Profile Download

1. Lara Magenta Mac Os Download
2. Lara Magenta Mac Os 11

CupsInteger16 CUPS 1.2/Mac OS X 10.5 User-defined integer values cupsMarkerType64 CUPS 1.2/Mac OS X 10.5 Ink/toner type cupsMediaType Media type code cupsNumColors CUPS 1.2/Mac OS X 10.5 Number of color compoents cupsPageSizeName64 CUPS 1.2/Mac OS X 10.5 PageSize name cupsPageSize2 CUPS 1.2/Mac OS X 10.5. Magenta is a research project exploring the role of machine learning in the process of creating art and music. Primarily this involves developing new deep learning and reinforcement learning algorithms for generating songs, images, drawings, and other materials. If you looking to play it on your Mac (I recommend that you do!!!) just make sure it’s compatible with your particular machine. My iMac is about 3 yrs. Old, OS 10.9.4 no Retina display, but excellent graphics just the same. I highly recommend getting a gamepad if you don’t already have one, we have the Sony for PS3 and it works great!!!

Lara Magenta Mac Os Download

Overview

The CUPS raster API provides a standard interface for reading and writingCUPS raster streams which are used for printing to raster printers. Because theraster format is updated from time to time, it is important to use this API toavoid incompatibilities with newer versions of CUPS.

Two kinds of CUPS filters use the CUPS raster API - raster image processor(RIP) filters such as pstoraster and cgpdftoraster(Mac OS X) that produce CUPS raster files and printer driver filters thatconvert CUPS raster files into a format usable by the printer. Printerdriver filters are by far the most common.

CUPS raster files (application/vnd.cups-raster) consists ofa stream of raster page descriptions produced by one of the RIP filters such aspstoraster, imagetoraster, orcgpdftoraster. CUPS raster files are referred to using thecups_raster_t type and areopened using the cupsRasterOpenfunction. For example, to read raster data from the standard input, openfile descriptor 0:

Each page of data begins with a page dictionary structure calledcups_page_header2_t. Thisstructure contains the colorspace, bits per color, media size, media type,hardware resolution, and so forth used for the page.

Do not confuse the colorspace in the page header with the PPD ColorModel keyword. ColorModel refers to the general type of color used for a device (Gray, RGB, CMYK, DeviceN) and is often used to select a particular colorspace for the page header along with the associate color profile. The page header colorspace (cupsColorSpace) describes both the type and organization of the color data, for example KCMY (black first) instead of CMYK and RGBA (RGB + alpha) instead of RGB.

You read the page header using thecupsRasterReadHeader2function:

After the page dictionary comes the page data which is a full-resolution,possibly compressed bitmap representing the page in the printer's outputcolorspace. You read uncompressed raster data using thecupsRasterReadPixelsfunction. A for loop is normally used to read the page one lineat a time:

When you are done reading the raster data, call thecupsRasterClose function to freethe memory used to read the raster file:

Functions by Task

Opening and Closing Raster Streams

Lara Magenta Mac Os 11

Reading Raster Streams

• cupsRasterReadHeaderDeprecated in CUPS 1.2/Mac OS X 10.5

Writing Raster Streams

• cupsRasterWriteHeaderDeprecated in CUPS 1.2/Mac OS X 10.5

Functions

cupsRasterClose

Close a raster stream.

void cupsRasterClose (
cups_raster_t *r
);

Parameters

r

Stream to close

Discussion

The file descriptor associated with the raster stream must be closedseparately as needed.

