Early Weight Averaging meets High Learning Rates for LLM Pre-training

• URL: http://arxiv.org/abs/2306.03241v2
• Date: Mon, 11 Dec 2023 22:31:12 GMT
• Title: Early Weight Averaging meets High Learning Rates for LLM Pre-training
• Authors: Sunny Sanyal, Atula Neerkaje, Jean Kaddour, Abhishek Kumar and Sujay Sanghavi
• Abstract summary: We show that models trained with high learning rates observe higher gains due to checkpoint averaging. Our training recipe outperforms conventional training and popular checkpoint averaging baselines.
• Score: 20.671831210738937
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Training Large Language Models (LLMs) incurs significant cost; hence, any strategy that accelerates model convergence is helpful. In this paper, we investigate the ability of a simple idea checkpoint averaging along the trajectory of a training run to improve both convergence and generalization quite early on during training. Here we show that models trained with high learning rates observe higher gains due to checkpoint averaging. Furthermore, these gains are amplified when checkpoints are sampled with considerable spacing in training steps. Our training recipe outperforms conventional training and popular checkpoint averaging baselines such as exponential moving average (EMA) and stochastic moving average (SWA). We evaluate our training recipe by pre-training LLMs, where high learning rates are inherently preferred due to extremely large batch sizes. Specifically, we pre-trained nanoGPT-2 models of varying sizes, small (125M), medium (335M), and large (770M)on the OpenWebText dataset, comprised of 9B tokens. Additionally, we present results for publicly available Pythia LLMs, ranging from 1B to 12B, which were trained on the PILE-deduped dataset containing 207B tokens.

