Characterizing Learning Curves During Language Model Pre-Training: Learning, Forgetting, and Stability

• URL: http://arxiv.org/abs/2308.15419v2
• Date: Tue, 30 Jul 2024 21:28:36 GMT
• Title: Characterizing Learning Curves During Language Model Pre-Training: Learning, Forgetting, and Stability
• Authors: Tyler A. Chang, Zhuowen Tu, Benjamin K. Bergen,
• Abstract summary: We observe that language models generate short repetitive phrases before learning to generate longer and more coherent text. Individual tokens often exhibit sudden increases or decreases in loss that are surprisingly consistent across pre-training runs. More frequent tokens reach lower final surprisals, exhibit less variability within and across pre-training runs, are learned earlier, and are less likely to be "forgotten" during pre-training.
• Score: 25.52470575274251
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: How do language models learn to make predictions during pre-training? To study this, we extract learning curves from five autoregressive English language model pre-training runs, for 1M unseen tokens in context. We observe that the language models generate short repetitive phrases before learning to generate longer and more coherent text. We also find that individual tokens often exhibit sudden increases or decreases in loss that are surprisingly consistent across pre-training runs. To better understand these fluctuations, we quantify the final surprisal, within-run variability, age of acquisition, forgettability, and cross-run variability of learning curves for individual tokens in context. More frequent tokens reach lower final surprisals, exhibit less variability within and across pre-training runs, are learned earlier, and are less likely to be "forgotten" during pre-training. Higher n-gram probabilities further accentuate these effects. Independent of the target token, shorter and more frequent contexts correlate with marginally more stable and quickly acquired predictions. Based on our results, we argue for the existence of sequential learning dependencies between different model capabilities, and we characterize language model learning as early n-gram learning before gradual refinement of tail n-gram predictions.

Related papers

• Frequency Explains the Inverse Correlation of Large Language Models' Size, Training Data Amount, and Surprisal's Fit to Reading Times [15.738530737312335]
  Recent studies have shown that as Transformer-based language models become larger and are trained on very large amounts of data, the fit of their surprisal estimates to naturalistic human reading times degrades. This paper presents a series of analyses showing that word frequency is a key explanatory factor underlying these two trends. The results indicate that Transformer-based language models' surprisal estimates diverge from human-like expectations due to the superhumanly complex associations they learn for predicting rare words.
  arXiv  Detail & Related papers  (2024-02-03T20:22:54Z)
• Are We Falling in a Middle-Intelligence Trap? An Analysis and Mitigation of the Reversal Curse [73.65112477688353]
  Recent studies have highlighted a phenomenon in large language models known as "the reversal curse" We contend that the reversal curse is partially a result of specific model training objectives. We propose a novel training method, BI Casual language modeling Optimization (BICO), designed to mitigate the reversal curse.
  arXiv  Detail & Related papers  (2023-11-13T17:01:12Z)
• MiLe Loss: a New Loss for Mitigating the Bias of Learning Difficulties in Generative Language Models [40.992566245706996]
  We propose a MiLe Loss function for mitigating the bias of learning difficulties with tokens. We train generative language models at different scales of 468M, 1.2B, and 6.7B parameters. Experiments reveal that models incorporating the proposed MiLe Loss can gain consistent performance improvement on downstream benchmarks.
  arXiv  Detail & Related papers  (2023-10-30T13:33:21Z)
• Uncertainty-Aware Unlikelihood Learning Improves Generative Aspect Sentiment Quad Prediction [52.05304897163256]
  We propose a template-agnostic method to control the token-level generation. Specifically, we introduce Monte Carlo dropout to understand the built-in uncertainty of pre-trained language models. We further propose marginalized unlikelihood learning to suppress the uncertainty-aware mistake tokens.
  arXiv  Detail & Related papers  (2023-06-01T07:49:06Z)
• Training Trajectories of Language Models Across Scales [99.38721327771208]
  Scaling up language models has led to unprecedented performance gains. How do language models of different sizes learn during pre-training? Why do larger language models demonstrate more desirable behaviors?
  arXiv  Detail & Related papers  (2022-12-19T19:16:29Z)
• Language Model Pre-Training with Sparse Latent Typing [66.75786739499604]
  We propose a new pre-training objective, Sparse Latent Typing, which enables the model to sparsely extract sentence-level keywords with diverse latent types. Experimental results show that our model is able to learn interpretable latent type categories in a self-supervised manner without using any external knowledge.
  arXiv  Detail & Related papers  (2022-10-23T00:37:08Z)
• What do Large Language Models Learn beyond Language? [10.9650651784511]
  We find that pretrained models significantly outperform comparable non-pretrained neural models. Experiments surprisingly reveal that the positive effects of pre-training persist even when pretraining on multi-lingual text or computer code. Our findings suggest a hitherto unexplored deep connection between pre-training and inductive learning abilities of language models.
  arXiv  Detail & Related papers  (2022-10-21T23:43:13Z)
• Token-wise Curriculum Learning for Neural Machine Translation [94.93133801641707]
  Existing curriculum learning approaches to Neural Machine Translation (NMT) require sufficient sampling amounts of "easy" samples from training data at the early training stage. We propose a novel token-wise curriculum learning approach that creates sufficient amounts of easy samples. Our approach can consistently outperform baselines on 5 language pairs, especially for low-resource languages.
  arXiv  Detail & Related papers  (2021-03-20T03:57:59Z)
• Pretrained Language Model Embryology: The Birth of ALBERT [68.5801642674541]
  We investigate the developmental process from a set of randomly parameters to a totipotent language model. Our results show that ALBERT learns to reconstruct and predict tokens of different parts of speech (POS) in different learning speeds during pretraining. These findings suggest that knowledge of a pretrained model varies during pretraining, and having more pretrain steps does not necessarily provide a model with more comprehensive knowledge.
  arXiv  Detail & Related papers  (2020-10-06T05:15:39Z)

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.