GLM

This is the landing page for GLM model guidance in NeMo RL. It links to version-specific subpages covering validated recipes, recommended training and generation settings, and known issues for each supported version.

For the full list of supported GLM models, see Model Support.

Version Guides

• GLM-5 — GLM-5.1 and GLM-5.2 GRPO recipes on the Megatron backend, including 131K-token training capacity, colocated and non-colocated vLLM, and cuDNN or TileLang DSA kernels.

Subpages for future GLM versions can be added here as distinct, recipe-backed guidance becomes available. Until then, the GRPO guide and recipe YAML files remain the source of truth for other supported GLM models.