CUPS 1.2/Mac OS X 10.5 cupsRasterInterpretPPD

Interpret PPD commands to create a page header.

int cupsRasterInterpretPPD (
cups_page_header2_t *h,
ppd_file_t *ppd,
int num_options,
cups_option_t *options,
cups_interpret_cb_t func
);

Parameters

h

Page header to create

ppd

PPD file

num_options

Number of options

options

Options

func

Optional page header callback (NULL for none)

Return Value

0 on success, -1 on failure

Discussion

This function is used by raster image processing (RIP) filters likecgpdftoraster and imagetoraster when writing CUPS raster data for a page.It is not used by raster printer driver filters which only read CUPSraster data.
cupsRasterInterpretPPD does not mark the options in the PPD usingthe 'num_options' and 'options' arguments. Instead, mark the options withcupsMarkOptions and ppdMarkOption prior to calling it -this allows for per-page options without manipulating the options array.
The 'func' argument specifies an optional callback function that iscalled prior to the computation of the final raster data. The functioncan make changes to the cups_page_header2_t data as needed to use asupported raster format and then returns 0 on success and -1 if therequested attributes cannot be supported.
cupsRasterInterpretPPD supports a subset of the PostScript language.Currently only the [, ], <<, >>, {,}, cleartomark, copy, dup, index,pop, roll, setpagedevice, and stopped operatorsare supported.

cupsRasterOpen

Open a raster stream.

cups_raster_t *cupsRasterOpen (
int fd,
cups_mode_t mode
);

Parameters

fd

File descriptor

mode

Mode - CUPS_RASTER_READ, CUPS_RASTER_WRITE, or CUPS_RASTER_WRITE_COMPRESSED

Return Value

New stream

Discussion

This function associates a raster stream with the given file descriptor.For most printer driver filters, 'fd' will be 0 (stdin). For most rasterimage processor (RIP) filters that generate raster data, 'fd' will be 1(stdout).
When writing raster data, the CUPS_RASTER_WRITE orCUPS_RASTER_WRITE_COMPRESS mode can be used - compressed outputis generally 25-50% smaller but adds a 100-300% execution time overhead.

DEPRECATED cupsRasterReadHeader

Read a raster page header and store it in aversion 1 page header structure.

unsigned cupsRasterReadHeader (
cups_raster_t *r,
cups_page_header_t *h
);

Parameters

r

Raster stream

h

Pointer to header data

Return Value

1 on success, 0 on failure/end-of-file

Discussion

This function is deprecated. Use cupsRasterReadHeader2 instead.
Version 1 page headers were used in CUPS 1.0 and 1.1 and contain a subsetof the version 2 page header data. This function handles reading version 2page headers and copying only the version 1 data into the provided buffer.

CUPS 1.2/Mac OS X 10.5 cupsRasterReadHeader2

Read a raster page header and store it in aversion 2 page header structure.

unsigned cupsRasterReadHeader2 (
cups_raster_t *r,
cups_page_header2_t *h
);

Parameters

r

Raster stream

h

Pointer to header data

Return Value

1 on success, 0 on failure/end-of-file

cupsRasterReadPixels

Read raster pixels.

unsigned cupsRasterReadPixels (
cups_raster_t *r,
unsigned char *p,
unsigned len
);

Parameters

r

Raster stream

p

Pointer to pixel buffer

len

Number of bytes to read

Return Value

Number of bytes read

Discussion

For best performance, filters should read one or more whole lines.The 'cupsBytesPerLine' value from the page header can be used to allocatethe line buffer and as the number of bytes to read.

DEPRECATED cupsRasterWriteHeader

Write a raster page header from a version 1 pageheader structure.

unsigned cupsRasterWriteHeader (
cups_raster_t *r,
cups_page_header_t *h
);

Parameters

r

Raster stream

h

Raster page header

Return Value

1 on success, 0 on failure

Discussion

This function is deprecated. Use cupsRasterWriteHeader2 instead.

CUPS 1.2/Mac OS X 10.5 cupsRasterWriteHeader2

Write a raster page header from a version 2page header structure.

unsigned cupsRasterWriteHeader2 (
cups_raster_t *r,
cups_page_header2_t *h
);

Parameters

r

Raster stream

h

Raster page header

Return Value

1 on success, 0 on failure

Discussion

The page header can be initialized using cupsRasterInterpretPPD.

cupsRasterWritePixels

Write raster pixels.

unsigned cupsRasterWritePixels (
cups_raster_t *r,
unsigned char *p,
unsigned len
);

Parameters

r

Raster stream

p

Bytes to write

len

Number of bytes to write

Return Value

Number of bytes written

Discussion

For best performance, filters should write one or more whole lines.The 'cupsBytesPerLine' value from the page header can be used to allocatethe line buffer and as the number of bytes to write.

