Hub Models](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/efficientnet_v2_s) 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 [EfficientNet-V2-s on GitHub](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/efficientnet_v2_s) for usage instructions.

	## Model Details

	Model Type: Model_use_case.image_classification

	Model Stats:
	- Model checkpoint: Imagenet
	- Input resolution: 384x384
	- Number of parameters: 21.4M
	- Model size (float): 81.7 MB
	- Model size (w8a16): 27.2 MB

	## Performance Summary
	\| Model \| Runtime \| Precision \| Chipset \| Inference Time (ms) \| Peak Memory Range (MB) \| Primary Compute Unit
	\|---\|---\|---\|---\|---\|---\|---
	\| EfficientNet-V2-s \| ONNX \| float \| Snapdragon® X Elite \| 2.668 ms \| 47 - 47 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| float \| Snapdragon® 8 Gen 3 Mobile \| 2.012 ms \| 0 - 211 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| float \| Qualcomm® QCS8550 (Proxy) \| 2.716 ms \| 0 - 134 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| float \| Qualcomm® QCS9075 \| 3.615 ms \| 0 - 4 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| float \| Snapdragon® 8 Elite For Galaxy Mobile \| 1.591 ms \| 0 - 134 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| float \| Snapdragon® 8 Elite Gen 5 Mobile \| 1.358 ms \| 0 - 134 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| w8a16 \| Snapdragon® X Elite \| 2.631 ms \| 24 - 24 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| w8a16 \| Snapdragon® 8 Gen 3 Mobile \| 1.826 ms \| 0 - 228 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| w8a16 \| Qualcomm® QCS6490 \| 311.643 ms \| 25 - 32 MB \| CPU
	\| EfficientNet-V2-s \| ONNX \| w8a16 \| Qualcomm® QCS8550 (Proxy) \| 2.624 ms \| 0 - 31 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| w8a16 \| Qualcomm® QCS9075 \| 2.989 ms \| 0 - 3 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| w8a16 \| Qualcomm® QCM6690 \| 142.621 ms \| 13 - 26 MB \| CPU
	\| EfficientNet-V2-s \| ONNX \| w8a16 \| Snapdragon® 8 Elite For Galaxy Mobile \| 1.335 ms \| 0 - 183 MB \| NPU
	\| EfficientNet-V2-s \| ONNX \| w8a16 \| Snapdragon® 7 Gen 4 Mobile \| 130.814 ms \| 15 - 28 MB \| CPU
	\| EfficientNet-V2-s \| ONNX \| w8a16 \| Snapdragon® 8 Elite Gen 5 Mobile \| 1.121 ms \| 0 - 183 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| float \| Snapdragon® X Elite \| 2.943 ms \| 1 - 1 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| float \| Snapdragon® 8 Gen 3 Mobile \| 1.959 ms \| 0 - 145 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| float \| Qualcomm® QCS8275 (Proxy) \| 10.921 ms \| 1 - 66 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| float \| Qualcomm® QCS8550 (Proxy) \| 2.671 ms \| 1 - 2 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| float \| Qualcomm® QCS9075 \| 3.659 ms \| 1 - 3 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| float \| Qualcomm® QCS8450 (Proxy) \| 5.634 ms \| 0 - 155 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| float \| Snapdragon® 8 Elite For Galaxy Mobile \| 1.549 ms \| 0 - 68 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| float \| Snapdragon® 8 Elite Gen 5 Mobile \| 1.218 ms \| 0 - 69 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Snapdragon® X Elite \| 2.929 ms \| 0 - 0 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Snapdragon® 8 Gen 3 Mobile \| 1.775 ms \| 0 - 147 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Qualcomm® QCS6490 \| 6.816 ms \| 0 - 2 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Qualcomm® QCS8275 (Proxy) \| 5.339 ms \| 0 - 105 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Qualcomm® QCS8550 (Proxy) \| 2.61 ms \| 0 - 2 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Qualcomm® QCS9075 \| 2.937 ms \| 0 - 2 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Qualcomm® QCM6690 \| 14.296 ms \| 0 - 226 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Qualcomm® QCS8450 (Proxy) \| 3.23 ms \| 0 - 153 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Snapdragon® 8 Elite For Galaxy Mobile \| 1.235 ms \| 0 - 108 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Snapdragon® 7 Gen 4 Mobile \| 2.919 ms \| 0 - 105 MB \| NPU
	\| EfficientNet-V2-s \| QNN_DLC \| w8a16 \| Snapdragon® 8 Elite Gen 5 Mobile \| 1.001 ms \| 0 - 109 MB \| NPU
	\| EfficientNet-V2-s \| TFLITE \| float \| Snapdragon® 8 Gen 3 Mobile \| 1.952 ms \| 0 - 196 MB \| NPU
	\| EfficientNet-V2-s \| TFLITE \| float \| Qualcomm® QCS8275 (Proxy) \| 10.991 ms \| 0 - 113 MB \| NPU
	\| EfficientNet-V2-s \| TFLITE \| float \| Qualcomm® QCS8550 (Proxy) \| 2.632 ms \| 0 - 2 MB \| NPU
	\| EfficientNet-V2-s \| TFLITE \| float \| Qualcomm® QCS9075 \| 3.673 ms \| 0 - 50 MB \| NPU
	\| EfficientNet-V2-s \| TFLITE \| float \| Qualcomm® QCS8450 (Proxy) \| 5.714 ms \| 0 - 205 MB \| NPU
	\| EfficientNet-V2-s \| TFLITE \| float \| Snapdragon® 8 Elite For Galaxy Mobile \| 1.504 ms \| 0 - 120 MB \| NPU
	\| EfficientNet-V2-s \| TFLITE \| float \| Snapdragon® 8 Elite Gen 5 Mobile \| 1.226 ms \| 0 - 117 MB \| NPU

	## License
	* The license for the original implementation of EfficientNet-V2-s can be found
	[here](https://github.com/pytorch/vision/blob/main/LICENSE).

	## References
	* [EfficientNetV2: Smaller Models and Faster Training](https://arxiv.org/abs/2104.00298)
	* [Source Model Implementation](https://github.com/pytorch/vision/blob/main/torchvision/models/efficientnet.py)

	## Community
	* Join [our AI Hub Slack community](https://aihub.qualcomm.com/community/slack) to collaborate, post questions and learn more about on-device AI.
	* For questions or feedback please [reach out to us](mailto:ai-hub-support@qti.qualcomm.com).