Project name

Project

Choose a name for your project.

project name

Project type

Choose custom or predefined project

predefined model - choose model from predefined quantization configurations (basic)

custom model - create new quantization configuration using wizard (advanced)

Select model

Predefined models

Choose input model from predefined examples.

Custom model

Choose input model that you would like to use.

model path
Set input & output nodes * Inputs* {{ getBoundaryNodes('inputs') }} Outputs* {{ getBoundaryNodes('outputs') }}

graph input

{{item}}

inputs to choose

{{item}}

graph input

graph input

Add custom input nodes

graph output

{{item}}

outputs to choose

{{item}}

graph output

graph output

Add custom output nodes

Order of inputs and outputs matters.

Model domain

Model domain helps to set up basic parameters for the model.

model domain NLP {{ domain.name | underscoreToSpace | titlecase }}

Dummy dataset

This is artificial dataset that can be used for fast experiments.

shape

Neural Coder will be used for PyTorch script-based model optimization and benchmark.