GLM Z1 9B 查看AI模型的详细信息和价格

模型介绍

GLM-Z1-9B-0414 is a 9B-parameter language model developed by THUDM as part of the GLM-4 family. It incorporates techniques originally applied to larger GLM-Z1 models, including extended reinforcement learning, pairwise ranking alignment, and training on reasoning-intensive tasks such as mathematics, code, and logic. Despite its smaller size, it demonstrates strong performance on general-purpose reasoning tasks and outperforms many open-source models in its weight class.