Rövid ismertetőm
#2
by
Emalog
- opened
The model definitely gives more detailed answers than a Llama3_8b model.
It requires more precise wording (instructions), to describe the "exclipt" or "Dirty languange"content than some other language models.
The big disadvantage of the model is that as soon as it reaches the set context text (token) lenght, it immediately loses the information given in the system prompt. (In other models, a process is used and the prompt information is slowly lost).