I have access to a tensorflow speech enhancement model running within a docker container on a remote machine with Ubuntu OS. The tensorflow application crashes everytime at interpreter.invoke() stage without any logs. I am not able to pin-point the exact reason for the crash even after trying various ways to access the logs(python logging package/tf.get_logger()). I am not sure if the containerization of the application is creating an issue in accessing the generated logs. As per a similar thread here https://github.com/tensorflow/tensorflow/issues/48106, the user was able to get the "python crash logs" and pinpoint the exact function creating the crash. Can you please let me know how I can access the logs in a way that has been obtained in the thread linked above given the application has been containerized?
Asked
Active
Viewed 19 times
