Error when attempting to install nv-wavenet

[Tylersuard]

Hello. I am attempting to run this project in Google Colab. When I run the command "!python setup.py install" in the src/nv-wavenet directory as directed, I get 28 errors (see below) .

Can you tell me what is wrong with my notebook, and how to fix it so that it runs? Link below.
Thank you for your help!
-Tyler Suard

https://colab.research.google.com/drive/1XqC2zeEkogtx8uGbXAmpFlt3FohnoQAs

running install
running bdist_egg
running egg_info
writing nv_wavenet_ext.egg-info/PKG-INFO
writing dependency_links to nv_wavenet_ext.egg-info/dependency_links.txt
writing top-level names to nv_wavenet_ext.egg-info/top_level.txt
writing manifest file 'nv_wavenet_ext.egg-info/SOURCES.txt'
installing library code to build/bdist.linux-x86_64/egg
running install_lib
running build_ext
building 'nv_wavenet_ext' extension
/usr/local/cuda/bin/nvcc -I/usr/local/lib/python3.6/dist-packages/torch/lib/include -I/usr/local/lib/python3.6/dist-packages/torch/lib/include/torch/csrc/api/include -I/usr/local/lib/python3.6/dist-packages/torch/lib/include/TH -I/usr/local/lib/python3.6/dist-packages/torch/lib/include/THC -I/usr/local/cuda/include -I/usr/include/python3.6m -c wavenet_infer.cu -o build/temp.linux-x86_64-3.6/wavenet_infer.o --compiler-options '-fPIC' -arch=sm_70 -std=c++14 --use_fast_math -maxrregcount 128 --ptxas-options=-v --expt-relaxed-constexpr -D__GNUC__=6 -DTORCH_API_INCLUDE_EXTENSION_H -DTORCH_EXTENSION_NAME=nv_wavenet_ext -D_GLIBCXX_USE_CXX11_ABI=0
:0:0: warning: "GNUC" redefined
: note: this is the location of the previous definition
nv_wavenet.cuh(116): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_prev<T_weight,T_data,R,BATCH_UNROLL>(int, int, int, int, int, int, T_weight *, T_data *, T_data *, T_data *, __nv_bool) [with T_weight=half2, T_data=half, R=128, BATCH_UNROLL=4]"
nv_wavenet_dualblock.cuh(108): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=4]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=4]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=4]"
(555): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(124): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_prev<T_weight,T_data,R,BATCH_UNROLL>(int, int, int, int, int, int, T_weight *, T_data *, T_data *, T_data *, __nv_bool) [with T_weight=half2, T_data=half, R=128, BATCH_UNROLL=4]"
nv_wavenet_dualblock.cuh(108): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=4]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=4]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=4]"
(555): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(141): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_cur<T_weight,T_data,R,BATCH_UNROLL>(int, int, int, int, int, T_weight *, T_data , T_data , T_data ()[R], T_data ()[], T_data *) [with T_weight=half2, T_data=half, R=128, BATCH_UNROLL=4]"
nv_wavenet_dualblock.cuh(142): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=4]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=4]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=4]"
(555): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(166): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_pointwise<T_weight,T_data,R,S,BATCH_UNROLL,DUAL_BLOCK>(int, int, int, int, int, T_data , T_data ()[R], T_data ()[], T_data ()[R], T_data *, volatile int *) [with T_weight=half2, T_data=half, R=128, S=128, BATCH_UNROLL=4, DUAL_BLOCK=true]"
nv_wavenet_dualblock.cuh(146): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=4]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=4]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=4]"
(555): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(203): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_res<T_weight,T_data,R,S,BATCH_UNROLL,DUAL_BLOCK>(int, int, int, int, int, int, T_weight , T_data , T_data ()[R], T_data ()[R], T_data *, T_data *, __nv_bool) [with T_weight=half2, T_data=half, R=128, S=128, BATCH_UNROLL=4, DUAL_BLOCK=true]"
nv_wavenet_dualblock.cuh(150): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=4]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=4]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=4]"
(555): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(116): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_prev<T_weight,T_data,R,BATCH_UNROLL>(int, int, int, int, int, int, T_weight *, T_data *, T_data *, T_data *, __nv_bool) [with T_weight=half2, T_data=half, R=128, BATCH_UNROLL=3]"
nv_wavenet_dualblock.cuh(108): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=3]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=3]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=3]"
(559): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(124): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_prev<T_weight,T_data,R,BATCH_UNROLL>(int, int, int, int, int, int, T_weight *, T_data *, T_data *, T_data *, __nv_bool) [with T_weight=half2, T_data=half, R=128, BATCH_UNROLL=3]"
nv_wavenet_dualblock.cuh(108): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=3]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=3]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=3]"
(559): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(141): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_cur<T_weight,T_data,R,BATCH_UNROLL>(int, int, int, int, int, T_weight *, T_data , T_data , T_data ()[R], T_data ()[], T_data *) [with T_weight=half2, T_data=half, R=128, BATCH_UNROLL=3]"
nv_wavenet_dualblock.cuh(142): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=3]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=3]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=3]"
