Open sleebapaul opened 4 years ago
My System
Trace of cat /etc/nv_tegra_release
cat /etc/nv_tegra_release
# R32 (release), REVISION: 4.2, GCID: 20074772, BOARD: t210ref, EABI: aarch64, DATE: Thu Apr 9 01:22:12 UTC 2020
What I've done?
I've downloaded the latest Jetpack SDk from here.
The file name was nv-jetson-nano-sd-card-image-r32.4.2.zip. Flashed it using Baleno Etcher.
nv-jetson-nano-sd-card-image-r32.4.2.zip
Baleno Etcher
Installed TensorFlow 1.13.1 using the command mentioned in this repository.
1.13.1
pip3 install --extra-index-url https://developer.download.nvidia.com/compute/redist/jp/v42 tensorflow-gpu==1.13.1+nv19.5 --user
Then comes the following error. TensorFlow 1.13.1 requires CUDA 10.0 and it can't find the same in the Jetson Nano.
10.0
ImportError: libcublas.so.10.0: cannot open shared object file: No such file or directory
Because, by default, the CUDA version installed in Jetson Nano is 10.2.
10.2
Another issue is TensorRT version. TensorFlow 1.13.1 is built for TensorRT 5.0. The default TensorRT version is 7.1.
5.0
7.1
Questions
Can I downgrade the CUDA version to 10.0 ? How to do that?
Can I downgrade the TensorRT version to 5? How to do that?
How to make this repository work on r32.4.2?
r32.4.2
No news on this issue?
My System
Trace of
cat /etc/nv_tegra_release
# R32 (release), REVISION: 4.2, GCID: 20074772, BOARD: t210ref, EABI: aarch64, DATE: Thu Apr 9 01:22:12 UTC 2020
What I've done?
I've downloaded the latest Jetpack SDk from here.
The file name was
nv-jetson-nano-sd-card-image-r32.4.2.zip
. Flashed it usingBaleno Etcher
.Installed TensorFlow
1.13.1
using the command mentioned in this repository.Then comes the following error. TensorFlow
1.13.1
requires CUDA10.0
and it can't find the same in the Jetson Nano.Because, by default, the CUDA version installed in Jetson Nano is
10.2
.Another issue is TensorRT version. TensorFlow
1.13.1
is built for TensorRT5.0
. The default TensorRT version is7.1
.Questions
Can I downgrade the CUDA version to 10.0 ? How to do that?
Can I downgrade the TensorRT version to 5? How to do that?
How to make this repository work on
r32.4.2
?