Hugging Face's logo

qualcomm
/
VIT

Image Classification
PyTorch
LiteRT
backbone
android
Model card Files
xet
Community

VIT: Optimized for Mobile Deployment

Imagenet classifier and general purpose backbone

VIT 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 model is an implementation of VIT found here.

This repository provides scripts to run VIT on Qualcomm® devices. More details on model performance across various devices, can be found here.

Model Details

  • Model Type: Model_use_case.image_classification
  • Model Stats:
    • Model checkpoint: Imagenet
    • Input resolution: 224x224
    • Number of parameters: 86.6M
    • Model size (float): 330 MB
    • Model size (w8a16): 86.2 MB
    • Model size (w8a8): 83.2 MB
Model Precision Device Chipset Target Runtime Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit Target Model
VIT float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) TFLITE 42.829 ms 0 - 306 MB NPU VIT.tflite
VIT float QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) QNN_DLC 44.895 ms 0 - 328 MB NPU VIT.dlc
VIT float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) TFLITE 17.158 ms 0 - 301 MB NPU VIT.tflite
VIT float QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) QNN_DLC 21.662 ms 0 - 317 MB NPU VIT.dlc
VIT float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) TFLITE 12.547 ms 0 - 16 MB NPU VIT.tflite
VIT float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN_DLC 14.024 ms 0 - 27 MB NPU VIT.dlc
VIT float QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) ONNX 13.399 ms 0 - 219 MB NPU VIT.onnx.zip
VIT float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) TFLITE 15.223 ms 0 - 306 MB NPU VIT.tflite
VIT float QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN_DLC 16.5 ms 0 - 332 MB NPU VIT.dlc
VIT float SA7255P ADP Qualcomm® SA7255P TFLITE 42.829 ms 0 - 306 MB NPU VIT.tflite
VIT float SA7255P ADP Qualcomm® SA7255P QNN_DLC 44.895 ms 0 - 328 MB NPU VIT.dlc
VIT float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) TFLITE 12.544 ms 0 - 14 MB NPU VIT.tflite
VIT float SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN_DLC 13.998 ms 1 - 28 MB NPU VIT.dlc
VIT float SA8295P ADP Qualcomm® SA8295P TFLITE 19.329 ms 0 - 290 MB NPU VIT.tflite
VIT float SA8295P ADP Qualcomm® SA8295P QNN_DLC 19.738 ms 1 - 324 MB NPU VIT.dlc
VIT float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) TFLITE 12.539 ms 0 - 15 MB NPU VIT.tflite
VIT float SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN_DLC 14.048 ms 1 - 26 MB NPU VIT.dlc
VIT float SA8775P ADP Qualcomm® SA8775P TFLITE 15.223 ms 0 - 306 MB NPU VIT.tflite
VIT float SA8775P ADP Qualcomm® SA8775P QNN_DLC 16.5 ms 0 - 332 MB NPU VIT.dlc
VIT float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile TFLITE 8.453 ms 0 - 312 MB NPU VIT.tflite
VIT float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN_DLC 9.56 ms 1 - 333 MB NPU VIT.dlc
VIT float Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile ONNX 9.118 ms 0 - 331 MB NPU VIT.onnx.zip
VIT float Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile TFLITE 6.18 ms 0 - 310 MB NPU VIT.tflite
VIT float Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile QNN_DLC 6.907 ms 1 - 322 MB NPU VIT.dlc
VIT float Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile ONNX 6.329 ms 0 - 325 MB NPU VIT.onnx.zip
VIT float Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile TFLITE 5.131 ms 0 - 310 MB NPU VIT.tflite
VIT float Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile QNN_DLC 5.348 ms 1 - 318 MB NPU VIT.dlc
