From Quantity to Quality: Boosting LLM Performance with Self-Guided Data Selection for Instruction Tuning

08/23/2023

∙

In the realm of Large Language Models, the balance between instruction data quality and quantity has become a focal point. Recognizing this, we introduce a self-guided methodology for LLMs to autonomously discern and select cherry samples from vast open-source datasets, effectively minimizing manual curation and potential cost for instruction tuning an LLM. Our key innovation, the Instruction-Following Difficulty (IFD) metric, emerges as a pivotal tool to identify discrepancies between a model's expected responses and its autonomous generation prowess. Through the adept application of IFD, cherry samples are pinpointed, leading to a marked uptick in model training efficiency. Empirical validations on renowned datasets like Alpaca and WizardLM underpin our findings; with a mere 10 improved results. This synthesis of self-guided cherry-picking and the IFD metric signifies a transformative leap in the optimization of LLMs, promising both efficiency and resource-conscious advancements.

READ FULL TEXT

From Quantity to Quality: Boosting LLM Performance with Self-Guided Data Selection for Instruction Tuning

Panda LLM: Training Data and Evaluation for Open-Sourced Chinese Instruction-Following Large Language Models

Self-QA: Unsupervised Knowledge Guided Language Model Alignment

Efficient Finetuning Large Language Models For Vietnamese Chatbot

Towards Better Instruction Following Language Models for Chinese: Investigating the Impact of Training Data and Evaluation

Time Travel in LLMs: Tracing Data Contamination in Large Language Models

Donkii: Can Annotation Error Detection Methods Find Errors in Instruction-Tuning Datasets?

Flesch or Fumble? Evaluating Readability Standard Alignment of Instruction-Tuned Language Models

From Quantity to Quality: Boosting LLM Performance with Self-Guided Data Selection for Instruction Tuning

Related Research

Panda LLM: Training Data and Evaluation for Open-Sourced Chinese Instruction-Following Large Language Models

Self-QA: Unsupervised Knowledge Guided Language Model Alignment

Efficient Finetuning Large Language Models For Vietnamese Chatbot

Towards Better Instruction Following Language Models for Chinese: Investigating the Impact of Training Data and Evaluation

Time Travel in LLMs: Tracing Data Contamination in Large Language Models

Donkii: Can Annotation Error Detection Methods Find Errors in Instruction-Tuning Datasets?

Flesch or Fumble? Evaluating Readability Standard Alignment of Instruction-Tuned Language Models