(559): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(166): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_pointwise<T_weight,T_data,R,S,BATCH_UNROLL,DUAL_BLOCK>(int, int, int, int, int, T_data , T_data ()[R], T_data ()[], T_data ()[R], T_data *, volatile int *) [with T_weight=half2, T_data=half, R=128, S=128, BATCH_UNROLL=3, DUAL_BLOCK=true]"
nv_wavenet_dualblock.cuh(146): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=3]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=3]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=3]"
(559): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(203): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_res<T_weight,T_data,R,S,BATCH_UNROLL,DUAL_BLOCK>(int, int, int, int, int, int, T_weight , T_data , T_data ()[R], T_data ()[R], T_data *, T_data *, __nv_bool) [with T_weight=half2, T_data=half, R=128, S=128, BATCH_UNROLL=3, DUAL_BLOCK=true]"
nv_wavenet_dualblock.cuh(150): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=3]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=3]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=3]"
(559): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(116): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_prev<T_weight,T_data,R,BATCH_UNROLL>(int, int, int, int, int, int, T_weight *, T_data *, T_data *, T_data *, __nv_bool) [with T_weight=half2, T_data=half, R=128, BATCH_UNROLL=2]"
nv_wavenet_dualblock.cuh(108): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=2]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=2]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=2]"
(563): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(124): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_prev<T_weight,T_data,R,BATCH_UNROLL>(int, int, int, int, int, int, T_weight *, T_data *, T_data *, T_data *, __nv_bool) [with T_weight=half2, T_data=half, R=128, BATCH_UNROLL=2]"
nv_wavenet_dualblock.cuh(108): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=2]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=2]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=2]"
(563): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(141): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_cur<T_weight,T_data,R,BATCH_UNROLL>(int, int, int, int, int, T_weight *, T_data , T_data , T_data ()[R], T_data ()[], T_data *) [with T_weight=half2, T_data=half, R=128, BATCH_UNROLL=2]"
nv_wavenet_dualblock.cuh(142): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=2]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=2]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=2]"
(563): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(166): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_pointwise<T_weight,T_data,R,S,BATCH_UNROLL,DUAL_BLOCK>(int, int, int, int, int, T_data , T_data ()[R], T_data ()[], T_data ()[R], T_data *, volatile int *) [with T_weight=half2, T_data=half, R=128, S=128, BATCH_UNROLL=2, DUAL_BLOCK=true]"
nv_wavenet_dualblock.cuh(146): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=2]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=2]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=2]"
(563): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(203): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_res<T_weight,T_data,R,S,BATCH_UNROLL,DUAL_BLOCK>(int, int, int, int, int, int, T_weight , T_data , T_data ()[R], T_data ()[R], T_data *, T_data *, __nv_bool) [with T_weight=half2, T_data=half, R=128, S=128, BATCH_UNROLL=2, DUAL_BLOCK=true]"
nv_wavenet_dualblock.cuh(150): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=2]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=2]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=2]"
(563): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(116): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_prev<T_weight,T_data,R,BATCH_UNROLL>(int, int, int, int, int, int, T_weight *, T_data *, T_data *, T_data *, __nv_bool) [with T_weight=half2, T_data=half, R=128, BATCH_UNROLL=1]"
nv_wavenet_dualblock.cuh(108): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=1]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=1]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=1]"
(566): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(124): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_prev<T_weight,T_data,R,BATCH_UNROLL>(int, int, int, int, int, int, T_weight *, T_data *, T_data *, T_data *, __nv_bool) [with T_weight=half2, T_data=half, R=128, BATCH_UNROLL=1]"
nv_wavenet_dualblock.cuh(108): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=1]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=1]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=1]"
(566): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(141): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_cur<T_weight,T_data,R,BATCH_UNROLL>(int, int, int, int, int, T_weight *, T_data , T_data , T_data ()[R], T_data ()[], T_data *) [with T_weight=half2, T_data=half, R=128, BATCH_UNROLL=1]"
nv_wavenet_dualblock.cuh(142): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=1]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=1]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=1]"
(566): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(166): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_pointwise<T_weight,T_data,R,S,BATCH_UNROLL,DUAL_BLOCK>(int, int, int, int, int, T_data , T_data ()[R], T_data ()[], T_data ()[R], T_data *, volatile int *) [with T_weight=half2, T_data=half, R=128, S=128, BATCH_UNROLL=1, DUAL_BLOCK=true]"