Data Types

cups_adv_t

AdvanceMedia attribute values

typedef enum cups_adv_e cups_adv_t;

cups_bool_t

Boolean type

typedef enum cups_bool_e cups_bool_t;

cups_cspace_t

cupsColorSpace attribute values

typedef enum cups_cspace_e cups_cspace_t;

cups_cut_t

CutMedia attribute values

typedef enum cups_cut_e cups_cut_t;

cups_edge_t

LeadingEdge attribute values

typedef enum cups_edge_e cups_edge_t;

cups_interpret_cb_t

cupsRasterInterpretPPD callback function

typedef int (*cups_interpret_cb_t)(cups_page_header2_t *header, int preferred_bits);

cups_jog_t

Jog attribute values

typedef enum cups_jog_e cups_jog_t;

cups_mode_t

cupsRasterOpen modes

typedef enum cups_mode_e cups_mode_t;

cups_order_t

cupsColorOrder attribute values

typedef enum cups_order_e cups_order_t;

cups_orient_t

Orientation attribute values

typedef enum cups_orient_e cups_orient_t;

CUPS 1.2/Mac OS X 10.5 cups_page_header2_t

Version 2 page header

typedef struct cups_page_header2_s cups_page_header2_t;

DEPRECATED cups_page_header_t

Version 1 page header

typedef struct cups_page_header_s cups_page_header_t;

cups_raster_t

Raster stream data

typedef struct _cups_raster_s cups_raster_t;

Structures

CUPS 1.2/Mac OS X 10.5 cups_page_header2_s

Version 2 page header

struct cups_page_header2_s {
unsigned AdvanceDistance;
cups_adv_t AdvanceMedia;
cups_bool_t Collate;
cups_cut_t CutMedia;
cups_bool_t Duplex;
unsigned HWResolution[2];
unsigned ImagingBoundingBox[4];
cups_bool_t InsertSheet;
cups_jog_t Jog;
cups_edge_t LeadingEdge;
cups_bool_t ManualFeed;
unsigned Margins[2];
char MediaClass[64];
char MediaColor[64];
unsigned MediaPosition;
char MediaType[64];
unsigned MediaWeight;
cups_bool_t MirrorPrint;
cups_bool_t NegativePrint;
unsigned NumCopies;
cups_orient_t Orientation;
cups_bool_t OutputFaceUp;
char OutputType[64];
unsigned PageSize[2];
cups_bool_t Separations;
cups_bool_t TraySwitch;
cups_bool_t Tumble;
unsigned cupsBitsPerColor;
unsigned cupsBitsPerPixel;
float cupsBorderlessScalingFactor;
unsigned cupsBytesPerLine;
cups_order_t cupsColorOrder;
cups_cspace_t cupsColorSpace;
unsigned cupsCompression;
unsigned cupsHeight;
float cupsImagingBBox[4];
unsigned cupsInteger[16];
char cupsMarkerType[64];
unsigned cupsMediaType;
unsigned cupsNumColors;
char cupsPageSizeName[64];
float cupsPageSize[2];
float cupsReal[16];
char cupsRenderingIntent[64];
unsigned cupsRowCount;
unsigned cupsRowFeed;
unsigned cupsRowStep;
char cupsString[16][64];
unsigned cupsWidth;
};

Members

AdvanceDistance

AdvanceDistance value in points

AdvanceMedia

AdvanceMedia value (cups_adv_t)

Collate

Collated copies value

CutMedia

CutMedia value (cups_cut_t)

Duplex

Duplexed (double-sided) value

HWResolution[2]

Resolution in dots-per-inch

ImagingBoundingBox[4]

Pixel region that is painted (points, left, bottom, right, top)

InsertSheet

InsertSheet value

Jog

Jog value (cups_jog_t)

