Run the Phi-3 vision model with the ONNX Runtime generate() API

The Phi-3 vision model is a small, but powerful multi modal model that allows you to use both image and text to output text. It is used in scenarios such as describing the content of images in detail.

The Phi-3 vision model is supported by versions of onnxruntime-genai 0.3.0-rc2 and later.

You can download the models here:

Support for DirectML is coming soon!

Setup

  1. Install the git large file system extension

    HuggingFace uses git for version control. To download the ONNX models you need git lfs to be installed, if you do not already have it.

    • Windows: winget install -e --id GitHub.GitLFS (If you don’t have winget, download and run the exe from the official source)
    • Linux: apt-get install git-lfs
    • MacOS: brew install git-lfs

    Then run git lfs install

  2. Install the HuggingFace CLI

    pip install huggingface-hub[cli]

Choose your platform

If you have an NVIDIA GPU, that will give the best performance right now.

The models will also run on CPU, but they will be slower.

Support for Windows machines with GPUs other than NVIDIA is coming soon!

Note: Only one package and model is required based on your hardware. That is, only execute the steps for one of the following sections

Run with NVIDIA CUDA

  1. Download the model

    huggingface-cli download microsoft/Phi-3-vision-128k-instruct-onnx-cuda --include cuda-int4-rtn-block-32/* --local-dir .

    This command downloads the model into a folder called cuda-int4-rtn-block-32.

  2. Setup your CUDA environment

    Install the CUDA toolkit.

    Ensure that the CUDA_PATH environment variable is set to the location of your CUDA installation.

  3. Install the generate() API

    • CUDA 11
    pip install numpy
pip install --pre onnxruntime-genai-cuda --index-url=https://aiinfra.pkgs.visualstudio.com/PublicPackages/_packaging/onnxruntime-genai/pypi/simple/
    • CUDA 12
    pip install numpy
pip install onnxruntime-genai-cuda --pre --index-url=https://aiinfra.pkgs.visualstudio.com/PublicPackages/_packaging/onnxruntime-cuda-12/pypi/simple/

  4. Run the model

    Run the model with phi3v.py.

    curl https://raw.githubusercontent.com/microsoft/onnxruntime-genai/main/examples/python/phi3v.py -o phi3v.py
python phi3v.py -m cuda-int4-rtn-block-32

    Enter the path to an image file and a prompt, and the model uses the image and prompt to give you an answer.

    For example: Describe the image

    coffee

    The image shows a cup of coffee with a latte art design on top. The coffee is a light brown color,
and the art is white with a leaf-like pattern. The cup is white and has a handle on one side.</s>

Run on CPU

  1. Download the model

    huggingface-cli download microsoft/Phi-3-vision-128k-instruct-onnx-cpu --include cpu-int4-rtn-block-32-acc-level-4/* --local-dir .

    This command downloads the model into a folder called cpu-int4-rtn-block-32-acc-level-4

  2. Install the generate() API for CPU

    pip install numpy
pip install --pre onnxruntime-genai

  3. Run the model

    Run the model with phi3v.py.

    curl https://raw.githubusercontent.com/microsoft/onnxruntime-genai/main/examples/python/phi3v.py -o phi3v.py
python phi3v.py -m cpu-int4-rtn-block-32-acc-level-4

    Enter the path to an image file and a prompt, and the model uses the image and prompt to give you an answer.

    For example: Convert this image to markdown format

    Excel table with cookie sales figures

    | Product             | Qtr 1      | Qtr 2      | Grand Total |
|---------------------|------------|------------|-------------|
| Chocolade           | $744.60    | $162.56    | $907.16     |
| Gummibarchen        | $5,079.60  | $1,249.20  | $6,328.80   |
| Scottish Longbreads | $1,267.50  | $1,062.50  | $2,330.00   |
| Sir Rodney's Scones | $1,418.00  | $756.00    | $2,174.00   |
| Tarte au sucre      | $4,728.00  | $4,547.92  | $9,275.92   |
| Chocolate Biscuits  | $943.89    | $349.60    | $1,293.49   |
| Total               | $14,181.59 | $8,127.78  | $22,309.37  |

The table lists various products along with their sales figures for Qtr 1, Qtr 2, and the Grand Total.
The products include Chocolade, Gummibarchen, Scottish Longbreads, Sir Rodney's Scones, Tarte au sucre,
and Chocolate Biscuits. The Grand Total column sums up the sales for each product across the two quarters.</s>