Scaling Parameter-Constrained Language Models with Quality Data

• URL: http://arxiv.org/abs/2410.03083v1
• Date: Fri, 4 Oct 2024 02:07:17 GMT
• Title: Scaling Parameter-Constrained Language Models with Quality Data
• Authors: Ernie Chang, Matteo Paltenghi, Yang Li, Pin-Jie Lin, Changsheng Zhao, Patrick Huber, Zechun Liu, Rastislav Rabatin, Yangyang Shi, Vikas Chandra,
• Abstract summary: Scaling laws in language modeling traditionally quantify training loss as a function of dataset size and model parameters. We extend the conventional understanding of scaling law by offering a microscopic view of data quality within the original formulation.
• Score: 32.35610029333478
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Scaling laws in language modeling traditionally quantify training loss as a function of dataset size and model parameters, providing compute-optimal estimates but often neglecting the impact of data quality on model generalization. In this paper, we extend the conventional understanding of scaling law by offering a microscopic view of data quality within the original formulation -- effective training tokens -- which we posit to be a critical determinant of performance for parameter-constrained language models. Specifically, we formulate the proposed term of effective training tokens to be a combination of two readily-computed indicators of text: (i) text diversity and (ii) syntheticity as measured by a teacher model. We pretrained over $200$ models of 25M to 1.5B parameters on a diverse set of sampled, synthetic data, and estimated the constants that relate text quality, model size, training tokens, and eight reasoning task accuracy scores. We demonstrated the estimated constants yield +0.83 Pearson correlation with true accuracies, and analyzed it in scenarios involving widely-used data techniques such as data sampling and synthesis which aim to improve data quality.

Related papers

• Self-calibration for Language Model Quantization and Pruning [38.00221764773372]
  Quantization and pruning are fundamental approaches for model compression. In a post-training setting, state-of-the-art quantization and pruning methods require calibration data. We propose self-calibration as a solution.
  arXiv  Detail & Related papers  (2024-10-22T16:50:00Z)
• From Text to Treatment Effects: A Meta-Learning Approach to Handling Text-Based Confounding [7.5348062792]
  This paper examines the performance of meta-learners when confounding variables are expressed in text. We show that learners using pre-trained text representations of confounders achieve improved CATE estimates. Due to the entangled nature of the text embeddings, these models do not fully match the performance of meta-learners with perfect confounder knowledge.
  arXiv  Detail & Related papers  (2024-09-23T19:46:19Z)
• QuRating: Selecting High-Quality Data for Training Language Models [64.83332850645074]
  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.
  arXiv  Detail & Related papers  (2024-02-15T06:36:07Z)
• Influence Scores at Scale for Efficient Language Data Sampling [3.072340427031969]
  "influence scores" are used to identify important subsets of data. In this paper, we explore the applicability of influence scores in language classification tasks.
  arXiv  Detail & Related papers  (2023-11-27T20:19:22Z)
• The Languini Kitchen: Enabling Language Modelling Research at Different Scales of Compute [66.84421705029624]
  We introduce an experimental protocol that enables model comparisons based on equivalent compute, measured in accelerator hours. We pre-process an existing large, diverse, and high-quality dataset of books that surpasses existing academic benchmarks in quality, diversity, and document length. This work also provides two baseline models: a feed-forward model derived from the GPT-2 architecture and a recurrent model in the form of a novel LSTM with ten-fold throughput.
  arXiv  Detail & Related papers  (2023-09-20T10:31:17Z)
• Post-training Model Quantization Using GANs for Synthetic Data Generation [57.40733249681334]
  We investigate the use of synthetic data as a substitute for the calibration with real data for the quantization method. We compare the performance of models quantized using data generated by StyleGAN2-ADA and our pre-trained DiStyleGAN, with quantization using real data and an alternative data generation method based on fractal images.
  arXiv  Detail & Related papers  (2023-05-10T11:10:09Z)
• Leveraging Synthetic Targets for Machine Translation [5.302421715411791]
  We show that training models on synthetic targets outperforms training on the actual ground-truth data. We provide preliminary analysis into whether this boost in performance is linked to ease of optimization or more deterministic nature of the predictions.
  arXiv  Detail & Related papers  (2023-05-07T07:42:22Z)
• SynBench: Task-Agnostic Benchmarking of Pretrained Representations using Synthetic Data [78.21197488065177]
  Recent success in fine-tuning large models, that are pretrained on broad data at scale, on downstream tasks has led to a significant paradigm shift in deep learning. This paper proposes a new task-agnostic framework, textitSynBench, to measure the quality of pretrained representations using synthetic data.
  arXiv  Detail & Related papers  (2022-10-06T15:25:00Z)
• An Empirical Investigation of Commonsense Self-Supervision with Knowledge Graphs [67.23285413610243]
  Self-supervision based on the information extracted from large knowledge graphs has been shown to improve the generalization of language models. We study the effect of knowledge sampling strategies and sizes that can be used to generate synthetic data for adapting language models.
  arXiv  Detail & Related papers  (2022-05-21T19:49:04Z)
• Evaluating natural language processing models with generalization metrics that do not need access to any training or testing data [66.11139091362078]
  We provide the first model selection results on large pretrained Transformers from Huggingface using generalization metrics. Despite their niche status, we find that metrics derived from the heavy-tail (HT) perspective are particularly useful in NLP tasks.
  arXiv  Detail & Related papers  (2022-02-06T20:07:35Z)
• How much pretraining data do language models need to learn syntax? [12.668478784932878]
  Transformers-based pretrained language models achieve outstanding results in many well-known NLU benchmarks. We study the impact of pretraining data size on the knowledge of the models using RoBERTa.
  arXiv  Detail & Related papers  (2021-09-07T15:51:39Z)

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.