LeadingEdge

LeadingEdge value (cups_edge_t)

ManualFeed

ManualFeed value

Margins[2]

Lower-lefthand margins in points

MediaClass[64]

MediaClass string

MediaColor[64]

MediaColor string

MediaPosition

MediaPosition value

MediaType[64]

MediaType string

MediaWeight

MediaWeight value in grams/m^2

MirrorPrint

MirrorPrint value

NegativePrint

NegativePrint value

NumCopies

Number of copies to produce

Orientation

Orientation value (cups_orient_t)

OutputFaceUp

OutputFaceUp value

OutputType[64]

OutputType string

PageSize[2]

Width and length of page in points

Separations

Separations value

TraySwitch

TraySwitch value

Tumble

Tumble value

cupsBitsPerColor

Number of bits for each color

cupsBitsPerPixel

Number of bits for each pixel

cupsBorderlessScalingFactor CUPS 1.2/Mac OS X 10.5

Scaling that was applied to page data

cupsBytesPerLine

Number of bytes per line

cupsColorOrder

Order of colors

cupsColorSpace

True colorspace

cupsCompression

Device compression to use

cupsHeight

Height of page image in pixels

cupsImagingBBox[4] CUPS 1.2/Mac OS X 10.5

Floating point ImagingBoundingBox(scaling factor not applied, left,bottom, right, top)

cupsInteger[16] CUPS 1.2/Mac OS X 10.5

User-defined integer values

cupsMarkerType[64] CUPS 1.2/Mac OS X 10.5

Ink/toner type

cupsMediaType

Media type code

cupsNumColors CUPS 1.2/Mac OS X 10.5

Number of color compoents

cupsPageSizeName[64] CUPS 1.2/Mac OS X 10.5

PageSize name

cupsPageSize[2] CUPS 1.2/Mac OS X 10.5

Floating point PageSize (scaling *factor not applied)

cupsReal[16] CUPS 1.2/Mac OS X 10.5

User-defined floating-point values

cupsRenderingIntent[64] CUPS 1.2/Mac OS X 10.5

Color rendering intent

cupsRowCount

Rows per band

cupsRowFeed

Feed between bands

cupsRowStep

Spacing between lines

cupsString[16][64] CUPS 1.2/Mac OS X 10.5

User-defined string values

cupsWidth

Width of page image in pixels

DEPRECATED cups_page_header_s

Version 1 page header

struct cups_page_header_s {
unsigned AdvanceDistance;
cups_adv_t AdvanceMedia;
cups_bool_t Collate;
cups_cut_t CutMedia;
cups_bool_t Duplex;
unsigned HWResolution[2];
unsigned ImagingBoundingBox[4];
cups_bool_t InsertSheet;
cups_jog_t Jog;
cups_edge_t LeadingEdge;
cups_bool_t ManualFeed;
unsigned Margins[2];
char MediaClass[64];
char MediaColor[64];
unsigned MediaPosition;
char MediaType[64];
unsigned MediaWeight;
cups_bool_t MirrorPrint;
cups_bool_t NegativePrint;
unsigned NumCopies;
cups_orient_t Orientation;
cups_bool_t OutputFaceUp;
char OutputType[64];
unsigned PageSize[2];
cups_bool_t Separations;
cups_bool_t TraySwitch;
cups_bool_t Tumble;
unsigned cupsBitsPerColor;
unsigned cupsBitsPerPixel;
unsigned cupsBytesPerLine;
cups_order_t cupsColorOrder;
cups_cspace_t cupsColorSpace;
unsigned cupsCompression;
unsigned cupsHeight;
unsigned cupsMediaType;
unsigned cupsRowCount;
unsigned cupsRowFeed;
unsigned cupsRowStep;
unsigned cupsWidth;
};

Members

AdvanceDistance

AdvanceDistance value in points

AdvanceMedia

AdvanceMedia value (cups_adv_t)

Collate

Collated copies value

CutMedia

CutMedia value (cups_cut_t)

Duplex

Duplexed (double-sided) value

HWResolution[2]

