Selecting a suitable pretraining dataset is crucial for both general-domain (e.g., GPT -

Selecting a suitable pretraining dataset is crucial for both general-domain (e.g., GPT -

As large language models (LLMs) are pretrained on massive web corpora, careful selection of data becomes essential to ensure effective and efficient learning. While perplexity (PPL)-based filtering has shown strong performance, it suffers from drawbacks: substantial time costs and inherent unreliability of the model when handling noisy or out-of-distribution samples. In this work, we propose a simple yet powerful alternative: a prior-based data filtering method that estimates token priors using corpus-level term frequency statistics, inspired by linguistic insights on word roles and lexical density. Our approach filters documents based on the mean and standard deviation of token priors, serving as a fast proxy to PPL while requiring no model inference. Despite its simplicity, the prior-based filter achieves the highest average performance across 20 downstream benchmarks, while reducing time cost by over 1000x compared to PPL-based filtering. We further demonstrate its applicability to symbolic languages such as code and math, and its dynamic adaptability to multilingual corpora without supervision

Selecting high-quality pre-training data for large language models (LLMs) is crucial for enhancing their overall performance under limited computation budget, improving both training and sample efficiency. Recent advancements in file selection primarily rely on using an existing or trained proxy model to assess the similarity of samples to a target domain, such as high quality sources BookCorpus and Wikipedia. However, upon revisiting these methods, the domain-similarity selection criteria demonstrates a diversity dilemma, i.e.dimensional collapse in the feature space, improving performance on the domain-related tasks but causing severe degradation on generic performance. To prevent collapse and enhance diversity, we propose a DiverSified File selection algorithm (DiSF), which selects the most decorrelated text files in the feature space. We approach this with a classical greedy algorithm to achieve more uniform eigenvalues in the feature covariance matrix of the selected texts, analyzing its approximation to the optimal solution under a formulation of $γ$-weakly submodular optimization problem. Empirically, we establish a benchmark and conduct extensive experiments on the TinyLlama architecture with models from 120M to 1.1B parameters. Evaluating across nine tasks from the Harness framework, DiSF demonstrates a significant improvement on overall performance. Specifically, DiSF saves 98.5% of 590M training files in SlimPajama, outperforming the full-data pre-training within a 50B training budget, and achieving about 1.5x training efficiency and 5x data efficiency.

Data selection is critical for enhancing the performance of language models, particularly when aligning training datasets with a desired target distribution. This study explores the effects of different data selection methods and feature types on model performance. We evaluate whether selecting data subsets can influence downstream tasks, whether n-gram features improve alignment with target distributions, and whether embedding-based neural features provide complementary benefits. Through comparative experiments using baseline random selection methods and distribution aligned approaches, we provide insights into the interplay between data selection strategies and model training efficacy. All code for this study can be found on \href{https://github.com/jgu13/HIR-Hybrid-Importance-Resampling-for-Language-Models}{github repository}.

Data selection is crucial for optimizing language model (LM) performance on specific tasks, yet most existing methods fail to effectively consider the target task distribution. Current approaches either ignore task-specific requirements entirely or rely on approximations that fail to capture the nuanced patterns needed for tasks like Autoformalization or code generation. Methods that do consider the target distribution often rely on simplistic, sometimes noisy, representations, like hashed n-gram features, which can lead to collisions and introduce noise. We introduce ZIP-FIT, a data selection framework that uses gzip compression to directly measure alignment between potential training data and the target task distribution. Our key insight is that compression-based similarity captures both syntactic and structural patterns relevant to the target task, enabling more precise selection of truly task-relevant data. In extensive evaluations on Autoformalization and Python code generation, ZIP-FIT significantly outperforms leading baselines like DSIR and D4. Models trained on ZIP-FIT-selected data achieve their lowest cross-entropy loss up to 85.1% faster than baselines, demonstrating that better task alignment leads to more efficient learning. In addition, ZIP-FIT performs selection up to 65.8% faster than DSIR and two orders of magnitude faster than D4. Notably, ZIP-FIT shows that smaller, well-aligned datasets often outperform larger but less targeted ones, demonstrating that a small amount of higher quality data is superior to a large amount of lower quality data. Our results imply that task-aware data selection is crucial for efficient domain adaptation, and that compression offers a principled way to measure task alignment. By showing that targeted data selection can dramatically improve task-specific performance, our work provides new insights into the relationship between data quality, task alignment, and model learning efficiency.

Selecting a suitable pretraining dataset is crucial for both general-domain (e.g., GPT-3) and domain-specific (e.g., Codex) language models (LMs). We formalize this problem as selecting a subset of a large raw unlabeled dataset to match a desired target distribution given unlabeled target samples. Due to the scale and dimensionality of the raw text data, existing methods use simple heuristics or require human experts to manually curate data. Instead, we extend the classic importance resampling approach used in low-dimensions for LM data selection. We propose Data Selection with Importance Resampling (DSIR), an efficient and scalable framework that estimates importance weights in a reduced feature space for tractability and selects data with importance resampling according to these weights. We instantiate the DSIR framework with hashed n-gram features for efficiency, enabling the selection of 100M documents from the full Pile dataset in 4.5 hours. To measure whether hashed n-gram features preserve the aspects of the data that are relevant to the target, we define KL reduction, a data metric that measures the proximity between the selected pretraining data and the target on some feature space. Across 8 data selection methods (including expert selection), KL reduction on hashed n-gram features highly correlates with average downstream accuracy (r=0.82). When selecting data for continued pretraining on a specific domain, DSIR performs comparably to expert curation across 8 target distributions. When pretraining general-domain models (target is Wikipedia and books), DSIR improves over random selection and heuristic filtering baselines by 2-2.5% on the GLUE benchmark. Code is available at https://github.com/p-lambda/dsir.