# Olive

[Olive (ONNX LIVE)](https://microsoft.github.io/Olive/index.html) is a state‑of‑the‑art model optimization toolkit and command‑line interface (CLI) designed to help developers produce high‑performance, production‑ready ONNX Runtime models optimized for acceleration on Snapdragon® chipsets.
Olive enables models to efficiently leverage the available Snapdragon AI hardware accelerators on the deployment platform, including the NPU, GPU, and CPU.
Olive executes a workflow defined as an ordered sequence of model optimization passes.

Model optimization in Olive uses a workflow defined as an ordered sequence of optimization passes. Each pass encapsulates a specific model transformation, such as model compression, graph capture, quantization, or graph‑level optimization.

This tutorial presents a practical example demonstrating the quantization of the MobileNet model. For information about additional optimization capabilities, refer to [Olive Passes](https://microsoft.github.io/Olive/reference/pass.html).

## Setup

This tutorial requires the following setup which takes approximately five minutes. For more information about the setup, see [olive_setup.ps1](https://raw.githubusercontent.com/quic/wos-ai/refs/heads/main/Scripts/olive_setup.ps1).

1. Open PowerShell in administrator mode and run the following commands to install the dependencies. You must run PowerShell in administrator mode to grant the elevated permissions required to automatically install the necessary applications and execute the PowerShell scripts.
2. Set `$DIR_PATH`. By default it is `C:\\WoS_AI` although you can change to your desired path.

$DIR_PATH  = "C:\WoS_AI"
        Copy to clipboard
3. Run the following command to download the setup script.

if (!(Test-Path $DIR_PATH\Downloads\Setup_Scripts)) {mkdir $DIR_PATH\Downloads\Setup_Scripts}
        Invoke-WebRequest -O olive_setup.ps1 https://raw.githubusercontent.com/quic/wos-ai/refs/heads/main/Scripts/olive_setup.ps1
        Move-Item -Path ".\olive_setup.ps1" -Destination $DIR_PATH\Downloads\Setup_Scripts -Force
        Copy to clipboard
4. If running scripts is disabled on the system, run the following command and enter ‘A’(Yes for all).

Set-ExecutionPolicy RemoteSigned
        Copy to clipboard
5. Run the following commands to install the dependencies for Olive.

cd $DIR_PATH
        Copy to clipboard

powershell -command "&{. .\Downloads\Setup_Scripts\olive_setup.ps1; OLIVE_Setup -rootDirPath $DIR_PATH}"
        Copy to clipboard

    The command downloads and installs the following dependencies:

> 
> 
> - Python amd64 version 3.12.6
>     - Olive artifacts include `olive_download_files.py` for model and data acquisition, `olive_user_script.py` for defining the dataloader within the Olive recipe, and Olive recipe for `sd-legacy-stable-diffusion-v1-5`.
>     - Visual Studio Redistributable Version 14.44.35208
>     - Creates a virtual environment (SDX\_OLIVE\_ENV) with the necessary libraries for the QNN EP.
>     - onnxruntime-qnn, olive-ai

Note

Avoid installing any other ONNX Runtime variant in this environment. The `setup.ps1` script installs the necessary onnxruntime-qnn. Installing a different variant may result in fetching the incorrect ONNX Runtime variant and could cause issues.
6. When setup is complete, the following folder structure is in `$DIR_PATH`:

> 
> 
> ![../../_images/olive_folder_structure_sd.png](data:image/png;base64,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)

## Olive tutorial - 1

This tutorial uses Olive for quantizing the traditional classification model (mobilenetv2-12).

1. Open a PowerShell and run the following commands to set `DIR_PATH` and activate the Python virtual environment. Ensure that `DIR_PATH` is the same as the one used during the setup.

${DIR_PATH} = "C:\WoS_AI"
        Copy to clipboard

powershell -NoExit -command "&{cd $DIR_PATH; . .\Downloads\Setup_Scripts\olive_setup.ps1; Activate_OLIVE_VENV -rootDirPath $DIR_PATH}"
        Copy to clipboard

Note

The following example is for the MobileNet model. If you need to try the same example for a different model, use the existing Python virtual environment (SDX\_OLIVE\_ENV) and handle the dependencies by updating the Olive recipe.
2. Run the following command to download the mobilenetv2-12.onnx model and data for quantization.

python .\olive_download_files.py
        Copy to clipboard

![../../_images/olive_data.png](data:image/png;base64,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)
3. The previous command creates a `model` folder containing the mobilenetv2-12.onnx model and a `Data` folder containing the dataset for the model’s optimization and inferencing.
4. Run the following to create the `olive_mobilenet_qnn_ep.json` file in the working directory.

notepad.exe olive_mobilenet_qnn_ep.json
        Copy to clipboard
5. Copy the following into the `olive_mobilenet_qnn_ep.json` file and save.

