[Build] my cuda is 12. which version shold i install , torch.__version__ '2.2.0+cu121'

[LucasLOOT]

Describe the issue

my cuda is 12.2 which version shold i install

Urgency

No response

Target platform

ubuntu22.04

Build script

none

Error / output

2024-02-22 17:02:52.455386347 [E:onnxruntime:Default, provider_bridge_ort.cc:1480 TryGetProviderInfo_CUDA] /onnxruntime_src/onnxruntime/core/session/provider_bridge_ort.cc:1193 onnxruntime::Provider& onnxruntime::ProviderLibrary::Get() [ONNXRuntimeError] : 1 : FAIL : Failed to load library libonnxruntime_providers_cuda.so with error: libcublasLt.so.11: cannot open shared object file: No such file or directory

2024-02-22 17:02:52.455408227 [W:onnxruntime:Default, onnxruntime_pybind_state.cc:747 CreateExecutionProviderInstance] Failed to create CUDAExecutionProvider. Please reference https://onnxruntime.ai/docs/execution-providers/CUDA-ExecutionProvider.html#requirements to ensure all dependencies are met.
2024-02-22 17:02:52.936640939 [E:onnxruntime:Default, provider_bridge_ort.cc:1480 TryGetProviderInfo_CUDA] /onnxruntime_src/onnxruntime/core/session/provider_bridge_ort.cc:1193 onnxruntime::Provider& onnxruntime::ProviderLibrary::Get() [ONNXRuntimeError] : 1 : FAIL : Failed to load library libonnxruntime_providers_cuda.so with error: libcublasLt.so.11: cannot open shared object file: No such file or directory

2024-02-22 17:02:52.936663879 [W:onnxruntime:Default, onnxruntime_pybind_state.cc:747 CreateExecutionProviderInstance] Failed to create CUDAExecutionProvider. Please reference https://onnxruntime.ai/docs/execution-providers/CUDA-ExecutionProvider.html#requirements to ensure all dependencies are met.

Visual Studio Version

No response

GCC / Compiler Version

No response

[snnn]

@jchen351 , would you please help?

[jchen351]

We are using cuda12.2 as well. PyTorch hasn't supported cuda 12 yet last time I checked.

[tianleiwu]

@Allfather9,
See https://onnxruntime.ai/docs/install/#install-onnx-runtime-gpu-cuda-12x
ORT for CUDA 12.x shall be compatible with torch 2.2.0+cu121, CUDA 12.x, cuDNN 8.x