Suggestion for Meta-Llama 3.1-8B Instruct change model

#8
by legolasyiu - opened

Hello unsloth,

quick suggestion,
Could you please fine tune the model with Meta-Llama 3.1-8B-Instruct instead of Meta-Llama 3.1-8B?

Seeing that your base model is unsloth/Meta-Llama-3.1-8B
Just want to help
Thomas

Oh, Thanks a lot.
I'm just a beginner

Just figure out, unsloth fine tune with instruct dataset and it will convert to Instruct. :)

Unsloth AI org

Hello unsloth,

quick suggestion,
Could you please fine tune the model with Meta-Llama 3.1-8B-Instruct instead of Meta-Llama 3.1-8B?

Seeing that your base model is unsloth/Meta-Llama-3.1-8B
Just want to help
Thomas

You can fine-tune using the instruct or base models whichever you prefer. Instruct is usually better if you have a smaller dataset!!

Thanks for the clarification!

What about unsloth/Meta-Llama-3.1-8B-Instruct? Is it same idea, to fine tune with smaller dataset?

Unsloth AI org

What about unsloth/Meta-Llama-3.1-8B-Instruct? Is it same idea, to fine tune with smaller dataset?

Yes correct

Thank you so much! So informative!

Sign up or log in to comment