and sadly it was unclear documentation that tf.logging.info() outputs to stderr not stdout, and thus popen does not work in the usual way, but simple fix, once known.
training_results = Popen(cmds,shell=False,stderr=PIPE,bufsize=1,executable="python")
resolved!
and sadly it was unclear documentation that tf.logging.info() outputs to stderr not stdout, and thus popen does not work in the usual way, but simple fix, once known.
training_results = Popen(cmds,shell=False,stderr=PIPE,bufsize=1,executable="python")