{
           "input_model": { "type": "OnnxModel", "model_path": "models/mobilenetv2-12.onnx" },
           "systems": {
              "local_system": {
                    "type": "LocalSystem",
                    "accelerators": [ { "execution_providers": [ "QNNExecutionProvider" ] } ]
              }
           },
           "data_configs": [
              {
                    "name": "mobilenet_data_config",
                    "user_script": "olive_user_script.py",
                    "load_dataset_config": { "type": "mobilenet_dataset", "data_dir": "data/eval" },
                    "post_process_data_config": { "type": "mobilenet_post_process" },
                    "dataloader_config": { "batch_size": 1 }
              }
           ],
           "evaluators": {
              "common_evaluator": {
                    "metrics": [
                       {
                          "name": "accuracy",
                          "type": "accuracy",
                          "data_config": "mobilenet_data_config",
                          "sub_types": [
                                {
                                   "name": "accuracy_score",
                                   "priority": 1,
                                   "metric_config": { "task": "multiclass", "num_classes": 1000 }
                                }
                          ]
                       },
                       {
                          "name": "latency",
                          "type": "latency",
                          "data_config": "mobilenet_data_config",
                          "sub_types": [ { "name": "avg", "priority": 2 } ]
                       }
                    ]
              }
           },
           "passes": {
              "dynamic_shape_to_fixed": { "type": "DynamicToFixedShape", "dim_param": [ "batch_size" ], "dim_value": [ 1 ] },
              "qnn_preprocess": { "type": "QNNPreprocess" },
              "quantization": {
                    "type": "OnnxStaticQuantization",
                    "data_config": "mobilenet_data_config",
                    "activation_type": "uint16",
                    "precision": "uint8"
              }
           },
           "host": "local_system",
           "target": "local_system",
           "evaluator": "common_evaluator",
           "evaluate_input_model": false,
           "cache_dir": "cache",
           "output_dir": "models/mobilenet_qnn_ep"
        }
        Copy to clipboard
6. The script in the previous step defines the quantization parameters and configures the optimization pipeline, serving as an essential input for Olive to perform model optimization.
7. Run the following command to quantize the MobileNet model.

> 
> 
> olive run --config .\olive_mobilenet_qnn_ep.json
>         Copy to clipboard
> 
> ![../../_images/olive_run.png](data:image/png;base64,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)
8. The previous command generates a quantized MobileNet ONNX model at the `models\mobilenet_qnn_ep\model.onnx` location.

Note

You can use the optimized mobilenetv2-12 model generated in the previous steps for inferencing using [ORT-QNN](https://docs.qualcomm.com/doc/80-62010-1/topic/ort-qnn-ep.html#ort-qnn-ep) or [WindowsML](https://docs.qualcomm.com/doc/80-62010-1/topic/winml-ep.html#winml-ep).

## Olive tutorial - 2

This tutorial uses the Olive recipe for quantizing and optimization of Stable Diffusion V-1.5.

1. Open a PowerShell and run the following commands to set `DIR_PATH` and activate the Python virtual environment. Ensure that `DIR_PATH` is the same as the one used during the setup.

${DIR_PATH} = "C:\WoS_AI"
        Copy to clipboard

powershell -NoExit -command "&{cd $DIR_PATH; . .\Downloads\Setup_Scripts\olive_setup.ps1; Activate_OLIVE_VENV -rootDirPath $DIR_PATH; cd olive-recipes\sd-legacy-stable-diffusion-v1-5\olive}"
        Copy to clipboard

Note

The following example is for the Stable Diffusion V-1.5 model. If you need to try the same example for a different model, use the existing Python virtual environment (SDX\_OLIVE\_ENV) and handle the dependencies by updating the Olive recipe.
2. Run the following command to download the stable-diffusion-v1-5 unoptimized onnx model (takes ~10mins).

python stable_diffusion.py --model_id sd-legacy/stable-diffusion-v1-5 --provider cpu --format qdq --optimize --only_conversion
        Copy to clipboard
3. Run the following command to generate data for quantization (takes ~3hrs).

python .\evaluation.py --save_data --model_id sd-legacy/stable-diffusion-v1-5 --num_inference_steps 25 --seed 0 --num_data 100 --guidance_scale 7.5
        Copy to clipboard
4. Run the following to optimize the ONNX models for performance improvements (takes ~1hr).

python stable_diffusion.py --model_id sd-legacy/stable-diffusion-v1-5 --provider qnn --format qdq --optimize
        Copy to clipboard

Note

vae\_decoder and unet are per-channel quantization and text\_encoder runs in fp16 precision.
5. Run the following command to try inference generating image.

python stable_diffusion.py --model_id sd-legacy/stable-diffusion-v1-5 --provider qnn --format qdq --guidance_scale 7.5 --seed 0 --num_inference_steps 25 --prompt "A baby is laying down with a teddy bear"
        Copy to clipboard

Note

To test and evaluate the generated optimized model, run the following command (takes ~3hrs).

python .\evaluation.py --model_id sd-legacy/stable-diffusion-v1-5 --num_inference_steps 25 --seed 0 --num_data 100 --guidance_scale 7.5 --provider QNNExecutionProvider --model_dir optimized-qnn_qdq
    Copy to clipboard

## More details

- Learn more about [Olive](https://microsoft.github.io/Olive/index.html).
- For more available optimization options, see [Olive Passes](https://microsoft.github.io/Olive/reference/pass.html).
- Several example models are provided as part of the tutorial. For more information, refer to [Olive Recipes](https://github.com/microsoft/olive-recipes).
- For details on creating optimized ONNX models using Olive, see [Democratizing AI Model optimization with the new Olive CLI](https://onnxruntime.ai/blogs/olive-cli).

Last Published: Jun 16, 2026

[Previous Topic
Windows AI Foundry](https://docs.qualcomm.com/bundle/publicresource/80-62010-1/topics/WinML.md) [Next Topic
Runtimes](https://docs.qualcomm.com/bundle/publicresource/80-62010-1/topics/runtimes.md)