README.md

---
library_name: pytorch
license: other
tags:
- bu_auto
- real_time
- android
pipeline_tag: object-detection

---

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

# Yolo-R: Optimized for Qualcomm Devices

YoloR is a machine learning model that predicts bounding boxes and classes of objects in an image.

This is based on the implementation of Yolo-R found [here](https://github.com/WongKinYiu/yolor.git).
This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/yolor) 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
Due to licensing restrictions, we cannot distribute pre-exported model assets for this model.
Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/yolor) 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

See our repository for [Yolo-R on GitHub](https://github.com/qualcomm/ai-hub-models/blob/main/qai_hub_models/models/yolor) for usage instructions.


## Model Details

**Model Type:** Model_use_case.object_detection

**Model Stats:**
- Model checkpoint: yolor_p6
- Input resolution: 640x640
- Number of parameters: 4.68M
- Model size (float): 17.9 MB

## Performance Summary
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
|---|---|---|---|---|---|---
| Yolo-R | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 26.596 ms | 6 - 318 MB | NPU
| Yolo-R | ONNX | float | Snapdragon® X2 Elite | 26.422 ms | 75 - 75 MB | NPU
| Yolo-R | ONNX | float | Snapdragon® X Elite | 58.203 ms | 74 - 74 MB | NPU
| Yolo-R | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 41.88 ms | 6 - 355 MB | NPU
| Yolo-R | ONNX | float | Qualcomm® QCS8550 (Proxy) | 56.589 ms | 0 - 78 MB | NPU
| Yolo-R | ONNX | float | Qualcomm® QCS9075 | 52.802 ms | 5 - 12 MB | NPU
| Yolo-R | ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 37.084 ms | 2 - 224 MB | NPU
| Yolo-R | ONNX | w8a16 | Snapdragon® 8 Elite Gen 5 Mobile | 17.989 ms | 3 - 429 MB | NPU
| Yolo-R | ONNX | w8a16 | Snapdragon® X2 Elite | 18.352 ms | 41 - 41 MB | NPU
| Yolo-R | ONNX | w8a16 | Snapdragon® X Elite | 30.487 ms | 40 - 40 MB | NPU
| Yolo-R | ONNX | w8a16 | Snapdragon® 8 Gen 3 Mobile | 20.937 ms | 0 - 493 MB | NPU
| Yolo-R | ONNX | w8a16 | Qualcomm® QCS6490 | 2273.766 ms | 128 - 134 MB | CPU
| Yolo-R | ONNX | w8a16 | Qualcomm® QCS8550 (Proxy) | 29.392 ms | 0 - 50 MB | NPU
| Yolo-R | ONNX | w8a16 | Qualcomm® QCS9075 | 29.42 ms | 1 - 6 MB | NPU
| Yolo-R | ONNX | w8a16 | Qualcomm® QCM6690 | 1174.879 ms | 46 - 59 MB | CPU
| Yolo-R | ONNX | w8a16 | Snapdragon® 8 Elite For Galaxy Mobile | 17.09 ms | 1 - 355 MB | NPU
| Yolo-R | ONNX | w8a16 | Snapdragon® 7 Gen 4 Mobile | 1117.064 ms | 123 - 136 MB | CPU
| Yolo-R | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 14.061 ms | 5 - 297 MB | NPU
| Yolo-R | QNN_DLC | float | Snapdragon® X2 Elite | 15.354 ms | 5 - 5 MB | NPU
| Yolo-R | QNN_DLC | float | Snapdragon® X Elite | 29.35 ms | 5 - 5 MB | NPU
| Yolo-R | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 22.723 ms | 5 - 324 MB | NPU
| Yolo-R | QNN_DLC | float | Qualcomm® QCS8275 (Proxy) | 100.21 ms | 1 - 253 MB | NPU
