Phainon/
margin_dpo_llama3_8b_harmless

Name: margin_dpo_llama3_8b_harmless
Availability: InStock

like

0

safetensors

llama

us

Views

No views yet

DPO fine-tuned LLaMA3 8B model (harmless setting).

Usage

python

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("Phainon/margin_dpo_llama3_8b_harmless")
tokenizer = AutoTokenizer.from_pretrained("Phainon/margin_dpo_llama3_8b_harmless")