For my 2nd (!) NaNoGenMo project this year, I am going to prompt a finetuned GPT-2 model with a sequence of manually written titles and original photographs.
My goal will be to create a generative work with more traditional / "direct" intent, and which (hopefully) will (more or less) represent a guided "expansion" of the supplied material.
For my 2nd (!) NaNoGenMo project this year, I am going to prompt a finetuned GPT-2 model with a sequence of manually written titles and original photographs.
My goal will be to create a generative work with more traditional / "direct" intent, and which (hopefully) will (more or less) represent a guided "expansion" of the supplied material.
I got started tonight with this....
[ uncurated / unedited / first run ]
Further.pdf
I will post the code later.