| Yolo-R | QNN_DLC | float | Qualcomm® QCS8550 (Proxy) | 29.639 ms | 5 - 7 MB | NPU
| Yolo-R | QNN_DLC | float | Qualcomm® SA8775P | 36.096 ms | 1 - 259 MB | NPU
| Yolo-R | QNN_DLC | float | Qualcomm® QCS9075 | 37.41 ms | 5 - 11 MB | NPU
| Yolo-R | QNN_DLC | float | Qualcomm® QCS8450 (Proxy) | 49.716 ms | 4 - 376 MB | NPU
| Yolo-R | QNN_DLC | float | Qualcomm® SA7255P | 100.21 ms | 1 - 253 MB | NPU
| Yolo-R | QNN_DLC | float | Qualcomm® SA8295P | 44.036 ms | 0 - 306 MB | NPU
| Yolo-R | QNN_DLC | float | Snapdragon® 8 Elite For Galaxy Mobile | 18.644 ms | 0 - 230 MB | NPU
| Yolo-R | QNN_DLC | w8a16 | Snapdragon® 8 Elite Gen 5 Mobile | 7.303 ms | 2 - 303 MB | NPU
| Yolo-R | QNN_DLC | w8a16 | Snapdragon® X2 Elite | 8.38 ms | 2 - 2 MB | NPU
| Yolo-R | QNN_DLC | w8a16 | Snapdragon® X Elite | 18.975 ms | 2 - 2 MB | NPU
| Yolo-R | QNN_DLC | w8a16 | Snapdragon® 8 Gen 3 Mobile | 12.442 ms | 2 - 367 MB | NPU
| Yolo-R | QNN_DLC | w8a16 | Qualcomm® QCS6490 | 77.67 ms | 1 - 5 MB | NPU
| Yolo-R | QNN_DLC | w8a16 | Qualcomm® QCS8275 (Proxy) | 37.386 ms | 1 - 289 MB | NPU
| Yolo-R | QNN_DLC | w8a16 | Qualcomm® QCS8550 (Proxy) | 18.338 ms | 2 - 5 MB | NPU
| Yolo-R | QNN_DLC | w8a16 | Qualcomm® SA8775P | 18.348 ms | 0 - 289 MB | NPU
| Yolo-R | QNN_DLC | w8a16 | Qualcomm® QCS9075 | 20.527 ms | 0 - 4 MB | NPU
| Yolo-R | QNN_DLC | w8a16 | Qualcomm® QCM6690 | 237.504 ms | 2 - 400 MB | NPU
| Yolo-R | QNN_DLC | w8a16 | Qualcomm® QCS8450 (Proxy) | 24.359 ms | 2 - 369 MB | NPU
| Yolo-R | QNN_DLC | w8a16 | Qualcomm® SA7255P | 37.386 ms | 1 - 289 MB | NPU
| Yolo-R | QNN_DLC | w8a16 | Qualcomm® SA8295P | 23.585 ms | 2 - 293 MB | NPU
| Yolo-R | QNN_DLC | w8a16 | Snapdragon® 8 Elite For Galaxy Mobile | 9.82 ms | 2 - 301 MB | NPU
| Yolo-R | QNN_DLC | w8a16 | Snapdragon® 7 Gen 4 Mobile | 26.793 ms | 2 - 312 MB | NPU
| Yolo-R | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 26.339 ms | 1 - 343 MB | NPU
| Yolo-R | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 38.872 ms | 34 - 478 MB | NPU
| Yolo-R | TFLITE | float | Qualcomm® QCS8275 (Proxy) | 137.805 ms | 1 - 328 MB | NPU
| Yolo-R | TFLITE | float | Qualcomm® QCS8550 (Proxy) | 59.03 ms | 0 - 3 MB | NPU
| Yolo-R | TFLITE | float | Qualcomm® SA8775P | 66.376 ms | 0 - 308 MB | NPU
| Yolo-R | TFLITE | float | Qualcomm® QCS9075 | 50.721 ms | 0 - 86 MB | NPU
| Yolo-R | TFLITE | float | Qualcomm® QCS8450 (Proxy) | 89.294 ms | 0 - 467 MB | NPU
| Yolo-R | TFLITE | float | Qualcomm® SA7255P | 137.805 ms | 1 - 328 MB | NPU
| Yolo-R | TFLITE | float | Qualcomm® SA8295P | 76.681 ms | 1 - 355 MB | NPU
| Yolo-R | TFLITE | float | Snapdragon® 8 Elite For Galaxy Mobile | 32.082 ms | 1 - 287 MB | NPU

## License
* The license for the original implementation of Yolo-R can be found
  [here](https://github.com/WongKinYiu/yolor/blob/main/LICENSE).

## References
* [You Only Learn One Representation: Unified Network for Multiple Tasks](https://arxiv.org/abs/2105.04206)
* [Source Model Implementation](https://github.com/WongKinYiu/yolor.git)

## 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).