[Nico Möller (nmoeller)]

Motivation and Context

Fixes : #13001

Description

Package version 0.9.0 introduced very helpful methods like applying chat templates automatically from the tokenizer and also support for multi audio & image files.

This PR adds the following functionalities:

• Add Inference with multiple Images
• Add Inference with multiple Audios
• Chat Templates for non Multimodal Models are read via onnx

Samples

Text Sample with ONNX:

Image Sample with ONNX :

Contribution Checklist

• The code builds clean without any errors or warnings
• The PR follows the SK Contribution Guidelines and the pre-submission formatting script raises no violations
• All unit tests pass, and I have added new tests where possible
• I didn't break anyone 😄

[Evan Mattson (moonbox3)]

Python Test Coverage Report •

File	Stmts	Miss	Cover	Missing
connectors/ai/onnx
utils.py	85	3	96%	55–56, 167
connectors/ai/onnx/services
onnx_gen_ai_chat_completion.py	90	13	85%	70–71, 85, 107, 134, 173, 179, 195, 205–208, 210
onnx_gen_ai_completion_base.py	62	23	62%	17–18, 40, 70–76, 80–85, 87–93
TOTAL	27402	4607	83%

Python Unit Test Overview

Tests	Skipped	Failures	Errors	Time
3765	23 💤	0 ❌	0 🔥	1m 44s ⏱️

[Evan Mattson (moonbox3)]

Nico Möller (@nmoeller) thanks for handling this. Looks like the uv.lock needs a local refresh, and a push upstream.

[Nico Möller (nmoeller)]

Evan Mattson (@moonbox3) always happy to help here ! I pushed an updated uv.lock file. Could you maybe restart the Pipelines here ? I think the error in the Check .md links pipeline was just temporary, since it says 403 forbidden to the .NET Download Link.

[Evan Mattson (moonbox3)]

Nico Möller (@nmoeller) there are some conflicts we should fix.

[Nico Möller (nmoeller)]

Evan Mattson (@moonbox3) my branch is updated with the latest changes of main now.