models VS onnxruntime

Uh, maybe they didn't feel the need to look. I already pointed you to the ONNX project. Here are some ONNX-based. These are just the ones being shared with the community. The limit of AMD's responsibility is writing the low-level libraries to support ONNX.

I am new to object detection models and I need help running my object detection Darknet YOLOv4-Tiny Model In Unity Barracuda. I trained my model and then i converted it to ONNX format with 2 methods. One method was using pytorch-YOLOv4 from github and the other by converting my model to tensorflow and then to onnx and shown here: "https://github.com/onnx/models/blob/main/vision/object_detection_segmentation/yolov4/dependencies/Conversion.ipynb"

Then i tried to convert it again using another method that i found here "https://github.com/onnx/models/blob/main/vision/object_detection_segmentation/yolov4/dependencies/Conversion.ipynb" in order to convert it from darknet to tensorflow and then to onnx but i didn't have any luck.

Here we use the GPT-2 model distributed by the ONNX official. Download GPT-2-LM-HEAD from the link.

Download pre-trained models from the ONNX Model Zoo

Instructions to reproduce the problem: I am trying to convert a proprietary model at work but for now i'll use mobilenetv2-7.onnx to explain/reproduce the issue.

One approach could be run a pre-trained object detector (like one of these) on each frame and then a simple object tracker on top of it (like this).

ONNX Runtime: ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator

Embedding generation uses the ONNX runtime under the hood. This is a cross-platform inferencing library that supports multiple execution providers from CPU to specialized GPUs.

tfjs is dead, looking at the commit history. The standard now is to convert PyTorch to onnx, then use onnxruntime (https://github.com/microsoft/onnxruntime/tree/main/js/web) to run the model on the browsdr.

- slang[2] has the potential, but the meta programming part is not as strong as C++, existing libraries cannot be used.

The above conclusion is drawn from my work https://github.com/microsoft/onnxruntime/tree/dev/opencl, purely nightmare to work with thoes drivers and jit compilers. Hopefully Vcc can take compute shader more seriously.

[1]: https://www.circle-lang.org/

I did. It depends what you want, for an overview of how ONNX Runtime works then Microsoft have a bunch of things on https://onnxruntime.ai, but the Java content is a bit lacking on there as I've not had time to write much. Eventually I'll probably write something similar to the C# SD tutorial they have on there but for the Java API.

For writing ONNX models from Java we added an ONNX export system to Tribuo in 2022 which can be used by anything on the JVM to export ONNX models in an easier way than writing a protobuf directly. Tribuo doesn't have full coverage of the ONNX spec, but we're happy to accept PRs to expand it, otherwise it'll fill out as we need it.

What about models besides GPT? Most of the popular vector encoding models aren't using this architecture.

If you really didn't want PyTorch/Transformers, you could consider exporting your models to ONNX (https://github.com/microsoft/onnxruntime).