on-policy-distillation
Train models using token-level KL distillation with teacher models. Use when you want true on-policy distillation (Thinking Machines approach), token-level KL penalty, or training with Qwen/Llama as teacher. Triggers on "on-policy", "token-level KL", "KL penalty", "Thinking Machines", "teacher logprobs", "distillation training".
Installation
Follow the repository instructions for this skill and your agent. Review the files and dependencies before adding it to your workflow.
View source ↗GuidesSkill instructions
Features and purpose
Train models using token-level KL distillation with teacher models. Use when you want true on-policy distillation (Thinking Machines approach), token-level KL penalty, or training with Qwen/Llama as teacher. Triggers on "on-policy", "token-level KL", "KL penalty", "Thinking Machines", "teacher logprobs", "distillation training".
Quick start
Follow the repository instructions for this skill and your agent. Review the files and dependencies before adding it to your workflow.
Guides ↗Examples
Read the usage examples in the source repository.
View source ↗Source history
- Indexed on
- 2026-01-23
- Index updated
- 2026-03-05
- Branch
- main
Index and sync dates describe this directory. View the source repository for code history.
GitHub ↗chat Comments (0)
Sign in to join the discussion and leave a comment.
No comments yet. Be the first to share your thoughts!