The Diminishing Returns of Masked Language Models to Science

• URL: http://arxiv.org/abs/2205.11342v2
• Date: Wed, 3 May 2023 15:21:40 GMT
• Title: The Diminishing Returns of Masked Language Models to Science
• Authors: Zhi Hong, Aswathy Ajith, Gregory Pauloski, Eamon Duede, Kyle Chard, Ian Foster
• Abstract summary: We evaluate the impact of training data, model size, pretraining and finetuning time on 12 downstream scientific tasks. We find that increasing model sizes, training data, or compute time does not always lead to significant improvements.
• Score: 0.7549732580284559
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Transformer-based masked language models such as BERT, trained on general corpora, have shown impressive performance on downstream tasks. It has also been demonstrated that the downstream task performance of such models can be improved by pretraining larger models for longer on more data. In this work, we empirically evaluate the extent to which these results extend to tasks in science. We use 14 domain-specific transformer-based models (including ScholarBERT, a new 770M-parameter science-focused masked language model pretrained on up to 225B tokens) to evaluate the impact of training data, model size, pretraining and finetuning time on 12 downstream scientific tasks. Interestingly, we find that increasing model sizes, training data, or compute time does not always lead to significant improvements (i.e., >1% F1), if at all, in scientific information extraction tasks and offered possible explanations for the surprising performance differences.

Related papers

• Uni-Mol2: Exploring Molecular Pretraining Model at Scale [27.172011090947823]
  We present Uni-Mol2, an innovative molecular pretraining model that integrates features at the atomic level, graph level, and geometry structure level. We successfully scale Uni-Mol2 to 1.1 billion parameters through pretraining on 800 million conformations, making it the largest molecular pretraining model to date.
  arXiv  Detail & Related papers  (2024-06-21T08:28:54Z)
• 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)
• Towards Foundation Models for Scientific Machine Learning: Characterizing Scaling and Transfer Behavior [32.74388989649232]
  We study how pre-training could be used for scientific machine learning (SciML) applications. We find that fine-tuning these models yields more performance gains as model size increases.
  arXiv  Detail & Related papers  (2023-06-01T00:32:59Z)
• INGENIOUS: Using Informative Data Subsets for Efficient Pre-Training of Language Models [40.54353850357839]
  We show how we can employ submodular optimization to select highly representative subsets of the training corpora. We show that the resulting models achieve up to $sim99%$ of the performance of the fully-trained models.
  arXiv  Detail & Related papers  (2023-05-11T09:24:41Z)
• Multi-granulariy Time-based Transformer for Knowledge Tracing [9.788039182463768]
  We leverage students historical data, including their past test scores, to create a personalized model for each student. We then use these models to predict their future performance on a given test.
  arXiv  Detail & Related papers  (2023-04-11T14:46:38Z)
• Exploring the Impact of Instruction Data Scaling on Large Language Models: An Empirical Study on Real-World Use Cases [17.431381376675432]
  In this paper we explore the performance of large language models based on instruction tuning across different scales of instruction data. With Bloomz-7B1-mt as the base model, the results show that merely increasing the amount of instruction data leads to continuous improvement in tasks such as open-ended generation. We propose potential future research directions such as effectively selecting high-quality training data, scaling base models and training methods specialized for hard tasks.
  arXiv  Detail & Related papers  (2023-03-26T14:49:37Z)
• The effectiveness of MAE pre-pretraining for billion-scale pretraining [65.98338857597935]
  We introduce an additional pre-pretraining stage that is simple and uses the self-supervised MAE technique to initialize the model. We measure the effectiveness of pre-pretraining on 10 different visual recognition tasks spanning image classification, video recognition, object detection, low-shot classification and zero-shot recognition.
  arXiv  Detail & Related papers  (2023-03-23T17:56:12Z)
• METRO: Efficient Denoising Pretraining of Large Scale Autoencoding Language Models with Model Generated Signals [151.3601429216877]
  We present an efficient method of pretraining large-scale autoencoding language models using training signals generated by an auxiliary model. We propose a recipe, namely "Model generated dEnoising TRaining Objective" (METRO) The resultant models, METRO-LM, consisting of up to 5.4 billion parameters, achieve new state-of-the-art on the GLUE, SuperGLUE, and SQuAD benchmarks.
  arXiv  Detail & Related papers  (2022-04-13T21:39:15Z)
• MoEfication: Conditional Computation of Transformer Models for Efficient Inference [66.56994436947441]
  Transformer-based pre-trained language models can achieve superior performance on most NLP tasks due to large parameter capacity, but also lead to huge computation cost. We explore to accelerate large-model inference by conditional computation based on the sparse activation phenomenon. We propose to transform a large model into its mixture-of-experts (MoE) version with equal model size, namely MoEfication.
  arXiv  Detail & Related papers  (2021-10-05T02:14:38Z)
• STAR: Sparse Transformer-based Action Recognition [61.490243467748314]
  This work proposes a novel skeleton-based human action recognition model with sparse attention on the spatial dimension and segmented linear attention on the temporal dimension of data. Experiments show that our model can achieve comparable performance while utilizing much less trainable parameters and achieve high speed in training and inference.
  arXiv  Detail & Related papers  (2021-07-15T02:53:11Z)
• Models, Pixels, and Rewards: Evaluating Design Trade-offs in Visual Model-Based Reinforcement Learning [109.74041512359476]
  We study a number of design decisions for the predictive model in visual MBRL algorithms. We find that a range of design decisions that are often considered crucial, such as the use of latent spaces, have little effect on task performance. We show how this phenomenon is related to exploration and how some of the lower-scoring models on standard benchmarks will perform the same as the best-performing models when trained on the same training data.
  arXiv  Detail & Related papers  (2020-12-08T18:03:21Z)

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.