https://github.com/thunlp/OPD

Rethinking On-Policy Distillation of Large Language Models: Phenomenology, Mechanism, and Recipe
https://github.com/thunlp/OPD

llms mechanism on-policy-distillation

Last synced: 25 days ago
JSON representation

Rethinking On-Policy Distillation of Large Language Models: Phenomenology, Mechanism, and Recipe

• Host: GitHub
• URL: https://github.com/thunlp/OPD
• Owner: thunlp
• Created: 2026-04-12T11:39:28.000Z (3 months ago)
• Default Branch: main
• Last Pushed: 2026-05-26T13:23:25.000Z (about 1 month ago)
• Last Synced: 2026-05-26T15:26:38.838Z (about 1 month ago)
• Topics: llms, mechanism, on-policy-distillation
• Language: Python
• Homepage:
• Size: 58.6 MB
• Stars: 474
• Watchers: 0
• Forks: 23
• Open Issues: 3
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• awesome-on-policy-distillation - Rethinking OPD - based scripts and top-k teacher–student overlap diagnostics merged upstream. (Frameworks and Implementations / Implementations)
• awesomeopd - OPD - the-badge&logo=github&logoColor=white&labelColor=181717&color=ffd700" alt="Stars"> | 2026.04 | Tsinghua THUNLP | [arXiv 2604.13016](https://arxiv.org/abs/2604.13016) | **Rethinking On-Policy Distillation: Phenomenology, Mechanism & Recipe** | (📚 Surveys, Foundations & Position Papers)