Finetuning further
#21
by
noobmldude
- opened
If I have to finetune this model further on my personal codebase, how should I go about doing it?
Should i continue the pretraining or should i create instruction response pairs?
How should I train for specific tool use or agentic workflows?