drums_rnn_generate: command not found

217 views
Skip to first unread message

田旺

unread,
Feb 26, 2017, 4:46:53 AM2/26/17
to Magenta Discuss

when I tried to use magenta models——"Drums RNN" to generate my drum tracks, I got this info:

drums_rnn_generate: command not found

and here is my .sh file

BUNDLE_PATH='/root/magenta/model/drum_kit_rnn.mag'
CONFIG='drum_kit'

drums_rnn_generate \
--config=${CONFIG} \
--bundle_file=${BUNDLE_PATH} \
--output_dir=/root/magenta/drums_rnn/generated \
--num_outputs=10 \
--num_steps=128 \
--primer_drums="[(36,)]"

I have installed magenta environment successfully,and activated magenta. it's strange that I can use command melody_rnn_generate which is from "melody_rnn" without any problem. Does someone get the same problem like mine? I want to know where the problem is and how to sovle it. Thanks anyway.

Ian Simon

unread,
Feb 26, 2017, 9:29:39 AM2/26/17
to 田旺, Magenta Discuss
Drums RNN is not part of the pip package just yet; it will be included in our next release.  You can use it in our development environment: https://github.com/tensorflow/magenta#development-environment

-Ian

--
Magenta project: magenta.tensorflow.org
To post to this group, send email to magenta...@tensorflow.org
To unsubscribe from this group, send email to magenta-discu...@tensorflow.org
---
You received this message because you are subscribed to the Google Groups "Magenta Discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to magenta-discu...@tensorflow.org.

Graham Parvinkarimi

unread,
Feb 26, 2017, 12:46:45 PM2/26/17
to Magenta Discuss, tianwa...@gmail.com
I have also been trying to use the drums_rnn model in the development environment but there is an issue with the most recent version of Bazel that is causing issues when trying to build Magenta: https://groups.google.com/forum/#!msg/bazel-discuss/NK4DuqDlY2w/5J2BmxgdAQAJ. Can anyone suggest a workaround for this or perhaps a way to build Magenta (including the drums_rnn) without Bazel for now?

-Graham

Graham Parvinkarimi

unread,
Feb 28, 2017, 12:30:31 AM2/28/17
to Magenta Discuss, tianwa...@gmail.com
The fix that worked for me was changing the line "import parser" in tensorflow/contrib/tensorboard/plugins/trace/trace.py to "import mido.parser" as suggested here: https://groups.google.com/forum/?utm_medium=email&utm_source=footer#!msg/bazel-discuss/NK4DuqDlY2w/EuwQT12SAQAJ.
Reply all
Reply to author
Forward
0 new messages