Demonstration tutorial of retraining OpenAI’s GPT-2-small (a text-generating Transformer neural network) on a large public domain Project Gutenberg poetry corpus to generate high-quality English verse.
https://jalammar.github.io/illustrated-gpt2/
Other tutorial : https://medium.com/@ngwaifoong92/beginners-guide-to-retrain-gpt-2-117m-to-generate-custom-text-content-8bb5363d8b7f
https://github.com/minimaxir/gpt-2-simple
Example : http://textsynth.org/
Datasets :
https://www.kaggle.com/datasets
https://github.com/awesomedata/awesome-public-datasets
Scrap webpage with python :
https://www.crummy.com/software/BeautifulSoup/
https://github.com/EugenHotaj/beatles/blob/master/scraper.py
Built by Adam King (@AdamDanielKing) as an easier way to play with OpenAI's new machine learning model. In February, OpenAI unveiled a language model called GPT-2 that generates coherent paragraphs of text one word at a time.
AI research that touches on dialogue and story generation. As before, I’m picking a few points of interest, summarizing highlights, and then linking through to the detailed research.
This one is about a couple of areas of natural language processing and generation, as well as sentiment understanding, relevant to how we might realize stories and dialogue with particular surface features and characteristics.
We call them "seeds". Each seed is a machine learning example you can start playing with. Explore, learn and grow them into whatever you like.
A Physical Book makes a digitized book “physical” by rendering it in a simulated space where properties like gravity, friction, and velocity all apply. The program randomly perturbs the letters, then takes a snapshot at a point in time, re-assembling the images into a new, “un-digitized” book.
The raw, uncorrected scanned text of The Up-To-Date Sandwich Book: 400 Ways to Make a Sandwich (1909) is re-imagined as this 251 page (50,964 words) book:
A Physical Book uses the web-based game engine Phaser. Each page of the book is rendered into an invisible <div>
(to produce correct leading and line-height), then copied into the Phaser game world with each letter instantiated as a distinct addressable sprite.
For each page, one of a dozen transformations is applied to give the text varying physical properties, such as mass, acceleration, collision, or opacity.
On page load, the requested chapter number is rendered, the transformation is assigned, and the world is allowed to run. A Selenium wrapper script calls all 500 pages in succession, taking a screenshot at a random point in the animation:
The final book is rendered in a two-page spread PDF.
The Library of Babel is a place for scholars to do research, for artists and writers to seek inspiration, for anyone with curiosity or a sense of humor to reflect on the weirdness of existence - in short, it’s just like any other library. If completed, it would contain every possible combination of 1,312,000 characters, including lower case letters, space, comma, and period. Thus, it would contain every book that ever has been written, and every book that ever could be - including every play, every song, every scientific paper, every legal decision, every constitution, every piece of scripture, and so on. At present it contains all possible pages of 3200 characters, about 104677 books.
Since I imagine the question will present itself in some visitors’ minds (a certain amount of distrust of the virtual is inevitable) I’ll head off any doubts: any text you find in any location of the library will be in the same place in perpetuity. We do not simply generate and store books as they are requested - in fact, the storage demands would make that impossible. Every possible permutation of letters is accessible at this very moment in one of the library's books, only awaiting its discovery. We encourage those who find strange concatenations among the variations of letters to write about their discoveries in the forum, so future generations may benefit from their research.
Do you hate having to write your artist statement? Generate your own here for free, and if you don't like it, generate another one. For use with funding applications, exhibitions, curriculum vitae, websites ...
The Story Starter provides 298 million ideas for writers of novels, short stories, mysteries, horror, science fiction or silly stories. If you have never written a story, The Story Starter will give you ideas that will help you get started.
genoTyp is an experiment regarding fonts under genetic aspects. Their characteristics are encoded in hereditary factors. Different fonts can be mixed as desired and their genomes can be manipulated. New fonts are generated according to genetic rules.