This repository contains the materials and code associated with my participation in the LLM Merging Competition: Building LLMs Efficiently through Merging at the NeurIPS 2024 Challenge. I am proud to announce that my team secured 2nd place in the competition!
You can access the paper related to this project through the following link:
For more details about the competition, you can visit the official competition website:
The official Kaggle competition page where the results and code are hosted:
The merged LLM model that I developed for this competition is hosted on Hugging Face. You can explore and use the model through the following link:
The LLM Merging Competition challenges participants to develop efficient techniques for merging large language models (LLMs). The goal is to improve the performance and efficiency of LLMs by merging different models, creating a more capable and cost-effective solution for future applications in natural language processing.
- 2nd Place in the LLM Merging Competition at NeurIPS 2024 Challenge
- Presented our work in a thematic presentation at the event