Resolution in dots-per-inch

ImagingBoundingBox[4]

Pixel region that is painted (points, left, bottom, right, top)

InsertSheet

InsertSheet value

Jog

Jog value (cups_jog_t)

LeadingEdge

LeadingEdge value (cups_edge_t)

ManualFeed

ManualFeed value

Margins[2]

Lower-lefthand margins in points

MediaClass[64]

MediaClass string

MediaColor[64]

MediaColor string

MediaPosition

MediaPosition value

MediaType[64]

MediaType string

MediaWeight

MediaWeight value in grams/m^2

MirrorPrint

MirrorPrint value

NegativePrint

NegativePrint value

NumCopies

Number of copies to produce

Orientation

Orientation value (cups_orient_t)

OutputFaceUp

OutputFaceUp value

OutputType[64]

OutputType string

PageSize[2]

Width and length of page in points

Separations

Separations value

TraySwitch

TraySwitch value

Tumble

Tumble value

cupsBitsPerColor

Number of bits for each color

cupsBitsPerPixel

Number of bits for each pixel

cupsBytesPerLine

Number of bytes per line

cupsColorOrder

Order of colors

cupsColorSpace

True colorspace

cupsCompression

Device compression to use

cupsHeight

Height of page image in pixels

cupsMediaType

Media type code

cupsRowCount

Rows per band

cupsRowFeed

Feed between bands

cupsRowStep

Spacing between lines

cupsWidth

Width of page image in pixels

Constants

cups_adv_e

AdvanceMedia attribute values

Constants

CUPS_ADVANCE_FILE

Advance the roll after this file

CUPS_ADVANCE_JOB

Advance the roll after this job

CUPS_ADVANCE_NONE

Never advance the roll

CUPS_ADVANCE_PAGE

Advance the roll after this page

CUPS_ADVANCE_SET

Advance the roll after this set

cups_bool_e

Boolean type

Constants

CUPS_FALSE

Logical false

CUPS_TRUE

Logical true

cups_cspace_e

cupsColorSpace attribute values

Constants

CUPS_CSPACE_CIELab CUPS 1.1.19/Mac OS X 10.3

CIE Lab

CUPS_CSPACE_CIEXYZ CUPS 1.1.19/Mac OS X 10.3

CIE XYZ

CUPS_CSPACE_CMY

Cyan, magenta, yellow

CUPS_CSPACE_CMYK

Cyan, magenta, yellow, black

CUPS_CSPACE_GMCK

Gold, magenta, yellow, black

CUPS_CSPACE_GMCS

Gold, magenta, yellow, silver

CUPS_CSPACE_GOLD

Gold foil

CUPS_CSPACE_ICC1 CUPS 1.1.19/Mac OS X 10.3

ICC-based, 1 color

CUPS_CSPACE_ICC2 CUPS 1.1.19/Mac OS X 10.3

ICC-based, 2 colors

CUPS_CSPACE_ICC3 CUPS 1.1.19/Mac OS X 10.3

ICC-based, 3 colors

CUPS_CSPACE_ICC4 CUPS 1.1.19/Mac OS X 10.3

ICC-based, 4 colors

CUPS_CSPACE_ICC5 CUPS 1.1.19/Mac OS X 10.3

ICC-based, 5 colors

CUPS_CSPACE_ICC6 CUPS 1.1.19/Mac OS X 10.3

ICC-based, 6 colors

CUPS_CSPACE_ICC7 CUPS 1.1.19/Mac OS X 10.3

ICC-based, 7 colors

CUPS_CSPACE_ICC8 CUPS 1.1.19/Mac OS X 10.3

ICC-based, 8 colors

CUPS_CSPACE_ICC9 CUPS 1.1.19/Mac OS X 10.3

ICC-based, 9 colors

CUPS_CSPACE_ICCA CUPS 1.1.19/Mac OS X 10.3

ICC-based, 10 colors

CUPS_CSPACE_ICCB CUPS 1.1.19/Mac OS X 10.3

ICC-based, 11 colors

CUPS_CSPACE_ICCC CUPS 1.1.19/Mac OS X 10.3

ICC-based, 12 colors

CUPS_CSPACE_ICCD CUPS 1.1.19/Mac OS X 10.3

ICC-based, 13 colors

CUPS_CSPACE_ICCE CUPS 1.1.19/Mac OS X 10.3

ICC-based, 14 colors

CUPS_CSPACE_ICCF CUPS 1.1.19/Mac OS X 10.3

ICC-based, 15 colors

CUPS_CSPACE_K

Black

CUPS_CSPACE_KCMY

Black, cyan, magenta, yellow

CUPS_CSPACE_KCMYcm

Black, cyan, magenta, yellow, *light-cyan, light-magenta

