search

HawkAaron / warp-transducer

A fast parallel implementation of RNN Transducer.
Apache License 2.0
307 stars 124 forks source link

nvcc fatal : Unsupported gpu architecture 'compute_30' while execute make #79

Closed GGzhangBOY closed 3 years ago

GGzhangBOY commented 3 years ago

The enviroument: RTX2070, Driver v450.80.02, and CUDA 11.0, ubuntu 20.04 I find that the compute capability of my GPU is 7.5, is there any modification in CMakeLists.txt? Can anyone help me out? I am new to DL under Linux enviroument :D

magdacisowska commented 3 years ago

just comment out those 3 lines in CMakeLists.txt:

# set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS} -gencode arch=compute_30,code=sm_30 -O2")
# set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS} -gencode arch=compute_35,code=sm_35")

set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS} -gencode arch=compute_50,code=sm_50")
# set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS} -gencode arch=compute_52,code=sm_52")
GGzhangBOY commented 3 years ago

just comment out those 3 lines in CMakeLists.txt:

# set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS} -gencode arch=compute_30,code=sm_30 -O2")
# set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS} -gencode arch=compute_35,code=sm_35")

set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS} -gencode arch=compute_50,code=sm_50")
# set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS} -gencode arch=compute_52,code=sm_52")

Thank you!!! Really save my day!!