Language models scale reliably with over-training and on downstream tasks
- URL: http://arxiv.org/abs/2403.08540v2
- Date: Fri, 14 Jun 2024 20:21:05 GMT
- Title: Language models scale reliably with over-training and on downstream tasks
- Authors: Samir Yitzhak Gadre, Georgios Smyrnis, Vaishaal Shankar, Suchin Gururangan, Mitchell Wortsman, Rulin Shao, Jean Mercat, Alex Fang, Jeffrey Li, Sedrick Keh, Rui Xin, Marianna Nezhurina, Igor Vasiljevic, Jenia Jitsev, Luca Soldaini, Alexandros G. Dimakis, Gabriel Ilharco, Pang Wei Koh, Shuran Song, Thomas Kollar, Yair Carmon, Achal Dave, Reinhard Heckel, Niklas Muennighoff, Ludwig Schmidt,
- Abstract summary: Scaling laws are useful guides for derisking expensive training runs.
However, there remain gaps between current studies and how language models are trained.
In contrast, scaling laws mostly predict loss on inference, but models are usually compared on downstream task performance.
- Score: 121.69867718185125
- License: http://creativecommons.org/licenses/by/4.0/
- Abstract: Scaling laws are useful guides for derisking expensive training runs, as they predict performance of large models using cheaper, small-scale experiments. However, there remain gaps between current scaling studies and how language models are ultimately trained and evaluated. For instance, scaling is usually studied in the compute-optimal training regime (i.e., "Chinchilla optimal" regime). In contrast, models are often over-trained to reduce inference costs. Moreover, scaling laws mostly predict loss on next-token prediction, but models are usually compared on downstream task performance. To address both shortcomings, we create a testbed of 104 models with 0.011B to 6.9B parameters trained with various numbers of tokens on three data distributions. First, we fit scaling laws that extrapolate in both the amount of over-training and the number of model parameters. This enables us to predict the validation loss of a 1.4B parameter, 900B token run (i.e., 32$\times$ over-trained) and a 6.9B parameter, 138B token run (i.e., a compute-optimal run)$\unicode{x2014}$each from experiments that take 300$\times$ less compute. Second, we relate the perplexity of a language model to its downstream task performance by proposing a power law. We use this law to predict top-1 error averaged over downstream tasks for the two aforementioned models, using experiments that take 20$\times$ less compute. Our experiments are available at https://github.com/mlfoundations/scaling.
Related papers
- A Hitchhiker's Guide to Scaling Law Estimation [56.06982415792523]
Scaling laws predict the loss of a target machine learning model by extrapolating from easier-to-train models with fewer parameters or smaller training sets.
We estimate more than 1000 scaling laws, then derive a set of best practices for estimating scaling laws in new model families.
arXiv Detail & Related papers (2024-10-15T17:59:10Z) - A Dynamical Model of Neural Scaling Laws [79.59705237659547]
We analyze a random feature model trained with gradient descent as a solvable model of network training and generalization.
Our theory shows how the gap between training and test loss can gradually build up over time due to repeated reuse of data.
arXiv Detail & Related papers (2024-02-02T01:41:38Z) - nanoLM: an Affordable LLM Pre-training Benchmark via Accurate Loss Prediction across Scales [65.01417261415833]
We present an approach to predict the pre-training loss based on our observations that Maximal Update Parametrization (muP) enables accurate fitting of scaling laws.
With around 14% of the one-time pre-training cost, we can accurately forecast the loss for models up to 52B.
Our goal with nanoLM is to empower researchers with limited resources to reach meaningful conclusions on large models.
arXiv Detail & Related papers (2023-04-14T00:45:01Z) - 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) - Training Compute-Optimal Large Language Models [54.00424650998489]
We train language models ranging from 70 million to over 16 billion parameters on 5 to 500 billion tokens.
For compute-optimal training, the model size and the number of training tokens should be scaled equally.
chinchilla uniformly and significantly outperforms Gopher (280B), GPT-3 (175B), Jurassic-1 (178B), and Megatron-Turing NLG (530B)
arXiv Detail & Related papers (2022-03-29T13:38:03Z) - Scaling Laws Under the Microscope: Predicting Transformer Performance
from Small Scale Experiments [42.793379799720434]
We investigate whether scaling laws can be used to accelerate model development.
We find that scaling laws emerge at finetuning time in some NLP tasks.
For tasks where scaling laws exist, they can be used to predict the performance of larger models.
arXiv Detail & Related papers (2022-02-13T19:13:00Z)
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.