VIT float Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile ONNX 4.866 ms 1 - 316 MB NPU VIT.onnx.zip
VIT float Snapdragon X Elite CRD Snapdragon® X Elite QNN_DLC 14.604 ms 1006 - 1006 MB NPU VIT.dlc
VIT float Snapdragon X Elite CRD Snapdragon® X Elite ONNX 13.833 ms 171 - 171 MB NPU VIT.onnx.zip
VIT w8a16 QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) QNN_DLC 65.468 ms 0 - 197 MB NPU VIT.dlc
VIT w8a16 QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) QNN_DLC 50.821 ms 0 - 224 MB NPU VIT.dlc
VIT w8a16 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN_DLC 25.829 ms 0 - 48 MB NPU VIT.dlc
VIT w8a16 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) ONNX 175.042 ms 494 - 724 MB NPU VIT.onnx.zip
VIT w8a16 QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN_DLC 22.996 ms 0 - 196 MB NPU VIT.dlc
VIT w8a16 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) QNN_DLC 196.707 ms 0 - 1572 MB NPU VIT.dlc
VIT w8a16 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) ONNX 528.728 ms 50 - 68 MB CPU VIT.onnx.zip
VIT w8a16 RB5 (Proxy) Qualcomm® QCS8250 (Proxy) ONNX 554.855 ms 47 - 123 MB CPU VIT.onnx.zip
VIT w8a16 SA7255P ADP Qualcomm® SA7255P QNN_DLC 65.468 ms 0 - 197 MB NPU VIT.dlc
VIT w8a16 SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN_DLC 25.759 ms 0 - 48 MB NPU VIT.dlc
VIT w8a16 SA8295P ADP Qualcomm® SA8295P QNN_DLC 36.873 ms 0 - 216 MB NPU VIT.dlc
VIT w8a16 SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN_DLC 25.626 ms 0 - 48 MB NPU VIT.dlc
VIT w8a16 SA8775P ADP Qualcomm® SA8775P QNN_DLC 22.996 ms 0 - 196 MB NPU VIT.dlc
VIT w8a16 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN_DLC 19.505 ms 140 - 342 MB NPU VIT.dlc
VIT w8a16 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile ONNX 132.975 ms 671 - 857 MB NPU VIT.onnx.zip
VIT w8a16 Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile QNN_DLC 15.548 ms 0 - 195 MB NPU VIT.dlc
VIT w8a16 Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile ONNX 102.313 ms 680 - 848 MB NPU VIT.onnx.zip
VIT w8a16 Snapdragon 7 Gen 5 QRD Snapdragon® 7 Gen 5 Mobile QNN_DLC 41.426 ms 0 - 277 MB NPU VIT.dlc
VIT w8a16 Snapdragon 7 Gen 5 QRD Snapdragon® 7 Gen 5 Mobile ONNX 603.447 ms 72 - 90 MB CPU VIT.onnx.zip
VIT w8a16 Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile QNN_DLC 12.446 ms 0 - 207 MB NPU VIT.dlc
VIT w8a16 Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile ONNX 119.229 ms 674 - 851 MB NPU VIT.onnx.zip
VIT w8a16 Snapdragon X Elite CRD Snapdragon® X Elite QNN_DLC 25.967 ms 317 - 317 MB NPU VIT.dlc
VIT w8a16 Snapdragon X Elite CRD Snapdragon® X Elite ONNX 156.229 ms 925 - 925 MB NPU VIT.onnx.zip
VIT w8a8 QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) TFLITE 15.845 ms 0 - 47 MB NPU VIT.tflite
VIT w8a8 QCS8450 (Proxy) Qualcomm® QCS8450 (Proxy) TFLITE 8.316 ms 0 - 57 MB NPU VIT.tflite
VIT w8a8 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) TFLITE 7.612 ms 0 - 19 MB NPU VIT.tflite
VIT w8a8 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) ONNX 193.957 ms 662 - 883 MB NPU VIT.onnx.zip
VIT w8a8 QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) TFLITE 7.99 ms 0 - 47 MB NPU VIT.tflite
VIT w8a8 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) TFLITE 60.311 ms 2 - 45 MB NPU VIT.tflite
VIT w8a8 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) ONNX 437.781 ms 31 - 48 MB CPU VIT.onnx.zip
VIT w8a8 RB5 (Proxy) Qualcomm® QCS8250 (Proxy) ONNX 422.631 ms 29 - 97 MB CPU VIT.onnx.zip
