Sorry the ignorance, it's better than the 8b version?

#5
by ZeroCool22 - opened

If yes, i could run it with:

Ryzen 5900x.
GPU: 1080 TI (11Gb of VRAM).
32Gb RAM.

And what mean "Instruct" and this can be used with Ollama APP?

Hi @ZeroCool22

This is an experimental model merged to have more parameters based on the Instruct Llama-3-8B. It is not that far off, but it's not better.
It is intended for others to use it in future merges and further fine-tuning.

I recommend looking here to find better models than the Llama-3-8B : https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=llama-3

Hi @ZeroCool22

This is an experimental model merged to have more parameters based on the Instruct Llama-3-8B. It is not that far off, but it's not better.
It is intended for others to use it in future merges and further fine-tuning.

I recommend looking here to find better models than the Llama-3-8B : https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard?query=llama-3

Thank you.

Sign up or log in to comment