README.md

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

---

![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/efficientvit_l2_cls/web-assets/model_demo.png)

# EfficientViT-l2-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-l2-cls found [here](https://github.com/CVHub520/efficientvit).
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/efficientvit_l2_cls) 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/efficientvit_l2_cls/releases/v0.46.0/efficientvit_l2_cls-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/efficientvit_l2_cls/releases/v0.46.0/efficientvit_l2_cls-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/efficientvit_l2_cls/releases/v0.46.0/efficientvit_l2_cls-qnn_dlc-float.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/efficientvit_l2_cls/releases/v0.46.0/efficientvit_l2_cls-tflite-float.zip)

For more device-specific assets and performance metrics, visit **[EfficientViT-l2-cls on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/efficientvit_l2_cls)**.


### 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/efficientvit_l2_cls) 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-l2-cls on GitHub](https://github.com/quic/ai-hub-models/blob/main/qai_hub_models/models/efficientvit_l2_cls) for usage instructions.

## Model Details

**Model Type:** Model_use_case.image_classification

**Model Stats:**
- Model checkpoint: Imagenet
- Input resolution: 224x224
- Number of parameters: 63.7M
- Model size (float): 243 MB

## Performance Summary
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
|---|---|---|---|---|---|---
| EfficientViT-l2-cls | ONNX | float | Snapdragon® X Elite | 7.908 ms | 132 - 132 MB | NPU
| EfficientViT-l2-cls | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 5.393 ms | 0 - 282 MB | NPU
| EfficientViT-l2-cls | ONNX | float | Qualcomm® QCS8550 (Proxy) | 7.558 ms | 0 - 162 MB | NPU
| EfficientViT-l2-cls | ONNX | float | Qualcomm® QCS9075 | 8.774 ms | 0 - 4 MB | NPU
| EfficientViT-l2-cls | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 4.115 ms | 0 - 272 MB | NPU
| EfficientViT-l2-cls | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 3.462 ms | 0 - 192 MB | NPU
| EfficientViT-l2-cls | QNN_DLC | float | Snapdragon® X Elite | 8.198 ms | 1 - 1 MB | NPU
| EfficientViT-l2-cls | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 5.405 ms | 0 - 234 MB | NPU
| EfficientViT-l2-cls | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 24.607 ms | 1 - 139 MB | NPU
| EfficientViT-l2-cls | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 7.536 ms | 1 - 250 MB | NPU
| EfficientViT-l2-cls | QNN_DLC | float | Qualcomm® QCS9075 | 8.599 ms | 1 - 3 MB | NPU
| EfficientViT-l2-cls | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 14.884 ms | 0 - 221 MB | NPU
| EfficientViT-l2-cls | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 3.997 ms | 0 - 219 MB | NPU
| EfficientViT-l2-cls | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 3.262 ms | 1 - 144 MB | NPU
| EfficientViT-l2-cls | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 5.345 ms | 0 - 372 MB | NPU
| EfficientViT-l2-cls | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 24.563 ms | 0 - 274 MB | NPU
| EfficientViT-l2-cls | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 7.475 ms | 0 - 3 MB | NPU
| EfficientViT-l2-cls | TFLITE | float | Qualcomm® QCS9075 | 8.565 ms | 0 - 134 MB | NPU
| EfficientViT-l2-cls | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 14.822 ms | 0 - 348 MB | NPU
| EfficientViT-l2-cls | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 3.984 ms | 0 - 275 MB | NPU
| EfficientViT-l2-cls | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 3.239 ms | 0 - 278 MB | NPU

## License
* The license for the original implementation of EfficientViT-l2-cls can be found
  [here](https://github.com/CVHub520/efficientvit/blob/main/LICENSE).

## References
* [EfficientViT: Multi-Scale Linear Attention for High-Resolution Dense Prediction](https://arxiv.org/abs/2205.14756)
* [Source Model Implementation](https://github.com/CVHub520/efficientvit)

## 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).