Closed Akashtyagi closed 5 years ago
Hi, could you please detail what version of python and tensorflow you're using?
It was due to the Tesnsorflow version. Earlier i was using Tensorflow 1.10.0 on Python3. Now i downgraded it to Tensorflow1.7.0 and that particular issue got resolved.
But now couldn't train the model due the following error !
That normally happens when you run out of memory - try adding the --testing
flag, which reduces the number of parameters and batch size. If that works, then RAM or lack of GPU is your limiting factor, and you'll need to train with a smaller model or smaller batch size.
Did you mean to add '--testing' flag in the command to terminal while running the script, because that didn't help ! Can you please help figure it out !
I am using the command - "
bash train.sh
Hi,
I can't help without knowing more about the system that you're attempting to run the model on. That error is being thrown because of some lower level system error, not because of the model - it normally happens when there's not sufficient memory to run the script. Have a look at this stackoverflow question for more details: https://stackoverflow.com/questions/45735858/why-tensorflow-just-outputs-killed
While trying to implement the code i ran into error when executing the train.sh, it shows an Attribute error as follow