[
https://issues.apache.org/jira/browse/MXNET-1083?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Amol Lele updated MXNET-1083:
-----------------------------
Description:
We are planning to provide examples that demonstrate the inference workflow
using C++ API.
These examples will demonstrate:
# How to load pre-trained network and its parameters.
# How to pre-process the data for validation/inference.
# Measure the throughput and validation accuracy.
These examples will be available under separate folder. We will primarily focus
on CiFar10 dataset. We will use models that are available at
http://data.mxnet.io/models/
In order to be consistent with python examples, the cpp-package examples will
demonstrate usage of
# 'imagenet1k-inception-bn'
# 'imagenet1k-resnet-18'
# 'imagenet1k-resnet-34'
# 'imagenet1k-resnet-50'
# 'imagenet1k-resnet-101'
# 'imagenet1k-resnet-152'
# 'imagenet1k-resnext-50'
# 'imagenet1k-resnext-101'
# 'imagenet1k-resnext-101-64x4d'
# 'imagenet11k-resnet-152'
# 'imagenet11k-place365ch-resnet-152'
# 'imagenet11k-place365ch-resnet-50'
> Write examples to demonstrate the inference workflow using C++ API.
> -------------------------------------------------------------------
>
> Key: MXNET-1083
> URL: https://issues.apache.org/jira/browse/MXNET-1083
> Project: Apache MXNet
> Issue Type: Story
> Components: Apache MXNet C/C++ API
> Reporter: Amol Lele
> Assignee: Amol Lele
> Priority: Major
>
> We are planning to provide examples that demonstrate the inference workflow
> using C++ API.
> These examples will demonstrate:
> # How to load pre-trained network and its parameters.
> # How to pre-process the data for validation/inference.
> # Measure the throughput and validation accuracy.
>
> These examples will be available under separate folder. We will primarily
> focus on CiFar10 dataset. We will use models that are available at
> http://data.mxnet.io/models/
> In order to be consistent with python examples, the cpp-package examples will
> demonstrate usage of
> # 'imagenet1k-inception-bn'
> # 'imagenet1k-resnet-18'
> # 'imagenet1k-resnet-34'
> # 'imagenet1k-resnet-50'
> # 'imagenet1k-resnet-101'
> # 'imagenet1k-resnet-152'
> # 'imagenet1k-resnext-50'
> # 'imagenet1k-resnext-101'
> # 'imagenet1k-resnext-101-64x4d'
> # 'imagenet11k-resnet-152'
> # 'imagenet11k-place365ch-resnet-152'
> # 'imagenet11k-place365ch-resnet-50'
>
>
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@mxnet.apache.org
For additional commands, e-mail: issues-h...@mxnet.apache.org
