Tenzokarton + Keras + ML Engine
I currently have the Google Cloud ML Engine installed to train Keras-built models. When using Keras, it seems like ML Engine does not automatically save logs to the storage bucket. I can see the logs on the ML Engine Jobs page, but they do not appear in my storage and therefore I cannot run the tensogram during training.
You can see the job completed successfully and logs generated: 
But then there are no logs in my repository: 
I followed this guide while setting up my environment: ( http://liufuyang.github.io/2017/04/02/just-another-tensorflow-beginner-guide-4.html )
So how do I get the logs and run a tensogram when training a Keras model on ML Engine? Has anyone else had any success with this?
source to share
You will need to create a keras.callbacks.TensorBoard (..) callback to write the logs. See Tensorboad Callback . You can also give the GCS path (gs: // path / to / my / logs) to the log_dir argument of the callback, and then point Tensorboard to that location. You will add a callback to the list when you call model.fit_generator (...) or model.fit (...).
tb_logs = callbacks.TensorBoard(
log_dir='gs://path/to/logs',
histogram_freq=0,
write_graph=True,
embeddings_freq=0)
model.fit_generator(..., callbacks=[tb_logs])
source to share