Hey, what's this?! A chatbot?

Jade is really simple to get up and running. Just get some data, such as these anime subtitles (ahem shameless plug), rename it to inputcln.txt, move it to the same directory as the .py file, and start the program with the -train flag.

BAM, it'll eat all your RAM and CPU until suddenly you have a 2-3 gb model (.p) file. Run without the -train flag for faster, more nyoom speeds.

Add Jade at https://discordbots.org/bot/410253782828449802

ALL BELOW IS OUTDATED, AND IS NO LONGER SUPPORTED

JadeV4 Usage

Jade's Prefix is JD, and is used to trigger all possible commands through NLP. Prefix is not required in DMs.

Example syntax:

Contextual Generative Conversation

Query:

Hi!
How old are you?

Response:

Hi, I don't know you!
I am 9 years old.

Jade can do basic math, with spaces as delimiters. Some basic hardcoded functions also exist.

Neural Style

Query:

Can you style this using un? [Attach Image]
Use exp to style [Image Link]

[Attach Image] and [Image Link] would reference This image in this demo. Response: Jade can extract style types (flags) such as exp and un and links from within the command phrase.

Listed below are flags, in the format FLAG ||| NAME OF STYLE IMAGE. Left images are the original, Right images are styled by Jade. un ||| Udnie dk ||| Dark Paint en ||| Enviornment rd||| Red lm ||| La Muse rp ||| Rainbow Princess sc ||| The Scream wr ||| Wreck (George Washington) wv ||| Wave ha1 ||| My artist's art ^-^ Works best with images that have a "focus" exp ||| Experimental WARNING: Most models in this folder are really... uh... weird

Object Recognition

Query:

What objects are in this image? [Attach Image]
Can you find the dog in this picture? [Image Link]

[Attach Image] would reference This image in this demo. [Image Link] would reference This image in this demo. Response:

Optical Character Recognition

Query:

What text is in this image? [Attach Image]
Could read this for me? [Image Link]

[Attach Image] would reference [This image](UPLOAD PENDING) in this demo. [Image Link] would reference [This image](UPLOAD PENDING) in this demo. Response:

Face Recognition

Query:

Who is this? [Attach Image]
Do know who the people in this image: [Image Link] are?

[Attach Image] and [Image Link] would reference This image in this demo. Response: Justin Timberlake Note: This operation works with multiple faces in an image.

DO NOT ADD FACES IF NOT NESSISARY. ONLY ADD FACES IF THE ORIGINAL IMAGE RETURNS "Unknown" Query:

Add "Justin Timberlake" using this: [Image Link]
This is "Justin Timberlake" [Attach Image]

The "" are REQUIRED around names to be added. [Attach Image] and [Image Link] would reference This image in this demo. Response: Added face "Justin Timberlake" to my library in 22742ms