Scaling laws for neural language models #507
Labels
Algorithms
Sorting, Learning or Classifying. All algorithms go here.
llm-experiments
experiments with large language models
MachineLearning
ML Models, Training and Inference
Papers
Research papers
Research
personal research notes for a topic
TITLE: Scaling laws for neural language models
DESCRIPTION:
Abstract
We study empirical scaling laws for language model performance on the cross-entropy loss. The loss scales as a power-law with model size, dataset size, and the amount of compute used for training, with some trends spanning more than seven orders of magnitude. Other architectural details such as network width or depth have minimal effects within a wide range. Simple equations govern the dependence of overfitting on model/dataset size and the dependence of training speed on model size. These relationships allow us to determine the optimal allocation of a fixed compute budget. Larger models are significantly more sample-efficient, such that optimally compute-efficient training involves training very large models on a relatively modest amount of data and stopping significantly before convergence.
URL: https://openai.com/research/scaling-laws-for-neural-language-models
Suggested labels
{ "label-name": "Scaling-Laws", "description": "Analysis of scaling laws for language models", "confidence": 69.89 }
The text was updated successfully, but these errors were encountered: