You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@beam.apache.org by OrielResearch Eila Arich-Landkof <ei...@orielresearch.org> on 2018/10/19 16:38:34 UTC
using tfma / ModelAnalysis with tensorflow (not estimator) model
Hello all,
I would like to use the modelAnalysis API for model debugging. I don't have
a background in model serving and generating the model eval graph for it -
so there might be basic background that I am missing.
as a start, I would like to add the tfma to this colab (open to other
suggestion that includes transfer learning)
Transfer Learning with TensorFlow
https://colab.research.google.com/github/tensorflow/hub/blob/master/examples/colab/image_feature_vector.ipynb
Could someone direct me, where should the graph eval generation call added?
The code below is for estimator model (taken from medium
<https://medium.com/tensorflow/introducing-tensorflow-model-analysis-scaleable-sliced-and-full-pass-metrics-5cde7baf0b7b>)
and, to my understanding, is not relevant for the colab implementation
# use TFMA to export an eval graph from the TensorFlow Estimator
tfma.export.export_eval_savedmodel(estimator=estimator,
eval_input_receiver_fn=eval_input_fn, …)
The colab requires python3 - I will make that the required code adaptation
will be applied to match apache beam python 2 requirement.
Thanks for any help,
--
Eila
www.orielresearch.org
https://www.meetu <https://www.meetup.com/Deep-Learning-In-Production/>p.co
<https://www.meetup.com/Deep-Learning-In-Production/>
m/Deep-Learning-In-Production/
<https://www.meetup.com/Deep-Learning-In-Production/>