CUPS_CSPACE_RGB

Red, green, blue

CUPS_CSPACE_RGBA

Red, green, blue, alpha

CUPS_CSPACE_RGBW CUPS 1.2/Mac OS X 10.5

Red, green, blue, white

CUPS_CSPACE_SILVER

Silver foil

CUPS_CSPACE_W

Luminance

CUPS_CSPACE_WHITE

White ink (as black)

CUPS_CSPACE_YMC

Yellow, magenta, cyan

CUPS_CSPACE_YMCK

Yellow, magenta, cyan, black

cups_cut_e

CutMedia attribute values

Constants

CUPS_CUT_FILE

Cut the roll after this file

CUPS_CUT_JOB

Cut the roll after this job

CUPS_CUT_NONE

Never cut the roll

CUPS_CUT_PAGE

Cut the roll after this page

CUPS_CUT_SET

Cut the roll after this set

cups_edge_e

LeadingEdge attribute values

Constants

CUPS_EDGE_BOTTOM

Leading edge is the bottom of the page

CUPS_EDGE_LEFT

Leading edge is the left of the page

CUPS_EDGE_RIGHT

Leading edge is the right of the page

CUPS_EDGE_TOP

Leading edge is the top of the page

cups_jog_e

Jog attribute values

Constants

CUPS_JOG_FILE

Move pages after this file

CUPS_JOG_JOB

Move pages after this job

CUPS_JOG_NONE

Never move pages

CUPS_JOG_SET

Move pages after this set

cups_mode_e

cupsRasterOpen modes

Constants

CUPS_RASTER_READ

Open stream for reading

CUPS_RASTER_WRITE

Open stream for writing

CUPS_RASTER_WRITE_COMPRESSED CUPS 1.3/Mac OS X 10.5

Open stream for compressed writing

cups_order_e

cupsColorOrder attribute values

Constants

CUPS_ORDER_BANDED

CCC MMM YYY KKK ...

CUPS_ORDER_CHUNKED

CMYK CMYK CMYK ...

CUPS_ORDER_PLANAR

CCC ... MMM ... YYY ... KKK ...

cups_orient_e

Orientation attribute values

Constants

CUPS_ORIENT_0

Don't rotate the page

CUPS_ORIENT_180

Turn the page upside down

CUPS_ORIENT_270

Rotate the page clockwise

CUPS_ORIENT_90

Rotate the page counter-clockwise

Machine Learning is all the rage these days, and with open source frameworks like TensorFlow developers have access to a range of APIs for using machine learning in their projects. Magenta, a Python library built by the TensorFlow team, makes it easier to process music and image data in particular.

Since I started learning how to code, one of the things that has always fascinated me was the concept of computers artificially creating music. I even published a paper talking about it in an undergrad research journal my freshman year of college.

Let's walk through the basics of setting up Magenta and programmatically generating some simple melodies in MIDI file format.

Installing Magenta

First we need to install Magenta, which can be done using pip. Make sure you create a virtual environment before installing. I am using Python 3.6.5, but Magenta is compatible with both Python 2 and 3.

Run the following command to install Magenta in your virtual environment, it's a pretty big library with a good amount of dependencies so it might take a bit of time:

Alternatively, if you want to install Magenta globally you can use the following shell commands to run an install script created by the Magenta team to simplify things:

This will give you access to both the Magenta and TensorFlow Python modules for development, as well as scripts to work with all of the models that Magenta has available. For this post, we're going to be using the Melody recurrent neural network model.

