Arpita1/sbs_convai2_llama3.1_lora

Model Card

Description

Llama-3.1-8B-Instruct finetuned on ConvAI2 using the SBS framework.

• Repository: GitHub
• Paper: https://arxiv.org/abs/2508.06886
• Funded by: UKRI AI-Medical CDT (Grant Reference: EP/S024336/1)
• Language(s) (NLP): English
• License: Llama 3.1

Note: These are just the LoRA weights and need to be merged with the Llama-3.1-8B-Instruct model before use.

BibTeX

@inproceedings{saggar2025,
  author    = {Saggar, Arpita and Darling, Jonathan C. and Dimitrova, Vania and Sarikaya, Duygu and Hogg, David C.},
  title     = {Score Before You Speak: Improving Persona Consistency in Dialogue Generation using Response Quality Scores},
  booktitle = {Proceedings of the 28th European Conference on Artificial Intelligence},
  year      = {2025},
}