search

nanoporetech / medaka

Sequence correction provided by ONT Research
https://nanoporetech.com
Other
391 stars 73 forks source link

Could not load symbol cublasGetSmCountTarget from libcublas.so.11. #472

Closed aksti closed 7 months ago

aksti commented 7 months ago

Hello,

While running Medaka (UBUNTU 18.04; Conda environment; Python 3.9.18; CUDA 11.2; cuBLAS 11.4), the following error occurred:

Could not load symbol cublasGetSmCountTarget from libcublas.so.11. Error: /usr/local/cuda/targets/x86_64-linux/lib/libcublas.so.11: undefined symbol: cublasGetSmCountTarget

The pipeline runs despite the error, but are there any drawbacks? How can I fix this error?

cjw85 commented 7 months ago

Hi @aksti,

I'm curious that you say you have CUDA 11.2 but cuBLAS 11.4. The version of tensorflow used by medaka is linked to CUDA 11.2.

aksti commented 7 months ago

Thanks for the answer. Here all versions installed automatically on my system. { "cuda" : { "name" : "CUDA SDK", "version" : "11.2.2" }, "cuda_cudart" : { "name" : "CUDA Runtime (cudart)", "version" : "11.2.152" }, "cuda_cuobjdump" : { "name" : "cuobjdump", "version" : "11.2.152" }, "cuda_cupti" : { "name" : "CUPTI", "version" : "11.2.152" }, "cuda_cuxxfilt" : { "name" : "CUDA cu++ filt", "version" : "11.2.152" }, "cuda_demo_suite" : { "name" : "CUDA Demo Suite", "version" : "11.2.152" }, "cuda_gdb" : { "name" : "CUDA GDB", "version" : "11.2.152" }, "cuda_memcheck" : { "name" : "CUDA Memcheck", "version" : "11.2.152" }, "cuda_nsight" : { "name" : "Nsight Eclipse Plugins", "version" : "11.2.152" }, "cuda_nvcc" : { "name" : "CUDA NVCC", "version" : "11.2.152" }, "cuda_nvdisasm" : { "name" : "CUDA nvdisasm", "version" : "11.2.152" }, "cuda_nvml_dev" : { "name" : "CUDA NVML Headers", "version" : "11.2.152" }, "cuda_nvprof" : { "name" : "CUDA nvprof", "version" : "11.2.152" }, "cuda_nvprune" : { "name" : "CUDA nvprune", "version" : "11.2.152" }, "cuda_nvrtc" : { "name" : "CUDA NVRTC", "version" : "11.2.152" }, "cuda_nvtx" : { "name" : "CUDA NVTX", "version" : "11.2.152" }, "cuda_nvvp" : { "name" : "CUDA NVVP", "version" : "11.2.152" }, "cuda_samples" : { "name" : "CUDA Samples", "version" : "11.2.152" }, "cuda_sanitizer_api" : { "name" : "CUDA Compute Sanitizer API", "version" : "11.2.152" }, "fabricmanager" : { "name" : "Fabric Manager", "version" : "460.32.03" }, "libcublas" : { "name" : "CUDA cuBLAS", "version" : "11.4.1.1043" }, "libcufft" : { "name" : "CUDA cuFFT", "version" : "10.4.1.152" }, "libcurand" : { "name" : "CUDA cuRAND", "version" : "10.2.3.152" }, "libcusolver" : { "name" : "CUDA cuSOLVER", "version" : "11.1.0.152" }, "libcusparse" : { "name" : "CUDA cuSPARSE", "version" : "11.4.1.1152" }, "libnpp" : { "name" : "CUDA NPP", "version" : "11.3.2.152" }, "libnvjpeg" : { "name" : "CUDA nvJPEG", "version" : "11.4.0.152" }, "nsight_compute" : { "name" : "Nsight Compute", "version" : "2020.3.1.4" }, "nsight_systems" : { "name" : "Nsight Systems", "version" : "2020.4.3.7" }, "nvidia_driver" : { "name" : "NVIDIA Linux Driver", "version" : "460.32.03" } }

Regards, Sascha

cjw85 commented 7 months ago

You appear to have a rather eclectic mix of versions of CUDA libraries being displayed here. I suggest talking to your system administrator to understand this more.