[
https://issues.apache.org/jira/browse/BEAM-7921?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16935152#comment-16935152
]
Neville Li commented on BEAM-7921:
----------------------------------
https://github.com/spotify/beam/commits/neville/tf
[~app-tarush] I have this WIP branch here but haven't got time to work on it
lately. Feel free to take over.
> Add Schema support for Tensorflow
> ---------------------------------
>
> Key: BEAM-7921
> URL: https://issues.apache.org/jira/browse/BEAM-7921
> Project: Beam
> Issue Type: Improvement
> Components: dsl-sql, io-java-tfrecord
> Affects Versions: 2.14.0
> Reporter: Neville Li
> Priority: Major
>
> Similar to BEAM-5807, Tensorflow's defacto storage format is {{TFRecord}}
> files with {{Example}} proto payload and its own schema.proto. We already
> have {{TFRecordIO}} support. Need to implement:
> * Conversion between Beam and TF schema
> * Conversion between Beam Row and TF Example proto
> * {{TFRecordTableProvider}}
> [https://github.com/tensorflow/metadata/blob/master/tensorflow_metadata/proto/v0/schema.proto]
> [https://github.com/tensorflow/tensorflow/blob/master/tensorflow/core/example/example.proto]
>
> Also it seems the metadata protos are not published as Java artifacts:
> [https://github.com/tensorflow/metadata/issues/5]
> My WIP branch: https://github.com/spotify/beam/tree/neville/tf
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
