training corpus / benchmarks
#1
by
medmac01
- opened
Hi there!
First, thank you for sharing this interesting model. I'm curious to know about the training corpus used to finetune this model, from both quantity and variety perspectives. Also, was there any evaluations done for assessing how accurate are the generated rules?
Have a good day :)
So I created the custom fine tuning data. It's got a mix of security topics.
I am working on creating benchmarks for some more focused security topics but it's still a work in progress.