scaling-laws-optimal-data-mixtures

【数据与研究】Predict optimal data mixture proportions for multi-domain LLM training using scaling laws that require only 10-20 small experiments. Determine which domains should contribute how much data across model sizes (186M-7B), reducing computational waste in foundation model pretraining.

独立 GitHub 快照 · 自动分类 · 以原作者文档为准