Merge Recipe??

#1
by fhai50032 - opened

Hi, If you can provide the merge recipe for listed models
and if any datasets used and how did you aligned a model to represent creative writing and another to show aggressive behaviour

PS: Model card is too funny

Check out mergekit_config.yml. During my tests I found out that Dolphin really likes to go aggro, spicyboros and xwin both are good at stylized writing, and I used lzlv as base because it's got a little something of everything. I just ties-merged them and didn't do any additional training. I you want to see the results of my tests, check out my meme benchmark.
Do you have any additional comments about the model, by the way?

Sign up or log in to comment