view release on metacpan or  search on metacpan

lib/AI/MXNet/Visualization.pm  view on Meta::CPAN

package AI::MXNet::Visualization;
use strict;
use warnings;
use AI::MXNet::Base;
use AI::MXNet::Function::Parameters;
use JSON::PP;

=encoding UTF-8

=head1 NAME

    AI::MXNet::Vizualization - Vizualization support for Perl interface to MXNet machine learning library

=head1 SYNOPSIS

    use strict;
    use warnings;
    use AI::MXNet qw(mx);

    ### model
    my $data = mx->symbol->Variable('data');
    my $conv1= mx->symbol->Convolution(data => $data, name => 'conv1', num_filter => 32, kernel => [3,3], stride => [2,2]);
    my $bn1  = mx->symbol->BatchNorm(data => $conv1, name => "bn1");
    my $act1 = mx->symbol->Activation(data => $bn1, name => 'relu1', act_type => "relu");
    my $mp1  = mx->symbol->Pooling(data => $act1, name => 'mp1', kernel => [2,2], stride =>[2,2], pool_type=>'max');

    my $conv2= mx->symbol->Convolution(data => $mp1, name => 'conv2', num_filter => 32, kernel=>[3,3], stride=>[2,2]);
    my $bn2  = mx->symbol->BatchNorm(data => $conv2, name=>"bn2");
    my $act2 = mx->symbol->Activation(data => $bn2, name=>'relu2', act_type=>"relu");
    my $mp2  = mx->symbol->Pooling(data => $act2, name => 'mp2', kernel=>[2,2], stride=>[2,2], pool_type=>'max');


    my $fl   = mx->symbol->Flatten(data => $mp2, name=>"flatten");
    my $fc1  = mx->symbol->FullyConnected(data => $fl,  name=>"fc1", num_hidden=>30);
    my $act3 = mx->symbol->Activation(data => $fc1, name=>'relu3', act_type=>"relu");
    my $fc2  = mx->symbol->FullyConnected(data => $act3, name=>'fc2', num_hidden=>10);
    my $softmax = mx->symbol->SoftmaxOutput(data => $fc2, name => 'softmax');

    ## creates the image file working directory
    mx->viz->plot_network($softmax, save_format => 'png')->render("network.png"); 

=head1 DESCRIPTION

     Vizualization support for Perl interface to MXNet machine learning library

=head1 Class methods

=head2 print_summary

    convert symbol for detail information

    Parameters
    ----------
    symbol: AI::MXNet::Symbol
        symbol to be visualized
    shape: hashref
        hashref of shapes, str->shape (arrayref[int]), given input shapes
    line_length: int
        total length of printed lines
    positions: arrayref[float]
        relative or absolute positions of log elements in each line
    Returns
    ------
        nothing
=cut

method print_summary(
    AI::MXNet::Symbol        $symbol,
    Maybe[HashRef[Shape]]    $shape=,
    Int                      $line_length=120,
    ArrayRef[Num]            $positions=[.44, .64, .74, 1]
)
{
    my $show_shape;
    my %shape_dict;
    if(defined $shape)
    {
        $show_shape = 1;
        my $interals = $symbol->get_internals;
        my (undef, $out_shapes, undef) = $interals->infer_shape(%{ $shape });
        Carp::confess("Input shape is incomplete")
            unless defined $out_shapes;
        @shape_dict{ @{ $interals->list_outputs } } = @{ $out_shapes };
    }
    my $conf = decode_json($symbol->tojson);
    my $nodes = $conf->{nodes};
    my %heads = map { $_ => 1 } @{ $conf->{heads}[0] };
    if($positions->[-1] <= 1)
    {
        $positions = [map { int($line_length * $_) } @{ $positions }];
    }
    # header names for the different log elements
    my $to_display = ['Layer (type)', 'Output Shape', 'Param #', 'Previous Layer'];
    my $print_row = sub { my ($fields, $positions) = @_;
        my $line = '';
        enumerate(sub {
            my ($i, $field) = @_;
            $line .= $field//'';
            $line = substr($line, 0, $positions->[$i]);
            $line .= ' ' x ($positions->[$i] - length($line));

        }, $fields);
        print $line,"\n";
    };
    print('_' x $line_length,"\n");
    $print_row->($to_display, $positions);
    print('=' x $line_length,"\n");
    my $print_layer_summary = sub { my ($node, $out_shape) = @_;
        my $op = $node->{op};
        my $pre_node = [];
        my $pre_filter = 0;
        if($op ne 'null')
        {
            my $inputs = $node->{inputs};
            for my $item (@{ $inputs })
            {
                my $input_node = $nodes->[$item->[0]];
                my $input_name = $input_node->{name};
                if($input_node->{op} ne 'null' or exists $heads{ $item->[0] })
                {
                    push @{ $pre_node }, $input_name;
                    if($show_shape)
                    {
                        my $key = $input_name;
                        $key .= '_output' if $input_node->{op} ne 'null';
                        if(exists $shape_dict{ $key })
                        {
                            $pre_filter = $pre_filter + int($shape_dict{$key}[1]//0);
                        }
                    }
                }
            }
        }
        my $cur_param = 0;
        if($op eq 'Convolution')
        {
            my $num_filter = $node->{attr}{num_filter};
            $cur_param = $pre_filter * $num_filter;
            while($node->{attr}{kernel} =~ /(\d+)/g)
            {
                $cur_param *= $1;
            }
            $cur_param += $num_filter;
        }
        elsif($op eq 'FullyConnected')
        {
            $cur_param = $pre_filter * ($node->{attr}{num_hidden} + 1);
        }
        elsif($op eq 'BatchNorm')
        {
            my $key = "$node->{name}_output";
            if($show_shape)