Generating a basic melody in MIDI format

Rather than training our own model, let's use one of the pre-trained melody models provided by the TensorFlow team.

First, download this file, which is a .mag bundle file for a recurrent neural network that has been trained on thousands of MIDI files. We're going to use this as a starting point to generate some melodies. Save it to the current directory you are working in.

When generating a melody, we have to provide a priming melody. This can be a MIDI file using the --prime_midi flag, or a format that Magenta uses which is a string representation of a Python list using the --prime_melody flag. Let's create some melodies using middle C as the starting note, which in this format would be '[60]'.

Each melody will be 8 measures in length, corresponding to the --num_steps flag. This refers to how many 16th step durations the generated tune will be.

With your virtual environment activated, run the following command, making sure to replace /path/to/basic_rnn.mag with an actual path to the .mag file you just downloaded:

This should output 10 MIDI files in the directory /tmp/melody_rnn/generated, or whichever directory you want using the --output_dir flag. It will take some time to execute, so be patient!

Navigate to the output directory, and try playing the MIDI files to see what kind of music you just created! If you are on Mac OS X, the GarageBand program can play MIDI files.

Here's an example of a melody that was generated when I ran this code:

Using different models for more structured compositions

Those melodies are cool for the novelty of a machine composing music, but to me it still sounds mostly like a bunch of random notes. The Magenta team provides two other pre-trained models we can use to generate melodies that have more structure.

The previous model worked by generating notes one by one, only keeping track of the most recent note. That's why a lot of the melodies sound all over the place. Among other things, this Lookback RNN keeps track of the most recent two bars, so it is able to add more repetition into the music.

Download the Lookback RNN model, and save it to the same directory you saved basic_rnn.mag.

Let's generate some melodies using the Lookback RNN, remembering to replace /path/to/lookback_rnn.mag with an actual path to the .mag file you downloaded:

You will likely notice that the melodies you generate with this one have a lot more repetition. Here's one of the ones I got:

Now let's try out the Attention RNN. Instead of just looking back at the last two measures, this one is designed to give more long term structure in generated compositions. You can read about the algorithm in this blog post.

Again download the model, and save it to the right directory, and then run the following:

For this example, we are generating melodies that are twice as long. One of the melodies I generated seems to even have a structure that could be repeated. To me, this one sounds like it could be turned into a long form song, complete with different sections that flow into each other.

Using a MIDI file as a priming melody

In the previous examples we have only used a single note, middle C, as the priming melody. But it's much more interesting to create music from an already existing melody that a human wrote. Let's use a MIDI file with a more complex melody that was written by a human to create a musical collaboration between man and machine.

I'm a guitarist and would like to hear a computer shred. So for this example I'm going to use the guitar solo from Omens of Love by the Japanese fusion band T-Square. We'll use the first four measures of this solo, which provide a nice melodic start, and see if we can generate four more measures to complement it.

Download this MIDI file containing a 4-bar melody, and save it to a directory of your choice.

Now use whichever model you want from the previous sections to create a computational jam session! I am going to use the Attention RNN because I liked some of the results I got before:

You might have to generate a ton of output melodies to get something that sounds human, but out of the 10 I generated, this one works really nicely!

Awesome! What else can I do?

We were able to generate some simple melodies with some pre-trained neural network models, and that's awesome! We're already off to a great start when it comes to using machines to create music.

It's a lot of fun to feed different MIDIs to a neural network model and see what comes out. Magenta offers a whole variety of models to work with, and in this post we've only covered the first steps to working with the Melody RNN model.

Keep an eye out for future Twilio blog posts on working with music data using Magenta, including how to train your own models.

Feel free to reach out for any questions or to show off any cool artificial creativity related projects you build or find out about:

• Email: Sagnew@twilio.com
• Twitter: @Sagnewshreds
• GitHub: Sagnew
• Twitch (streaming live code): Sagnewshreds