Self-Directed Synthetic Dialogues and Revisions Technical Report

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lambert, Nathan, Schoelkopf, Hailey, Gokaslan, Aaron, Soldaini, Luca, Pyatkin, Valentina, Castricato, Louis
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914888209661952
author Lambert, Nathan
Schoelkopf, Hailey
Gokaslan, Aaron
Soldaini, Luca
Pyatkin, Valentina
Castricato, Louis
author_facet Lambert, Nathan
Schoelkopf, Hailey
Gokaslan, Aaron
Soldaini, Luca
Pyatkin, Valentina
Castricato, Louis
contents Synthetic data has become an important tool in the fine-tuning of language models to follow instructions and solve complex problems. Nevertheless, the majority of open data to date is often lacking multi-turn data and collected on closed models, limiting progress on advancing open fine-tuning methods. We introduce Self Directed Synthetic Dialogues (SDSD), an experimental dataset consisting of guided conversations of language models talking to themselves. The dataset consists of multi-turn conversations generated with DBRX, Llama 2 70B, and Mistral Large, all instructed to follow a conversation plan generated prior to the conversation. We also explore including principles from Constitutional AI and other related works to create synthetic preference data via revisions to the final conversation turn. We hope this work encourages further exploration in multi-turn data and the use of open models for expanding the impact of synthetic data.
format Preprint
id arxiv_https___arxiv_org_abs_2407_18421
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Self-Directed Synthetic Dialogues and Revisions Technical Report
Lambert, Nathan
Schoelkopf, Hailey
Gokaslan, Aaron
Soldaini, Luca
Pyatkin, Valentina
Castricato, Louis
Computation and Language
Machine Learning
Synthetic data has become an important tool in the fine-tuning of language models to follow instructions and solve complex problems. Nevertheless, the majority of open data to date is often lacking multi-turn data and collected on closed models, limiting progress on advancing open fine-tuning methods. We introduce Self Directed Synthetic Dialogues (SDSD), an experimental dataset consisting of guided conversations of language models talking to themselves. The dataset consists of multi-turn conversations generated with DBRX, Llama 2 70B, and Mistral Large, all instructed to follow a conversation plan generated prior to the conversation. We also explore including principles from Constitutional AI and other related works to create synthetic preference data via revisions to the final conversation turn. We hope this work encourages further exploration in multi-turn data and the use of open models for expanding the impact of synthetic data.
title Self-Directed Synthetic Dialogues and Revisions Technical Report
topic Computation and Language
Machine Learning
url https://arxiv.org/abs/2407.18421