Modelpublic

basab1142/hawkes-nsgo-llama3-hawkes-dpo

sourceHugging Faceupdated 2mo agoView on Hugging Face

0likes3downloads

Model Card

basab1142/hawkes-nsgo-llama3-hawkes-dpo

LoRA adapter trained with Hawkes-DPO on the kC-NSGO synthetic preference dataset (Hawkes-process-driven opinion shifts).

Field	Value
Base model	`meta-llama/Meta-Llama-3.1-8B-Instruct`
Method	Hawkes-DPO
Regime	`adversarial_ns_dpo`
Discount gap	4.249
Final reward accuracy	nan
LoRA rank	16
LoRA targets	qproj, kproj, vproj, oproj

Load

python

from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel

base = AutoModelForCausalLM.from_pretrained(
    "meta-llama/Meta-Llama-3.1-8B-Instruct",
    torch_dtype="bfloat16", device_map="auto"
)
model = PeftModel.from_pretrained(base, "basab1142/hawkes-nsgo-llama3-hawkes-dpo")
tokenizer = AutoTokenizer.from_pretrained("meta-llama/Meta-Llama-3.1-8B-Instruct")