VIT w8a8 SA7255P ADP Qualcomm® SA7255P TFLITE 15.845 ms 0 - 47 MB NPU VIT.tflite
VIT w8a8 SA8255 (Proxy) Qualcomm® SA8255P (Proxy) TFLITE 7.607 ms 0 - 62 MB NPU VIT.tflite
VIT w8a8 SA8295P ADP Qualcomm® SA8295P TFLITE 9.95 ms 0 - 50 MB NPU VIT.tflite
VIT w8a8 SA8650 (Proxy) Qualcomm® SA8650P (Proxy) TFLITE 7.62 ms 0 - 22 MB NPU VIT.tflite
VIT w8a8 SA8775P ADP Qualcomm® SA8775P TFLITE 7.99 ms 0 - 47 MB NPU VIT.tflite
VIT w8a8 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile TFLITE 5.386 ms 0 - 56 MB NPU VIT.tflite
VIT w8a8 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile ONNX 132.061 ms 671 - 851 MB NPU VIT.onnx.zip
VIT w8a8 Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile TFLITE 4.23 ms 0 - 56 MB NPU VIT.tflite
VIT w8a8 Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile ONNX 97.949 ms 657 - 804 MB NPU VIT.onnx.zip
VIT w8a8 Snapdragon 7 Gen 5 QRD Snapdragon® 7 Gen 5 Mobile TFLITE 22.854 ms 0 - 30 MB NPU VIT.tflite
VIT w8a8 Snapdragon 7 Gen 5 QRD Snapdragon® 7 Gen 5 Mobile ONNX 492.001 ms 37 - 51 MB CPU VIT.onnx.zip
VIT w8a8 Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile TFLITE 3.4 ms 0 - 58 MB NPU VIT.tflite
VIT w8a8 Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile ONNX 114.491 ms 675 - 889 MB NPU VIT.onnx.zip
VIT w8a8 Snapdragon X Elite CRD Snapdragon® X Elite ONNX 158.207 ms 926 - 926 MB NPU VIT.onnx.zip
VIT w8a8_mixed_int16 QCS8275 (Proxy) Qualcomm® QCS8275 (Proxy) QNN_DLC 53.711 ms 0 - 244 MB NPU VIT.dlc
VIT w8a8_mixed_int16 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) QNN_DLC 21.253 ms 0 - 42 MB NPU VIT.dlc
VIT w8a8_mixed_int16 QCS8550 (Proxy) Qualcomm® QCS8550 (Proxy) ONNX 195.052 ms 420 - 665 MB NPU VIT.onnx.zip
VIT w8a8_mixed_int16 QCS9075 (Proxy) Qualcomm® QCS9075 (Proxy) QNN_DLC 20.116 ms 0 - 230 MB NPU VIT.dlc
VIT w8a8_mixed_int16 RB3 Gen 2 (Proxy) Qualcomm® QCS6490 (Proxy) ONNX 457.445 ms 52 - 70 MB CPU VIT.onnx.zip
VIT w8a8_mixed_int16 RB5 (Proxy) Qualcomm® QCS8250 (Proxy) ONNX 472.605 ms 37 - 114 MB CPU VIT.onnx.zip
VIT w8a8_mixed_int16 SA7255P ADP Qualcomm® SA7255P QNN_DLC 53.711 ms 0 - 244 MB NPU VIT.dlc
VIT w8a8_mixed_int16 SA8255 (Proxy) Qualcomm® SA8255P (Proxy) QNN_DLC 21.192 ms 0 - 41 MB NPU VIT.dlc
VIT w8a8_mixed_int16 SA8650 (Proxy) Qualcomm® SA8650P (Proxy) QNN_DLC 21.047 ms 0 - 42 MB NPU VIT.dlc
VIT w8a8_mixed_int16 SA8775P ADP Qualcomm® SA8775P QNN_DLC 20.116 ms 0 - 230 MB NPU VIT.dlc
VIT w8a8_mixed_int16 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile QNN_DLC 16.416 ms 0 - 251 MB NPU VIT.dlc
VIT w8a8_mixed_int16 Samsung Galaxy S24 Snapdragon® 8 Gen 3 Mobile ONNX 160.386 ms 542 - 752 MB NPU VIT.onnx.zip
VIT w8a8_mixed_int16 Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile QNN_DLC 13.097 ms 0 - 266 MB NPU VIT.dlc
VIT w8a8_mixed_int16 Samsung Galaxy S25 Snapdragon® 8 Elite For Galaxy Mobile ONNX 118.291 ms 553 - 733 MB NPU VIT.onnx.zip
VIT w8a8_mixed_int16 Snapdragon 7 Gen 5 QRD Snapdragon® 7 Gen 5 Mobile QNN_DLC 33.635 ms 1 - 241 MB NPU VIT.dlc
VIT w8a8_mixed_int16 Snapdragon 7 Gen 5 QRD Snapdragon® 7 Gen 5 Mobile ONNX 515.589 ms 85 - 104 MB CPU VIT.onnx.zip
VIT w8a8_mixed_int16 Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile QNN_DLC 9.838 ms 0 - 197 MB NPU VIT.dlc
VIT w8a8_mixed_int16 Snapdragon 8 Elite Gen 5 QRD Snapdragon® 8 Elite Gen5 Mobile ONNX 142.396 ms 539 - 759 MB NPU VIT.onnx.zip
VIT w8a8_mixed_int16 Snapdragon X Elite CRD Snapdragon® X Elite QNN_DLC 20.939 ms 350 - 350 MB NPU VIT.dlc
VIT w8a8_mixed_int16 Snapdragon X Elite CRD Snapdragon® X Elite ONNX 196.143 ms 926 - 926 MB NPU VIT.onnx.zip

