TiC-LM: A Web-Scale Benchmark for Time-Continual LLM Pretraining

• URL: http://arxiv.org/abs/2504.02107v1
• Date: Wed, 02 Apr 2025 20:11:54 GMT
• Title: TiC-LM: A Web-Scale Benchmark for Time-Continual LLM Pretraining
• Authors: Jeffrey Li, Mohammadreza Armandpour, Iman Mirzadeh, Sachin Mehta, Vaishaal Shankar, Raviteja Vemulapalli, Samy Bengio, Oncel Tuzel, Mehrdad Farajtabar, Hadi Pouransari, Fartash Faghri,
• Abstract summary: Large Language Models (LLMs) trained on historical web data inevitably become outdated.<n>We introduce a web-scale dataset for time-continual pretraining of LLMs derived from 114 dumps of Common Crawl (CC)<n>We also design time-stratified evaluations across both general CC data and specific domains.
• Score: 41.629324249810054
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large Language Models (LLMs) trained on historical web data inevitably become outdated. We investigate evaluation strategies and update methods for LLMs as new data becomes available. We introduce a web-scale dataset for time-continual pretraining of LLMs derived from 114 dumps of Common Crawl (CC) - orders of magnitude larger than previous continual language modeling benchmarks. We also design time-stratified evaluations across both general CC data and specific domains (Wikipedia, StackExchange, and code documentation) to assess how well various continual learning methods adapt to new data while retaining past knowledge. Our findings demonstrate that, on general CC data, autoregressive meta-schedules combined with a fixed-ratio replay of older data can achieve comparable held-out loss to re-training from scratch, while requiring significantly less computation (2.6x). However, the optimal balance between incorporating new data and replaying old data differs as replay is crucial to avoid forgetting on generic web data but less so on specific domains.

Related papers

• Large Language Models are Few-shot Multivariate Time Series Classifiers [23.045734479292356]
  Large Language Models (LLMs) have been extensively applied in time series analysis. Yet, their utility in the few-shot classification (i.e., a crucial training scenario) is underexplored. We aim to leverage the extensive pre-trained knowledge in LLMs to overcome the data scarcity problem.
  arXiv  Detail & Related papers  (2025-01-30T03:59:59Z)
• Formality is Favored: Unraveling the Learning Preferences of Large Language Models on Data with Conflicting Knowledge [55.65162959527848]
  Large language models have shown excellent performance on many knowledge-intensive tasks. However, pretraining data tends to contain misleading and even conflicting information. This study systematically analyze LLMs' learning preferences for data with conflicting knowledge.
  arXiv  Detail & Related papers  (2024-10-07T06:49:41Z)
• Accelerating Large Language Model Pretraining via LFR Pedagogy: Learn, Focus, and Review [50.78587571704713]
  Learn-Focus-Review (LFR) is a dynamic training approach that adapts to the model's learning progress.<n>LFR tracks the model's learning performance across data blocks (sequences of tokens) and prioritizes revisiting challenging regions of the dataset.<n>Compared to baseline models trained on the full datasets, LFR consistently achieved lower perplexity and higher accuracy.
  arXiv  Detail & Related papers  (2024-09-10T00:59:18Z)
• Improving Pretraining Data Using Perplexity Correlations [56.41097718862742]
  We present a framework that selects high-quality pretraining data without any LLM training of our own.<n>We build a new statistical framework for data selection centered around estimates of perplexity-benchmark correlations.<n>Our approach outperforms DSIR on every benchmark, while matching the best data selector found in DataComp-LM.
  arXiv  Detail & Related papers  (2024-09-09T17:23:29Z)
• Simple and Scalable Strategies to Continually Pre-train Large Language Models [20.643648785602462]
  Large language models (LLMs) are routinely pre-trained on billions of tokens, only to start the process over again once new data becomes available. We show that a simple and scalable combination of learning rate re-warming, LR re-decaying, and replay of previous data is sufficient to match the performance of fully re-training from scratch.
  arXiv  Detail & Related papers  (2024-03-13T17:58:57Z)
• Enhancing Consistency and Mitigating Bias: A Data Replay Approach for Incremental Learning [93.90047628101155]
  Deep learning systems are prone to catastrophic forgetting when learning from a sequence of tasks.<n>To address this, some methods propose replaying data from previous tasks during new task learning.<n>However, it is not expected in practice due to memory constraints and data privacy issues.
  arXiv  Detail & Related papers  (2024-01-12T12:51:12Z)
• Data Contamination Through the Lens of Time [21.933771085956426]
  Large language models (LLMs) are often supported by evaluating publicly available benchmarks. This practice raises concerns of data contamination, i.e., evaluating on examples that are explicitly or implicitly included in the training data. We conduct the first thorough longitudinal analysis of data contamination in LLMs by using the natural experiment of training cutoffs in GPT models.
  arXiv  Detail & Related papers  (2023-10-16T17:51:29Z)
• When Less is More: Investigating Data Pruning for Pretraining LLMs at Scale [12.94829977468838]
  Large volumes of text data have contributed significantly to the development of large language models. To date, efforts to prune datasets down to a higher quality subset have relied on hand-crafteds encoded as rule-based filters. We take a wider view and explore scalable estimates of data quality that can be used to measure the quality of pretraining data.
  arXiv  Detail & Related papers  (2023-09-08T19:34:05Z)
• Don't Memorize; Mimic The Past: Federated Class Incremental Learning Without Episodic Memory [36.4406505365313]
  This paper presents a framework for federated class incremental learning that utilizes a generative model to synthesize samples from past distributions instead of storing part of past data. The generative model is trained on the server using data-free methods at the end of each task without requesting data from clients.
  arXiv  Detail & Related papers  (2023-07-02T07:06:45Z)
• To Repeat or Not To Repeat: Insights from Scaling LLM under Token-Crisis [50.31589712761807]
  Large language models (LLMs) are notoriously token-hungry during pre-training, and high-quality text data on the web is approaching its scaling limit for LLMs. We investigate the consequences of repeating pre-training data, revealing that the model is susceptible to overfitting. Second, we examine the key factors contributing to multi-epoch degradation, finding that significant factors include dataset size, model parameters, and training objectives.
  arXiv  Detail & Related papers  (2023-05-22T17:02:15Z)
• On the Costs and Benefits of Adopting Lifelong Learning for Software Analytics -- Empirical Study on Brown Build and Risk Prediction [17.502553991799832]
  This paper evaluates the use of lifelong learning (LL) for industrial use cases at Ubisoft. LL is used to continuously build and maintain ML-based software analytics tools using an incremental learner that progressively updates the old model using new data.
  arXiv  Detail & Related papers  (2023-05-16T21:57:16Z)
• Improving information retention in large scale online continual learning [99.73847522194549]
  Online continual learning aims to adapt efficiently to new data while retaining existing knowledge. Recent work suggests that information retention remains a problem in large scale OCL even when the replay buffer is unlimited. We propose using a moving average family of methods to improve optimization for non-stationary objectives.
  arXiv  Detail & Related papers  (2022-10-12T16:59:43Z)

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.