Baseten adds GLM-5.2 fine-tuning and production deployment in Loops SDK

baseten · x · 2026-07-24

Baseten says users can fine-tune GLM-5.2 on the Loops SDK and deploy the resulting checkpoint to a production inference endpoint.

The company adds that SFT runs at 262K sequence length, matching the model’s long-horizon capabilities, and says RL support for GLM-5.2 is coming to Loops soon.

Original post →

More from Infra

Infra channel →