QuRating: Selecting High-Quality Data for Training Language Models

• URL: http://arxiv.org/abs/2402.09739v3
• Date: Wed, 17 Jul 2024 20:50:11 GMT
• Title: QuRating: Selecting High-Quality Data for Training Language Models
• Authors: Alexander Wettig, Aatmik Gupta, Saumya Malik, Danqi Chen,
• Abstract summary: We introduce QuRating, a method for selecting pre-training data that can capture human intuitions about data quality. In this paper, we investigate four qualities - writing style, required expertise, facts & trivia, and educational value. We train a Qur model to learn scalar ratings from pairwise judgments, and use it to annotate a 260B training corpus with quality ratings for each of the four criteria.
• Score: 64.83332850645074
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Selecting high-quality pre-training data is important for creating capable language models, but existing methods rely on simple heuristics. We introduce QuRating, a method for selecting pre-training data that can capture human intuitions about data quality. In this paper, we investigate four qualities - writing style, required expertise, facts & trivia, and educational value - and find that LLMs are able to discern these qualities, especially when making pairwise judgments of texts. We train a QuRater model to learn scalar ratings from pairwise judgments, and use it to annotate a 260B training corpus with quality ratings for each of the four criteria. In our experiments, we select 30B tokens according to the different quality ratings and train 1.3B-parameter language models on the selected data. We find that it is important to balance quality and diversity. When we sample using quality ratings as logits over documents, our models obtain lower perplexity and stronger in-context learning performance than baselines. Our best model is based on educational value and performs similarly to a model trained with uniform sampling for 50% more steps. Beyond data selection, we use the quality ratings to construct a training curriculum which improves performance without changing the training dataset. We extensively analyze the quality ratings and discuss their characteristics, biases, and wider implications.

Related papers

• SCAR: Efficient Instruction-Tuning for Large Language Models via Style Consistency-Aware Response Ranking [56.93151679231602]
  This research identifies two key stylistic elements in responses: linguistic form and semantic surprisal. Inspired by this, we introduce Style Consistency-Aware Response Ranking (SCAR) SCAR prioritizes instruction-response pairs in the training set based on their response stylistic consistency.
  arXiv  Detail & Related papers  (2024-06-16T10:10:37Z)
• Text Quality-Based Pruning for Efficient Training of Language Models [66.66259229732121]
  We propose a novel method for numerically evaluating text quality in large unlabelled NLP datasets. By proposing the text quality metric, the paper establishes a framework to identify and eliminate low-quality text instances. Experimental results over multiple models and datasets demonstrate the efficacy of this approach.
  arXiv  Detail & Related papers  (2024-04-26T18:01:25Z)
• Smaller Language Models are capable of selecting Instruction-Tuning Training Data for Larger Language Models [39.65879784788677]
  We introduce a novel training data selection based on the learning percentage of the samples. We assert that current language models possess the capability to autonomously select high-quality training data. Our paper introduces a novel approach to training data selection, showcasing a more efficient alternative.
  arXiv  Detail & Related papers  (2024-02-16T03:39:37Z)
• DsDm: Model-Aware Dataset Selection with Datamodels [81.01744199870043]
  Standard practice is to filter for examples that match human notions of data quality. We find that selecting according to similarity with "high quality" data sources may not increase (and can even hurt) performance compared to randomly selecting data. Our framework avoids handpicked notions of data quality, and instead models explicitly how the learning process uses train datapoints to predict on the target tasks.
  arXiv  Detail & Related papers  (2024-01-23T17:22:00Z)
• A Novel Metric for Measuring Data Quality in Classification Applications (extended version) [0.0]
  We introduce and explain a novel metric to measure data quality. This metric is based on the correlated evolution between the classification performance and the deterioration of data. We provide an interpretation of each criterion and examples of assessment levels.
  arXiv  Detail & Related papers  (2023-12-13T11:20:09Z)
• Calibrating LLM-Based Evaluator [92.17397504834825]
  We propose AutoCalibrate, a multi-stage, gradient-free approach to calibrate and align an LLM-based evaluator toward human preference. Instead of explicitly modeling human preferences, we first implicitly encompass them within a set of human labels. Our experiments on multiple text quality evaluation datasets illustrate a significant improvement in correlation with expert evaluation through calibration.
  arXiv  Detail & Related papers  (2023-09-23T08:46:11Z)
• Skill-it! A Data-Driven Skills Framework for Understanding and Training Language Models [29.17711426767209]
  We study how to best select data that leads to good downstream model performance across tasks. We develop a new framework based on a simple hypothesis: just as humans acquire interdependent skills in a deliberate order, language models also follow a natural order when learning a set of skills from their training data.
  arXiv  Detail & Related papers  (2023-07-26T18:01:49Z)
• Skill-Based Few-Shot Selection for In-Context Learning [123.26522773708683]
  Skill-KNN is a skill-based few-shot selection method for in-context learning. It does not require training or fine-tuning of any models, making it suitable for frequently expanding or changing example banks. Experimental results across five cross-domain semantic parsing datasets and six backbone models show that Skill-KNN significantly outperforms existing methods.
  arXiv  Detail & Related papers  (2023-05-23T16:28:29Z)
• Bias-Aware Loss for Training Image and Speech Quality Prediction Models from Multiple Datasets [13.132388683797503]
  We propose a bias-aware loss function that estimates each dataset's biases during training with a linear function. We prove the efficiency of the proposed method by training and validating quality prediction models on synthetic and subjective image and speech quality datasets.
  arXiv  Detail & Related papers  (2021-04-20T19:20:11Z)
• Reinforced Curriculum Learning on Pre-trained Neural Machine Translation Models [20.976165305749777]
  We learn a curriculum for improving a pre-trained NMT model by re-selecting influential data samples from the original training set. We propose a data selection framework based on Deterministic Actor-Critic, in which a critic network predicts the expected change of model performance.
  arXiv  Detail & Related papers  (2020-04-13T03:40:44Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.