All issues
Issue creation is restricted in this repository
问题
is:issue state:open
is:issue state:open
搜索 results
softmax_warp_forward overflows int32 above 2^31 elements and can return silently wrong results (CUDA and ROCm EPs)
ep:CUDAissues related to the CUDA execution providerissues related to the CUDA execution providerStatus: Open.#32299 In microsoft/onnxruntime;XNNPACK EP on Android arm64 aborts the process (Abort message: 'terminating') instead of returning an error — Resize with empty scales, still in 1.26.0
.NET拉取请求 that update .net code拉取请求 that update .net codeapi:CSharpissues related to the C# APIissues related to the C# APIep:Xnnpackissues related to XNNPACK EPissues related to XNNPACK EPplatform:mobileissues related to ONNX Runtime mobile; typically submitted using templateissues related to ONNX Runtime mobile; typically submitted using templateStatus: Open.#32298 In microsoft/onnxruntime;[Build] Align CPython 3.14 CUDA 13 wheels with TensorRT 11
ep:CUDAissues related to the CUDA execution providerissues related to the CUDA execution providerep:TensorRTissues related to TensorRT execution providerissues related to TensorRT execution providerStatus: Open.#32278 In microsoft/onnxruntime;TensorRT EP Resize with +inf input returns NaN while CPU/CUDA EP return +inf
ep:CUDAissues related to the CUDA execution providerissues related to the CUDA execution providerep:TensorRTissues related to TensorRT execution providerissues related to TensorRT execution providerStatus: Open.#32258 In microsoft/onnxruntime;CUDA EP ReduceMax/ReduceMin return finite limits instead of +/-inf for all-infinite inputs
ep:CUDAissues related to the CUDA execution providerissues related to the CUDA execution providerStatus: Open.#32256 In microsoft/onnxruntime;[Performance] ORT 1.29 CPU FP16 Gemm is ~3500x slower than 1.28 on Windows x64
api:CSharpissues related to the C# APIissues related to the C# APIperformanceissues related to performance regressionsissues related to performance regressionsplatform:windowsissues related to the Windows platformissues related to the Windows platformStatus: Open.#32255 In microsoft/onnxruntime;TensorRT EP engine cache reuses engine for same graph name with different static input shape
ep:TensorRTissues related to TensorRT execution providerissues related to TensorRT execution providerStatus: Open.#32254 In microsoft/onnxruntime;cuda graph bug for recurrent state inplace update
ep:CUDAissues related to the CUDA execution providerissues related to the CUDA execution providerStatus: Open.#32243 In microsoft/onnxruntime;[Web] Deprecate the onnxruntime-web WebGL backend
ep:WebGPUort-web webgpu providerort-web webgpu providerplatform:webissues related to ONNX Runtime web; typically submitted using templateissues related to ONNX Runtime web; typically submitted using templateStatus: Open.#32241 In microsoft/onnxruntime;[Build] onnxruntime-node install fails: "Failed to download build list. HTTP status code = 302" (script doesn't follow redirects)
.NET拉取请求 that update .net code拉取请求 that update .net codebuildbuild issues; typically submitted using templatebuild issues; typically submitted using templateplatform:webissues related to ONNX Runtime web; typically submitted using templateissues related to ONNX Runtime web; typically submitted using templateStatus: Open.#32233 In microsoft/onnxruntime;[Feature Request] WebGPU LinearAttention state_window for speculative/MTP rollback
ep:WebGPUort-web webgpu providerort-web webgpu providerplatform:webissues related to ONNX Runtime web; typically submitted using templateissues related to ONNX Runtime web; typically submitted using templateStatus: Open.#32232 In microsoft/onnxruntime;[Bug] GemmTransposeFusion folds an identity Transpose (perm=[0,1]) into Gemm transA/transB, producing silently wrong results
ep:CoreMLissues related to CoreML execution providerissues related to CoreML execution providerStatus: Open.#32228 In microsoft/onnxruntime;