Add support for OpenELM models from apple?

unslothai / unsloth

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

https://unsloth.ai

Apache License 2.0

15.54k stars 1.04k forks source link

Add support for OpenELM models from apple? #379

Open NilanEkanayake opened 4 months ago

NilanEkanayake commented 4 months ago

https://huggingface.co/apple/OpenELM

Has models ranging from 270M to 3B parameters. Would love to see more support for small models, since I'm stuck with 4gb VRAM currently. Tinyllama can't fill every niche I try to wedge it into.

danielhanchen commented 4 months ago

Interesting. I'll see what I can do!

FarhanAnis005 commented 4 months ago

Hey any updates on OpenELM