Installation

Install the package via pip:

pip install qai-hub-models

Configure Qualcomm® AI Hub to run this model on a cloud-hosted device

Sign-in to Qualcomm® AI Hub with your Qualcomm® ID. Once signed in navigate to Account -> Settings -> API Token.

With this API token, you can configure your client to run models on the cloud hosted devices.

qai-hub configure --api_token API_TOKEN

Navigate to docs for more information.

Demo off target

The package contains a simple end-to-end demo that downloads pre-trained weights and runs this model on a sample input.

python -m qai_hub_models.models.vit.demo

The above demo runs a reference implementation of pre-processing, model inference, and post processing.

NOTE: If you want running in a Jupyter Notebook or Google Colab like environment, please add the following to your cell (instead of the above).

%run -m qai_hub_models.models.vit.demo

Run model on a cloud-hosted device

In addition to the demo, you can also run the model on a cloud-hosted Qualcomm® device. This script does the following:

  • Performance check on-device on a cloud-hosted device
  • Downloads compiled assets that can be deployed on-device for Android.
  • Accuracy check between PyTorch and on-device outputs.
python -m qai_hub_models.models.vit.export

How does this work?

This export script leverages Qualcomm® AI Hub to optimize, validate, and deploy this model on-device. Lets go through each step below in detail:

Step 1: Compile model for on-device deployment

To compile a PyTorch model for on-device deployment, we first trace the model in memory using the jit.trace and then call the submit_compile_job API.

import torch

import qai_hub as hub
from qai_hub_models.models.vit import Model

# Load the model
torch_model = Model.from_pretrained()

# Device
device = hub.Device("Samsung Galaxy S25")

# Trace model
input_shape = torch_model.get_input_spec()
sample_inputs = torch_model.sample_inputs()

pt_model = torch.jit.trace(torch_model, [torch.tensor(data[0]) for _, data in sample_inputs.items()])

# Compile model on a specific device
compile_job = hub.submit_compile_job(
    model=pt_model,
    device=device,
    input_specs=torch_model.get_input_spec(),
)

# Get target model to run on-device
target_model = compile_job.get_target_model()

Step 2: Performance profiling on cloud-hosted device

After compiling models from step 1. Models can be profiled model on-device using the target_model. Note that this scripts runs the model on a device automatically provisioned in the cloud. Once the job is submitted, you can navigate to a provided job URL to view a variety of on-device performance metrics.

profile_job = hub.submit_profile_job(
    model=target_model,
    device=device,
)

Step 3: Verify on-device accuracy

To verify the accuracy of the model on-device, you can run on-device inference on sample input data on the same cloud hosted device.

input_data = torch_model.sample_inputs()
inference_job = hub.submit_inference_job(
    model=target_model,
    device=device,
    inputs=input_data,
)
    on_device_output = inference_job.download_output_data()

With the output of the model, you can compute like PSNR, relative errors or spot check the output with expected output.

Note: This on-device profiling and inference requires access to Qualcomm® AI Hub. Sign up for access.

Run demo on a cloud-hosted device

You can also run the demo on-device.

python -m qai_hub_models.models.vit.demo --eval-mode on-device

NOTE: If you want running in a Jupyter Notebook or Google Colab like environment, please add the following to your cell (instead of the above).

%run -m qai_hub_models.models.vit.demo -- --eval-mode on-device

Deploying compiled model to Android

The models can be deployed using multiple runtimes:

  • TensorFlow Lite (.tflite export): This tutorial provides a guide to deploy the .tflite model in an Android application.

  • QNN (.so export ): This sample app provides instructions on how to use the .so shared library in an Android application.

View on Qualcomm® AI Hub

Get more details on VIT's performance across various devices here. Explore all available models on Qualcomm® AI Hub

License

  • The license for the original implementation of VIT can be found here.
  • The license for the compiled assets for on-device deployment can be found here

References

Community

Downloads last month
298
Inference Providers NEW
Image Classification
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support