[Question] Plans for Phi3-mini-128k-instruct?

unslothai / unsloth

Finetune Llama 3.1, Mistral, Phi & Gemma LLMs 2-5x faster with 80% less memory

https://unsloth.ai

Apache License 2.0

15.17k stars 1.01k forks source link

[Question] Plans for Phi3-mini-128k-instruct? #488

Open chemwolf6922 opened 3 months ago

chemwolf6922 commented 3 months ago

Hi unsloth team. I'm wondering if you have plans for supporting the microsoft/Phi-3-mini-128k-instruct? Also, is it viable for a average user like me to help porting a model to unsloth? What might process be like if that is something that you could share.

Feel free to close this if this is not appropriate for this project's issue tab.

danielhanchen commented 3 months ago

Oh yes someone did ask on that - we're planning to add automatic model support so any model becomes supported!

findalexli commented 1 month ago

Any update on this?

danielhanchen commented 1 month ago

No sorry currently not