Cannot compile samples from amd64 ngc tensorRT containers

I am using this ngc image:
nvcr.io/nvidia/tensorrt:20.03-py3

When running container just -as it’s explained at the overview -and trying to compile a demo,I get:
…/Makefile.config:7: CUDA_INSTALL_DIR variable is not specified, using /usr/local/cuda by default, use CUDA_INSTALL_DIR=<cuda_directory> to change.
…/Makefile.config:10: CUDNN_INSTALL_DIR variable is not specified, using $CUDA_INSTALL_DIR by default, use CUDNN_INSTALL_DIR=<cudnn_directory> to change.

export CUDA_INSTALL_DIR=/usr/local/cuda/bin
export CUDNN_INSTALL_DIR=/usr/include/cudnn.h

Trying to make with these changes I get:

/usr/bin/ld: cannot find -lcudart
collect2: error: ld returned 1 exit status
../Makefile.config:176: recipe for target '../../bin/sample_char_rnn_debug' failed
make[1]: *** [../../bin/sample_char_rnn_debug] Error 1
make[1]: *** Waiting for unfinished jobs....
/usr/bin/ld: cannot find -lcudart
collect2: error: ld returned 1 exit status
../Makefile.config:170: recipe for target '../../bin/sample_char_rnn' failed
make[1]: *** [../../bin/sample_char_rnn] Error 1
make[1]: Leaving directory '/workspace/tensorrt/samples/sampleCharRNN'
Makefile:56: recipe for target 'all' failed
make: *** [all] Error 2

I’ve looking around other post with this issue but non of them solved me this problem. Otherwise, Should this be straight forward? It’s clearly explained that I can compile examples (and I am in a new fresh container):
https://docs.nvidia.com/deeplearning/tensorrt/container-release-notes/rel_21-03.html#rel_21-03

Am I doing something wrong?