5 Pull requests merged by 3 people

• Add time to Slurm

  #639 merged May 9, 2025

• Use pass@1 for all evals

  #633 merged May 9, 2025

• soft_overlong_punishment from DAPO paper

  #638 merged May 9, 2025

• Fix style again :)

  #636 merged May 8, 2025

• Code Execution using Morph Cloud

  #614 merged May 8, 2025

1 Pull request opened by 1 person

• Add dataset filtering script

  #637 opened May 9, 2025

1 Issue closed by 1 person

• Release 32B math-220k supervised fine-tuned weights

  #634 closed May 8, 2025

6 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• Sequence length problem

  #579 commented on May 8, 2025 • 0 new comments

• Reproducing GRPO based on Qwen2.5-1.5B-Instruct and using Math-220K dataset Yields Unexpected Results

  #538 commented on May 9, 2025 • 0 new comments

• The kl divergence collapses but the format reward becomes larger

  #373 commented on May 9, 2025 • 0 new comments

• OpenR1-Qwen-7B achieves 47.40 on AIME24, better than reported!

  #622 commented on May 9, 2025 • 0 new comments

• When I run the GRPO demo, I find that format_reward is always 0！！！

  #235 commented on May 9, 2025 • 0 new comments

• [WIP] R1-Zero-like experiments

  #569 commented on May 9, 2025 • 0 new comments