[Feature Request]: Request for NumPy 2.1.2 support in OpenVINO 2024.4.0

[avi333learn]

Request Description

I am encountering an issue where OpenVINO 2024.4.0 is incompatible with NumPy 2.1.2. The current release of OpenVINO requires NumPy versions <2.1.0, but my project requires NumPy 2.1.2 due to dependencies from other components (specifically for setting up FaceFusion). This results in conflicts when installing OpenVINO alongside other packages.

ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
openvino 2024.4.0 requires numpy<2.1.0,>=1.16.6, but you have numpy 2.1.2 which is incompatible.

Request:
I would like to request that OpenVINO adds support for NumPy 2.1.2 in future releases, as NumPy 2.x is becoming more common in various projects. Supporting the latest versions would help avoid these kinds of conflicts and make OpenVINO more compatible with modern workflows.

Steps to Reproduce:

Install OpenVINO 2024.4.0 with NumPy 2.1.2 already installed in the environment.
Run any code that attempts to use OpenVINO, and the dependency conflict occurs.

System Information:

OpenVINO Version: 2024.4.0
NumPy Version: 2.1.2
Python Version: Python 3.10.6
OS: Windows 11
Processor: 13th Gen Intel(R) Core(TM) i7-1355U
Graphics: Intel Iris Xe Graphics

Impact:
This issue is blocking the installation of OpenVINO alongside other components in the FaceFusion setup, where NumPy 2.1.2 is required. Downgrading NumPy is not an option due to dependency conflicts with other libraries.

Feature Use Case

Supporting NumPy 2.1.2 in OpenVINO would greatly improve compatibility with modern machine learning frameworks and tools that rely on the latest versions of NumPy. In my case, I am setting up FaceFusion, which requires NumPy 2.1.2 due to dependencies from other components. Downgrading NumPy introduces compatibility issues with those components, making it difficult to work in a unified environment.

By adding support for NumPy 2.1.2, OpenVINO will be more flexible and easier to integrate into a variety of machine learning workflows where the latest versions of dependencies are crucial for performance and functionality. This will reduce friction for users who rely on multiple frameworks and help avoid the need for isolating environments or downgrading packages, leading to smoother setups and better performance.

Issue submission checklist

• The feature request or improvement must be related to OpenVINO