ainewsblitz.com

Breaking

THUDM Open-Sources 'slime,' the Reinforcement Learning Framework Used to Post-Train GLM-5.2

THUDM Open-Sources 'slime,' the Reinforcement Learning Framework Used to Post-Train GLM-5.2
  • Open Source
  • Research & Papers
  • Foundation Models

The Tsinghua University Data Mining Group (THUDM) has released slime, the reinforcement learning framework used to post-train its GLM-5.2 model, as open source under the Apache-2.0 license. The move puts a production-grade RL post-training stack—one that has already trained a frontier-class model—into the hands of any developer or research group.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.

$20
Read this article
$29/month
Unlimited — all 27,934 articles, the full archive, and comprehension quizzes
Save 72%
$98/year
≈ $8.17/month
Unlimited, billed once a year