Mehdi Bahri

pip package not up to date?

8

Hi First of all, thank you for sharing this package! I'm installing spherecluster with pip and had to manually edit spherical_kmeans.py to fix the import of _k_means (I changed it...

[C++] Symbols not shown for class implementations

2

Hi, First of all, thank you for providing this plugin! It really fills a gap. There is an issue with C++ code. Consider the below minimal working example: ``` //...

Fully-working example with dynamic batching

5

Hi Thanks a lot for providing this backend. I have tried to use it and I have had some trouble getting Triton to load and run my OpenVINO models. I...

Hello The latest Triton image from NGC (24.04) includes the DALI backend, but the conda-packed environment that used to be shipped with the backend is no longer there. This breaks...

Confusion about versions and NGC images

4

Hi Thank you for the great work you're doing on TensorRT-LLM and the Triton backend. I have some questions on matching versions between the tensorrt-llm python package, the backend, and...

[Transformers Optimizer] CLIP-ViT encoder attention not getting fused

2

### Describe the issue I'm trying to optimize the vision encoder of a CLIP model exported from HuggingFace Transformers, but the attention subgraphs don't get fused. I tried with a...

OpenVINO EP doesn't respect threading parameters

1

**Description** In ONNXRuntime, the OpenVINO EP accepts configuration options to set the number of threads and number of streams documented [here](https://onnxruntime.ai/docs/execution-providers/OpenVINO-ExecutionProvider.html#cc-api-20), but these are ignored when passed to the EP...

Python backend SHM memory leak

24

**Description** I am encountering two possibly related issues with the Python backend and shared memory: 1. During operation, the shared memory usage keeps growing, leading to errors. It looks like...