Метка: mid-training

minnesotanlp/Finch-8B

Evolution Fine-Tuning: Learning to Discover Across 371 Optimization Tasks A mid-training «practice phase» that teaches small open-source LLMs...