Open TPIOS opened 3 years ago
Hi, Can you post some more information? Does your school have WML-CE or conda itself centrally installed? The license accept should save a file in your home that marks your install has accepted the license. If your home dir, or conda, or conda package locations are on various remote file locations that could complicate things. Post a bit more info and we can try to help.
I don't think our school uses WML-CE before. I try to run the License Acceptance Process directly by command lapc. But when I type "1" and enter. It said "Exited with 15. The license is not accepted." What is the exit code "15" meaning?
I think its not able to write out the accept file. Where do you have conda installed? What is your home dir like?
So, is that a permission error? The anaconda is installed by the administrator of our school server. I just export the installation path to my home dir and "/home/myusername/". I create .conda directory under my home dir and create Python virtual environment there.
Here is the full error output
Exited with: 15 License NOT accepted. Installation will not proceed. b'License NOT accepted. Installation will not proceed.\n' done ERROR conda.core.link:_execute(700): An error occurred while installing package 'https://public.dhe.ibm.com/ibmdl/export/pub/software/server/ibm-ai/conda::powerai-license-1.6.0-536.10b2576'. Rolling back transaction: done
LinkError: post-link script failed for package https://public.dhe.ibm.com/ibmdl/export/pub/software/server/ibm-ai/conda::powerai-license-1.6.0-536.10b2576 location of failed script: $environment_path$/bin/.powerai-license-post-link.sh ==> script messages <== License NOT accepted. Installation will not proceed.
==> script output <== stdout: stderr: return code: 1
()
By the way, can I use Tensorflow 2.1 with CUDA 10.1 provided by powerai, since our school server didn't provide CUDA10.2 driver?
Hi, I am very interested in using TensorFlow with large model support. When I tried to install the tensorflow-gpu from WML CE, I always get stuck with reading the IBM license, which requires me to:
Press Enter to continue viewing the license agreement, or, Enter "1" to accept the agreement, "2" to decline it or "99" to go back to the previous screen, "3" Print, "4" Read non-IBM terms.
I press "1" many times but the Executing transaction process seems running forever so that I can't install the packages.
Since I try to install this package on my school server, I am wondering if I should have a root permission to accept the license when using conda to install the tensorflow-gpu from WML CE.
Thank you very much!