Predicting Baby Weight with TensorFlow on Cloud ML EngineIr al Laboratorio
InvalidArgumentErrorTraceback (most recent call last) <ipython-input-21-a10b616ab2a1> in <module>() 26 27 shutil.rmtree('babyweight_trained', ignore_errors=True) # start fresh each time ---> 28 learn_runner.run(experiment_fn, 'babyweight_trained') /usr/local/lib/python2.7/dist-packages/tensorflow/contrib/learn/python/learn/learn_runner.pyc in run(experiment_fn, output_dir, schedule, run_config, hparams) 216 schedule = schedule or _get_default_schedule(run_config) 217 --> 218 return _execute_schedule(experiment, schedule) 219 220 /usr/local/lib/python2.7/dist-packages/tensorflow/contrib/learn/python/learn/learn_runner.pyc in _execute_schedule(experiment, schedule) 44 logging.error('Allowed values for this experiment are: %s', valid_tasks) 45 raise TypeError('Schedule references non-callable member %s' % schedule) ---> 46 return task() 47 48
Would have been good to get notice of zone limitations before completing preparation steps. Sample code returns lots of deprecation warnings. InvalidArgumentError in training. Jobs training on Cloud ML Engine fail after 5-7 minutes.
Couldn't deploy training on Cloud ML engine. Job failed with the following log entry: ... Traceback (most recent call last): File "/usr/lib/python2.7/runpy.py", line 162, in _run_module_as_main "__main__", fname, loader, pkg_name) File "/usr/lib/python2.7/runpy.py", line 72, in _run_code exec code in run_globals File "/root/.local/lib/python2.7/site-packages/trainer/task.py", line 79, in <module> model.train_and_evaluate(output_dir) File "/root/.local/lib/python2.7/site-packages/trainer/model.py", line 138, in train_and_evaluate wide, deep = get_wide_deep() File "/root/.local/lib/python2.7/site-packages/trainer/model.py", line 72, in get_wide_deep tf.feature_column.categorical_column_with_vocabulary_list('is_male', AttributeError: 'module' object has no attribute 'feature_column'
While I was making the bucket, the project stopped connecting to the cloud even though my internet connection was fine. I could not listen on 8081, and there was no information on the internet in the usual places about what the errors meant. I could not go back and start over, because the instance already existed. I wanted to learn something but this was a big waste of time and credits.
Notebook code didn't work.