Issues

• Question about synthetic preference generation for continual RLHF

  ComplexData-MILA/AIF-Gen#211 · Filter by author aus10mccaffs2026 opened on Jul 14, 2026

• Make "chosen" vs. "reject" divergence more obvious semantically

  ComplexData-MILA/AIF-Gen#205 · Filter by author EMZEDI opened on Jul 21, 2025
  

• Multi-gpu PPO

  bug
  ComplexData-MILA/AIF-Gen#194 · Filter by author Jacob-Chmura opened on May 7, 2025

• HF Dataset and Model cards and Croissant files

  ComplexData-MILA/AIF-Gen#139 · Filter by author EMZEDI opened on Apr 25, 2025 · 0.1.0 Release
  

• Integrate PPO training with vllm for generation [if possible]

  ComplexData-MILA/AIF-Gen#95 · Filter by author EMZEDI opened on Mar 13, 2025

• evaluate, log, and save_model functions unified implementation across all continual RLHF methods

  enhancement
  Task

  ComplexData-MILA/AIF-Gen#83 · Filter by author EMZEDI opened on Mar 7, 2025 · 0.1.1 Release
  

• Sparse Sequential Merging Algorithm

  ComplexData-MILA/AIF-Gen#78 · Filter by author EMZEDI opened on Mar 6, 2025
  1

  

• Add benchmarks/common or a class for common functionalities between the Continual Algorithms

  benchmarks
  Feature

  ComplexData-MILA/AIF-Gen#77 · Filter by author EMZEDI opened on Mar 6, 2025 · 0.1.1 Release
  1

  

• Clustering and Visualization for diversity | Paper write-up priority

  enhancement
  Feature

  ComplexData-MILA/AIF-Gen#72 · Filter by author EMZEDI opened on Feb 28, 2025 · 0.1.1 Release