onnxruntime

onnxruntime copied to clipboard

VitisAI EP: Add compilation timeout and better error handling for transformer models

3

## Problem Statement VitisAI execution provider hangs indefinitely when attempting to compile transformer models (BERT) on AMD Ryzen AI hardware, requiring manual process termination. This affects developer experience and makes...

cornerstone-report

### Description Fix a critical typo. ### Motivation and Context This misalignment would cause the OGA PR to fail.

BoarQing

[CPU][ConvTranspose] Nondeterministic outputs across runs (incl. NaNs) on WSL

5

### Describe the issue A tiny ONNX model with a single ConvTranspose node produces different results between runs on CPU (WSL). The input is all zeros, weights/bias are constants, yet...

faresernez

Expose start profiling API

2

### Description Add start profiling API in ORT. With this, we can profile for a time span. Based on this, we have [another genAI PR](https://github.com/microsoft/onnxruntime-genai/pull/1898) to support start/end profiling in...

xiaofeihan1

[Build] Eigen dependency download fails with SHA1 hash mismatch - build broken for v1.19.2

1

### Describe the issue ONNX Runtime 1.19.2 build fails during CMake configure when FetchContent attempts to download the Eigen library. The SHA1 hash of the downloaded archive from GitLab no...

1saeed

EP_FAIL : Non-zero status code returned while running Conv node. Name:'/features/features.0/Conv' Status Message: Failed to initialize CUDNN Frontend

13

I have an EC2 instance of type g5g.xlarge. I have installed the following: ``` CUDA-Toolit: Cuda compilation tools, release 12.4, V12.4.131 CUDNN Version: 9.6.0 Python: 3.12 Pytorch: Compiled from source...

m0hammadjaan

Pip cannot find package for Nvidia DGX Spark (arm linux)

6

### Describe the issue Trying to setup ai libraries for training, running into issue with pip install onnxruntime-gpu==1.21 Could not find a version that satisfies the requirement. Likely because its...

stierma1

[Performance][CPU] High CPU usage in onnxruntime webassembly build for CPP

13

### Describe the issue I’m seeing **significantly higher CPU usage** when running model inference in the browser via ONNX Runtime compiled for WebAssembly. While adding threads *does* reduce latency (great!),...

shriramrsr

TypeScript: Allow constructing float16 tensors using Float16Array

4

Fixes #26741 This change updates the TypeScript definitions to allow constructing `float16` tensors using `Float16Array` in environments where it is available. Runtime behavior remains unchanged (`float16` is still represented internally...

rockygeekz

[Performance] cudaMemcpyAsync dominates runtime for batch inference with FP32 inputs

3

### Describe the issue ### Describe When running a batch (batch=6, size=[6\*3\*1280\*1280]) FP32 inference on GPU (EP: CUDA Provider) with ONNX Runtime, the majority of the time is spent in...

GilbertPan97