Skip to content

Latest commit

 

History

History
22 lines (17 loc) · 898 Bytes

File metadata and controls

22 lines (17 loc) · 898 Bytes

Repository Manifest

This repository contains the Skill-RM reproduction code:

  • reward-judging runners for RewardBench2, JudgeBench, and RM-Bench;
  • skill packages for the standard-input and resource-available settings;
  • resource-use ablation runners;
  • JETTS sequential knockout evaluation code;
  • IF-RewardBench evaluation code;
  • pointwise instruction-following RL recipe for external verl;
  • configuration templates, documentation, tests, and Apache-2.0 metadata.

Runtime inputs and outputs are prepared separately:

  • benchmark data or cached expanded datasets;
  • model checkpoints or generated model outputs;
  • processed RL parquet files;
  • endpoint addresses or local secrets;
  • run logs, validation files, or generated experiment outputs.

Prepare benchmark data in a working directory and point the runners at those paths with environment variables. See docs/data_preparation.md.