We study recent research advances that improve large language models through efficient pre-training and scaling, and open datasets and tools. We combine these advances to introduce Cerebras-GPT, a family of open compute-optimal language models scaled from 111M to 13B parameters. We train Cerebras-GPT models on the Eleuther Pile dataset following DeepMind Chinchilla scaling rules for efficient pre-training (highest accuracy for a given compute budget). We characterize the predictable power-law scaling and compare Cerebras-GPT with other publicly-available models to show all Cerebras-GPT models have state-of-the-art training efficiency on both pre-training and downstream objectives. We describe our learnings including how Maximal Update Parameterization ($\mu$P) can further improve large model scaling, improving accuracy and hyperparameter predictability at scale. We release our pre-trained models and code, making this paper the first open and reproducible work comparing compute-optimal model scaling to models trained on fixed dataset sizes. Cerebras-GPT models are available on HuggingFace: https://huggingface.co/cerebras.

翻译：我们研究了近期通过高效预训练、扩展以及开放数据集与工具来改进大型语言模型的研究进展。结合这些进展，我们推出了Cerebras-GPT系列——一组从1.11亿到130亿参数的开源计算最优语言模型。我们遵循DeepMind Chinchilla缩放规则，在Eleuther Pile数据集上训练Cerebras-GPT模型，以实现高效的预训练（在给定计算预算下达到最高精度）。我们刻画了可预测的幂律缩放行为，并将Cerebras-GPT与其他公开模型进行对比，结果表明所有Cerebras-GPT模型在预训练及下游任务目标上均达到了最先进的训练效率。我们总结了关键经验，包括最大更新参数化（$\mu$P）如何进一步提升大规模模型扩展性能，以及如何在扩展过程中提升精度与超参数可预测性。我们发布了预训练模型和代码，使本文成为首个开源且可复现的、对比计算最优模型缩放与固定数据集大小训练模型的研究。Cerebras-GPT模型可在HuggingFace获取：https://huggingface.co/cerebras。