Issues

• Model does not learn with u-mup

  graphcore-research/unit-scaling#88 · Filter by author SwayStar123 opened on May 26, 2025
  2

  

• Add uu.Conv2d

  graphcore-research/unit-scaling#87 · Filter by author SwayStar123 opened on May 24, 2025
  2

• Scaling self-Attention when inference

  graphcore-research/unit-scaling#86 · Filter by author mmorinag127 opened on Apr 9, 2025
  1

• Support DTensor for distributed training

  graphcore-research/unit-scaling#85 · Filter by author rlrs opened on Dec 17, 2024
  1

• Attention Issue for DiT-like model

  graphcore-research/unit-scaling#84 · Filter by author kliyer-ai opened on Dec 10, 2024
  6

• Fix wandb logging bug

  bug
  graphcore-research/unit-scaling#82 · Filter by author thecharlieblake opened on Nov 11, 2024

• Adam and SGD update learning rate scaling.

  graphcore-research/unit-scaling#73 · Filter by author norpadon opened on Oct 14, 2024
  11

• MLP scaling w.r.t. expansion_factor

  graphcore-research/unit-scaling#70 · Filter by author EIFY opened on Sep 6, 2024
  1

• [In Progress] Sharing Unit-Mamba Implementation

  graphcore-research/unit-scaling#68 · Filter by author norikazu99 opened on Aug 25, 2024
  1

• [Question] What is 'mult' parameter?

  graphcore-research/unit-scaling#62 · Filter by author mmorinag127 opened on Aug 1, 2024
  4

  

• [Feature Request] Any plan to support fp8 type at latest torch version?

  graphcore-research/unit-scaling#45 · Filter by author MoFHeka opened on Sep 5, 2023
  5