We investigate the optimal model size and number of tokens for training a
transformer language model under a given compute budget. We find that current
large language models are significantly undertrained, a consequence of the
recent focus on scaling language models whilst keeping the amount of training
data constant. By training over 400 language models ranging from 70 million to
over 16 billion parameters on 5 to 500 billion tokens, we find that for
compute-optimal training, the model size and the number of training tokens
should be scaled equally: for every doubling of model size the number of
training tokens should also be doubled. We test this hypothesis by training a
predicted compute-optimal model, Chinchilla, that uses the same compute budget
as Gopher but with 70B parameters and 4$\times$ more more data. Chinchilla
uniformly and significantly outperforms Gopher (280B), GPT-3 (175B), Jurassic-1
(178B), and Megatron-Turing NLG (530B) on a large range of downstream
evaluation tasks. This also means that Chinchilla uses substantially less
compute for fine-tuning and inference, greatly facilitating downstream usage.
As a highlight, Chinchilla reaches a state-of-the-art average accuracy of 67.5%
on the MMLU benchmark, greater than a 7% improvement over Gopher.

本文研究在给定计算预算下训练变换器语言模型的最佳模型大小和标记数量，并发现对于计算优化的训练，模型大小和训练标记数量应该等量缩放。同时测试了此假设通过训练一个预测计算最优模型的 Chinchilla，该模型在大范围的下游评估任务中均匀且显著地优于 Gopher，GPT-3，Jurassic-1 和 Megatron-Turing NLG。