nv_wavenet_dualblock.cuh(146): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=1]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=1]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=1]"
(566): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(203): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_res<T_weight,T_data,R,S,BATCH_UNROLL,DUAL_BLOCK>(int, int, int, int, int, int, T_weight , T_data , T_data ()[R], T_data ()[R], T_data *, T_data *, __nv_bool) [with T_weight=half2, T_data=half, R=128, S=128, BATCH_UNROLL=1, DUAL_BLOCK=true]"
nv_wavenet_dualblock.cuh(150): here
instantiation of "void nv_wavenet_dualBlock_A<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, int) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=1]"
nv_wavenet_dualblock.cuh(337): here
instantiation of "void nv_wavenet_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=1]"
nv_wavenet_dualblock.cuh(347): here
instantiation of "__nv_bool launch_dualBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=1]"
(566): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(166): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_pointwise<T_weight,T_data,R,S,BATCH_UNROLL,DUAL_BLOCK>(int, int, int, int, int, T_data , T_data ()[R], T_data ()[], T_data ()[R], T_data *, volatile int *) [with T_weight=half2, T_data=half, R=128, S=128, BATCH_UNROLL=4, DUAL_BLOCK=false]"
nv_wavenet_singleblock.cuh(122): here
instantiation of "void nv_wavenet_singleBlock_8R<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=4]"
nv_wavenet_singleblock.cuh(265): here
instantiation of "__nv_bool launch_singleBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=4]"
(574): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(203): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_res<T_weight,T_data,R,S,BATCH_UNROLL,DUAL_BLOCK>(int, int, int, int, int, int, T_weight , T_data , T_data ()[R], T_data ()[R], T_data *, T_data *, __nv_bool) [with T_weight=half2, T_data=half, R=128, S=128, BATCH_UNROLL=4, DUAL_BLOCK=false]"
nv_wavenet_singleblock.cuh(126): here
instantiation of "void nv_wavenet_singleBlock_8R<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=4]"
nv_wavenet_singleblock.cuh(265): here
instantiation of "__nv_bool launch_singleBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=4]"
(574): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(166): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_pointwise<T_weight,T_data,R,S,BATCH_UNROLL,DUAL_BLOCK>(int, int, int, int, int, T_data , T_data ()[R], T_data ()[], T_data ()[R], T_data *, volatile int *) [with T_weight=half2, T_data=half, R=128, S=128, BATCH_UNROLL=3, DUAL_BLOCK=false]"
nv_wavenet_singleblock.cuh(122): here
instantiation of "void nv_wavenet_singleBlock_8R<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=3]"
nv_wavenet_singleblock.cuh(265): here
instantiation of "__nv_bool launch_singleBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=3]"
(578): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(203): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_res<T_weight,T_data,R,S,BATCH_UNROLL,DUAL_BLOCK>(int, int, int, int, int, int, T_weight , T_data , T_data ()[R], T_data ()[R], T_data *, T_data *, __nv_bool) [with T_weight=half2, T_data=half, R=128, S=128, BATCH_UNROLL=3, DUAL_BLOCK=false]"
nv_wavenet_singleblock.cuh(126): here
instantiation of "void nv_wavenet_singleBlock_8R<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=3]"
nv_wavenet_singleblock.cuh(265): here
instantiation of "__nv_bool launch_singleBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=3]"
(578): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(166): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_pointwise<T_weight,T_data,R,S,BATCH_UNROLL,DUAL_BLOCK>(int, int, int, int, int, T_data , T_data ()[R], T_data ()[], T_data ()[R], T_data *, volatile int *) [with T_weight=half2, T_data=half, R=128, S=128, BATCH_UNROLL=2, DUAL_BLOCK=false]"
nv_wavenet_singleblock.cuh(122): here
instantiation of "void nv_wavenet_singleBlock_8R<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=2]"
nv_wavenet_singleblock.cuh(265): here
instantiation of "__nv_bool launch_singleBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=2]"
(582): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(203): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_res<T_weight,T_data,R,S,BATCH_UNROLL,DUAL_BLOCK>(int, int, int, int, int, int, T_weight , T_data , T_data ()[R], T_data ()[R], T_data *, T_data *, __nv_bool) [with T_weight=half2, T_data=half, R=128, S=128, BATCH_UNROLL=2, DUAL_BLOCK=false]"
nv_wavenet_singleblock.cuh(126): here
instantiation of "void nv_wavenet_singleBlock_8R<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=2]"
nv_wavenet_singleblock.cuh(265): here
instantiation of "__nv_bool launch_singleBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=2]"
(582): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(166): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_pointwise<T_weight,T_data,R,S,BATCH_UNROLL,DUAL_BLOCK>(int, int, int, int, int, T_data , T_data ()[R], T_data ()[], T_data ()[R], T_data *, volatile int *) [with T_weight=half2, T_data=half, R=128, S=128, BATCH_UNROLL=1, DUAL_BLOCK=false]"
nv_wavenet_singleblock.cuh(122): here
instantiation of "void nv_wavenet_singleBlock_8R<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=1]"
nv_wavenet_singleblock.cuh(265): here
instantiation of "__nv_bool launch_singleBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=1]"
(585): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

nv_wavenet.cuh(203): error: identifier "namedBarrierSync" is undefined
detected during:
instantiation of "void nv_wavenet_res<T_weight,T_data,R,S,BATCH_UNROLL,DUAL_BLOCK>(int, int, int, int, int, int, T_weight , T_data , T_data ()[R], T_data ()[R], T_data *, T_data *, __nv_bool) [with T_weight=half2, T_data=half, R=128, S=128, BATCH_UNROLL=1, DUAL_BLOCK=false]"
nv_wavenet_singleblock.cuh(126): here
instantiation of "void nv_wavenet_singleBlock_8R<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=1]"
nv_wavenet_singleblock.cuh(265): here
instantiation of "__nv_bool launch_singleBlock<T_weight,T_data,R,S,A,BATCH_UNROLL>(nv_wavenet_params<T_weight, T_data>, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, BATCH_UNROLL=1]"
(585): here
instantiation of "__nv_bool nvWavenetInfer<T_weight, T_data, R, S, A, COND_REPEAT>::run(int, int, int *, int, __nv_bool, cudaStream_t) [with T_weight=half2, T_data=half, R=128, S=128, A=256, COND_REPEAT=800]"
wavenet_infer.cu(108): here

28 errors detected in the compilation of "/tmp/tmpxft_0000023e_00000000-6_wavenet_infer.cpp1.ii".
error: command '/usr/local/cuda/bin/nvcc' failed with exit status 1

[henrhoi]

Getting same error while trying to build nv-wavenet on a Tesla V100.

[SashaBurashnikova]

I also got the same error. Did you find the solution?

[KieRigby]

I managed to fix this by adding the line #include "softmax.cuh" below line 35 in the file src/nv-wavenet/nv_wavenet.cuh

This is just a quickfix though. Maybe a better fix would be to bring in the latest version of nv-wavenet which seems to have this line in that file.