You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@mxnet.apache.org by "Piyush Ghai (JIRA)" <ji...@apache.org> on 2019/01/02 22:55:00 UTC
[jira] [Commented] (MXNET-1286) Train a Float64 model in Python in
order to verify MXNET-1278
[ https://issues.apache.org/jira/browse/MXNET-1286?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16732495#comment-16732495 ]
Piyush Ghai commented on MXNET-1286:
------------------------------------
An easy workaround is to take any pre-trained model, and use the Cast operator of MXNet to cast it to a different DType.
Example snippet : [https://github.com/apache/incubator-mxnet/pull/12412/files#diff-33e25bba6c65aa67b8d00cb07bd7fc4cR206]
> Train a Float64 model in Python in order to verify MXNET-1278
> -------------------------------------------------------------
>
> Key: MXNET-1286
> URL: https://issues.apache.org/jira/browse/MXNET-1286
> Project: Apache MXNet
> Issue Type: Sub-task
> Components: Apache MXNet Scala API
> Reporter: Piyush Ghai
> Assignee: Piyush Ghai
> Priority: Minor
>
> To load a pre-trained model, we need to access a model or try to train a simple model using Float64 as the input types.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@mxnet.apache.org
For additional commands, e-mail: issues-help@mxnet.apache.org