Hugging Face's logo

Join the conversation

Join the community of Machine Learners and AI enthusiasts.

Sign Up
Back to feed
sometimesanotion 
posted an update Jan 22, 2025
Post
2764
I've managed a #1 score of 41.22% average for 14B parameter models on the Open LLM Leaderboard. As of this writing, sometimesanotion/Lamarck-14B-v0.7 is #8 for all models up to 70B parameters.

It took a custom toolchain around Arcee AI's mergekit to manage the complex merges, gradients, and LoRAs required to make this happen. I really like seeing features of many quality finetunes in one solid generalist model.
suayptalha

Appreciate your work!

·
sometimesanotion

Yours too! You get a lot out of the 7B parameter models!
sthenno

Congratulations! BTW, I'm still waiting for your response to https://huggingface.co/bamec66557/Qwen-2.5-14B-MINUS/discussions/1#678250364248fde89ea918f7 :)

·
sometimesanotion

Thank you, I somehow missed that notification! You can hit me up to discuss model merges anytime. You too, @CultriX .
minpeter

This is an impressive model. Try deploying it to your Friendli endpoints via the "Deploy" button at https://huggingface.co/sometimesanotion/Lamarck-14B-v0.7 and experimenting!
lubdhak98

peak 🥶
although generates around 8 tokesn/s on my 3070 lol
CultriX

Congrats! Very nicely done :)!

·
sometimesanotion

You helped this project get started, validating merge methods. Thank you!

In this post