Search results

Support group query attention in Attention(23) CUDA

ep:CUDAissues related to the CUDA execution providerissues related to the CUDA execution provider

Status: Open.

#27081 In microsoft/onnxruntime;

· titaiwangms opened on Jan 20, 2026

[CUDA] Qwen3 Model quality regression

ep:CUDAissues related to the CUDA execution providerissues related to the CUDA execution provider

Status: Open.

#27079 In microsoft/onnxruntime;

· jixiongdeng opened on Jan 20, 2026

[Performance] Performance regression in ReduceMean operator for int64 inputs between v1.20.0 and v1.21.0

performanceissues related to performance regressionsissues related to performance regressions

Status: Open.

#27076 In microsoft/onnxruntime;

· junghyunpark2001 opened on Jan 20, 2026

[Performance] Severe performance regression in Sigmoid operator for float64 inputs between v1.20.0 and v1.21.0

performanceissues related to performance regressionsissues related to performance regressions

Status: Open.

#27074 In microsoft/onnxruntime;

· junghyunpark2001 opened on Jan 20, 2026

[Performance] Performance regression in Hardmax operator for float32 inputs between v1.22.0 and v1.23.0

performanceissues related to performance regressionsissues related to performance regressions

Status: Open.

#27073 In microsoft/onnxruntime;

· junghyunpark2001 opened on Jan 20, 2026

Is it possible to build ONNX Runtime on IBM AIX?

Status: Open.

#27072 In microsoft/onnxruntime;

· rishabhmadan19 opened on Jan 20, 2026

[Web] Report <code>Buffer (unlabled) used in submit while destoryed</code> when enabling SDXL Turbo demo

ep:WebGPUort-web webgpu providerort-web webgpu provider

ep:WebNNWebNN execution providerWebNN execution provider

platform:webissues related to ONNX Runtime web; typically submitted using templateissues related to ONNX Runtime web; typically submitted using template

Status: Open.

#27068 In microsoft/onnxruntime;

· xhcao opened on Jan 20, 2026

[Performance] Performance regression in And operator for boolean broadcast operations between v1.18.0 and v1.19.0

performanceissues related to performance regressionsissues related to performance regressions

Status: Open.

#27065 In microsoft/onnxruntime;

· junghyunpark2001 opened on Jan 20, 2026

Feature Request: Release GIL during session initialization to enable parallel multi-GPU loading

Status: Open.

#27063 In microsoft/onnxruntime;

· tolleybot opened on Jan 19, 2026

[Mobile] [Bug] onnxruntime-react-native fails to load model in Expo 54 standalone iOS

ep:CoreMLissues related to CoreML execution providerissues related to CoreML execution provider

platform:mobileissues related to ONNX Runtime mobile; typically submitted using templateissues related to ONNX Runtime mobile; typically submitted using template

platform:webissues related to ONNX Runtime web; typically submitted using templateissues related to ONNX Runtime web; typically submitted using template

Status: Open.

#27062 In microsoft/onnxruntime;

· shaneeelliott opened on Jan 19, 2026

[Feature Request] Support optrace profiling with QNNExecutionProvider in the published onnxruntime-qnn wheels

ep:QNNissues related to QNN exeution providerissues related to QNN exeution provider

feature requestrequest for unsupported feature or enhancementrequest for unsupported feature or enhancement

Status: Open.

#27060 In microsoft/onnxruntime;

· martinResearch opened on Jan 19, 2026

[Performance] Performance regression in GatherND operator between v1.20.0 and v1.21.0

performanceissues related to performance regressionsissues related to performance regressions

Status: Open.

#27053 In microsoft/onnxruntime;

· junghyunpark2001 opened on Jan 18, 2026