torch-ort cannot be installed on windows: onnxruntime-training not found

[alpatots]

When running pip install torch-ort in a conda environment on Windows, I get the following error:

ERROR: Could not find a version that satisfies the requirement onnxruntime-training (from versions: none) ERROR: No matching distribution found for onnxruntime-training

However, if I run the same command in a conda environment in WSL, it works just fine. Other people could repro. Seems to be a Windows issue.

[baijumeswani]

Yes, as of now, onnxruntime-training (ORTModule) is not available on Windows. We do plan to change this in the near future. But I don't have a timeline to share with you yet.

[askhade]

@alpatots : Can you add some details for your scenario? Are you looking for Windows packages with Nvidia GPU support?

[alpatots]

@askhade I don't think I have more details than that... Our package takes a dependency on torch-ort, and we discovered it made our package un-installable on Windows. It does not affect runtime because we run jobs on linux anyways.

[dlod-openvino]

@askhade @baijumeswani There are many machine vision application based on Windows, currently the machine vision application will add Pytorch model (eg, Mask RCNN)for defects detection. We retain the Pytorch model on Windows and want to deploy it by torch-ort on Intel ARC GPU on Windows as well. We meet the same problem when install the torch-ort-infer[openvino] on Windows

Looking forward to add Windows support ASAP! Thank you very much in advance!!

[dlod-openvino]

Add error screenshot when execute:

from torch_ort import ORTInferenceModule, OpenVINOProviderOptions