Please add model knowledge other than code/math
#2
by
CoffeeBliss
- opened
Hello Xiaomi LLM-Core Team,
I’d like to request broadening the model’s expertise beyond coding/math.
Right now, it’s good for those topics but struggles with medical, humanities, arts, history, and everyday topics.
Suggestions:
Mix in training data from non-tech sources (textbooks, journals, medical books)
Test how well it handles things like healthcare, history, or storytelling.
Why this matters:
Makes the model more useful for students, writers, and casual users.
Keeps open-source AI accessible to everyone, not just tech/math users.