Related papers

• The Journey Matters: Average Parameter Count over Pre-training Unifies Sparse and Dense Scaling Laws [51.608402959163925]
  We present the first systematic exploration of optimal sparse pre-training configurations for large language models. We find that initiating pruning at 25% of total training compute and concluding at 75% achieves near-optimal final evaluation loss. We propose a new scaling law that modifies the Chinchilla scaling law to use the average parameter count over pre-training.
  arXiv  Detail & Related papers  (2025-01-21T20:23:22Z)
• Preference Curriculum: LLMs Should Always Be Pretrained on Their Preferred Data [19.221998577357713]
  Large language models (LLMs) generally utilize a consistent data distribution throughout the pretraining process. As the model's capability improves, it is intuitive that its data preferences dynamically change, indicating the need for pretraining with different data at various training stages. We propose the Perplexity Difference (PD) based Preference Curriculum learning framework, which always perceives and uses the data preferred by LLMs to train and boost them.
  arXiv  Detail & Related papers  (2025-01-21T13:12:13Z)
• Training Dynamics of a 1.7B LLaMa Model: A Data-Efficient Approach [10.39475177812483]
  We share insights gained from training DMaS-LLaMa-Lite on approximately 20 billion tokens of data. We chronicle the full training trajectory, documenting how evolving validation loss levels and downstream benchmarks reflect transitions from incoherent text to fluent, contextually grounded output. By detailing these experiences and offering training logs, checkpoints, and sample outputs, we aim to guide future researchers and practitioners in refining their pretraining strategies.
  arXiv  Detail & Related papers  (2024-12-17T21:15:52Z)
• What Do Learning Dynamics Reveal About Generalization in LLM Reasoning? [83.83230167222852]
  We find that a model's generalization behavior can be effectively characterized by a training metric we call pre-memorization train accuracy. By connecting a model's learning behavior to its generalization, pre-memorization train accuracy can guide targeted improvements to training strategies.
  arXiv  Detail & Related papers  (2024-11-12T09:52:40Z)
• 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. We build a new statistical framework for data selection centered around estimates of perplexity-benchmark correlations. 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)
• AutoScale: Scale-Aware Data Mixing for Pre-Training LLMs [61.13296177652599]
  We show that data mixtures that perform well at smaller scales may not retain their advantage at larger scales. We propose AutoScale, a two-stage, scale-aware data composition framework.
  arXiv  Detail & Related papers  (2024-07-29T17:06:30Z)
• Efficient Continual Pre-training by Mitigating the Stability Gap [68.49269649759005]
  We study the behavior of Large Language Models (LLMs) during continual pre-training. We propose three effective strategies to enhance LLM performance within a fixed compute budget. Our strategies improve the average medical task performance of the OpenLlama-3B model from 36.2% to 40.7% with only 40% of the original training budget.
  arXiv  Detail & Related papers  (2024-06-21T02:28:37Z)
• The Fine Line: Navigating Large Language Model Pretraining with Down-streaming Capability Analysis [27.310894780313618]
  This paper undertakes a comprehensive comparison of model capabilities at various pretraining intermediate checkpoints. We confirm that specific downstream metrics exhibit similar training dynamics across models of different sizes. In addition to our core findings, we've reproduced Amber and OpenLLaMA, releasing their intermediate checkpoints.
  arXiv  Detail & Related papers  (2024-04-01T16:00:01Z)
• SwiftLearn: A Data-Efficient Training Method of Deep Learning Models using Importance Sampling [3.8330834108666667]
  We present SwiftLearn, a data-efficient approach to accelerate training of deep learning models. This subset is selected based on an importance criteria measured over the entire dataset during warm-up stages. We show that almost 90% of the data can be dropped achieving an end-to-end average speedup of 3.36x while keeping the average accuracy drop less than 0.92%.
  arXiv  Detail & Related papers  (2023-11-25T22:51:01Z)
• D4: Improving LLM Pretraining via Document De-Duplication and Diversification [38.84592304799403]
  We show that careful data selection via pre-trained model embeddings can speed up training. We also show that repeating data intelligently consistently outperforms baseline training.
  arXiv  Detail & Related papers  (2023-08-23T17:58:14Z)
• bert2BERT: Towards Reusable Pretrained Language Models [51.078081486422896]
  We propose bert2BERT, which can effectively transfer the knowledge of an existing smaller pre-trained model to a large model. bert2BERT saves about 45% and 47% computational cost of pre-training BERT_BASE and GPT_BASE by reusing the models of almost their half sizes.
  arXiv  Detail & Related papers  (2021-10-14T04:05:25Z)
• Curriculum Learning: A Regularization Method for Efficient and Stable Billion-Scale GPT Model Pre-Training [18.640076155697415]
  We present a study of a curriculum learning based approach, which helps improve the pre-training convergence speed of autoregressive models. Our evaluations demonstrate that curriculum learning enables training GPT-2 models with 8x larger batch size and 4x larger learning rate.
  arXiv  Detail & Related papers  (2021-08-13T06:32:53Z)
• Graph Sampling Based Deep Metric Learning for Generalizable Person Re-Identification [114.56752624945142]
  We argue that the most popular random sampling method, the well-known PK sampler, is not informative and efficient for deep metric learning. We propose an efficient mini batch sampling method called Graph Sampling (GS) for large-scale metric learning.
  arXiv  Detail & Related papers  (2021-04-04T06:44:15Z)
• EarlyBERT: Efficient BERT Training via Early-bird Lottery Tickets [106.79387235014379]
  EarlyBERT is a general computationally-efficient training algorithm applicable to both pre-training and fine-tuning of large-scale language models. We are the first to identify structured winning tickets in the early stage of BERT training, and use them for efficient training. EarlyBERT easily achieves comparable performance to standard BERT with 3545% less training time.
  arXiv  Detail & Related papers  (2020-12-31T20:38:20Z)
• Neural Semi-supervised Learning for Text Classification Under Large-Scale Pretraining [51.19885385587916]
  We conduct studies on semi-supervised learning in the task of text classification under the context of large-scale LM pretraining. Our work marks an initial step in understanding the behavior of semi-supervised learning models under the context of large-scale pretraining.
  arXiv  Detail & Related papers  (2020-11-17T13:39:05Z)
• To Pretrain or Not to Pretrain: Examining the Benefits of Pretraining on Resource Rich Tasks [25.05882459314221]
  We show that as the number of training examples grow into the millions, the accuracy gap between finetuning BERT-based model and training vanilla LSTM from scratch narrows to within 1%. Our findings indicate that pre-trained models might reach a diminishing return point as the supervised data size increases significantly.
  arXiv  Detail & Related papers  (2020-06-15T18:18:59Z)

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.