GLM-4-32B-0414: New MIT-licensed SOTA LLM from Zhipu AI

GLM-4-32B-0414: New MIT-licensed SOTA LLM from Zhipu AI(huggingface.co)

3 points by wizee 1 year ago | 1 comment

wizee 1 year ago |

In my experiments, its non-reasoning variant is the strongest non-reasoning LLM in its size class, and outperforms much larger LLMs like Llama 4 Maverick in programming and mechanical engineering problems.

wizee 1 year ago |

Demo available at https://chat.z.ai/

In my experiments, its non-reasoning variant is the strongest non-reasoning LLM in its size class, and outperforms much larger LLMs like Llama 4 Maverick in programming and mechanical engineering problems.