paper-with-me

SurgeGlobal/LaMini

홈페이지 · 논문 1편

## Overview The LaMini Dataset is an instruction dataset generated using h2ogpt-gm-oasst1-en-2048-falcon-40b-v2. It is designed for instruction-tuning pre-trained models to specialize them in a variety of downstream tasks. ## Dataset Generation - Base Model: h2oai/h2ogpt-gm-oasst1-en-2048-falcon-40b-v2. - Seed Instructions: Sourced from databricks/databricks-dolly-15k dataset. - Generation Approach: Example-guided and topic-guided strategies. - Total Instructions: 1,504 unique instruction examples. ### Dataset Sources - Repository: Bitbucket Project - Paper : Pre-Print ## Structure Each entry in the dataset contains: - Instruction - Response ## Usage The LaMini Dataset can be used to fine-tune language models to improve their ability to follow instructions and generate relevant responses. ## Access The dataset is available on HuggingFace at the following link: https://huggingface.co/datasets/SurgeGlobal/LaMini ## Citation If you find our work useful, please cite our paper as follows: `` @misc{surge2024openbezoar, title={OpenBezoar: Small, Cost-Effective and Open Models Trained on Mixes of Instruction Data}, author={Chandeepa Dissanayake and Lahiru Lowe and Sachith Gunasekara and Yasiru Ratnayake}, year={2024}, eprint={2404.12195}, archivePrefix={arXiv}, primaryClass={cs.CL} } `` ## Dataset Authors Chandeepa Dissanayake, Lahiru Lowe, Sachith Gunasekara, and Yasiru Ratnayake

Texts English