RepoJournal
Hugging Face

@huggingface

Transformers, Datasets, and the open AI-model layer

Pick a date

  1. Wed 20 may
  2. Thu 21 may
  3. Sat 23 may
  4. Sun 24 may
  5. Mon 25 may
  6. Tue 26 may
  7. Wed 27 may
  8. Thu 28 may
  9. Fri 29 may
  10. Sat 30 may
  11. Sun 31 may
  12. Mon 1 jun
  13. Mon 8 jun
  14. Tue 9 jun
  15. Wed 10 jun
  16. Thu 11 jun
  17. Fri 12 jun
  18. Sat 13 jun
  19. Sun 14 jun
  20. Mon 15 jun
  21. Tue 16 jun
  22. Wed 17 jun
  23. Thu 18 jun
  24. Sat 27 jun
  25. Sun 28 jun
  26. Mon 29 jun
  27. Tue 30 jun
  28. Wed 1 jul
  29. Thu 2 jul
  30. Today 3 jul

The Wire · Showcase

DIFFUSERS FIXES KOHYA LORA COMPATIBILITY BUG BLOCKING MODEL LOADING

By RepoJournal · Filed · About Hugging Face

Kohya-format LoRA weights for conv_in, conv_out, and time_embedding layers were silently failing to load in diffusers, now fixed.

The diffusers library was dropping Kohya-format UNet LoRA keys on three critical top-level submodules [1], causing them to be reported as unexpected rather than applied to the model. This affected anyone fine-tuning these layers with Kohya-ss tooling and trying to load checkpoints into diffusers. The fix adds proper key conversion mapping so lora_unet_conv_in and similar patterns now correctly map to diffusers naming conventions [2]. Meanwhile on the transformers side, auto-mappings registration for remote code is getting fixed [3] to handle model loading edge cases that surfaced in earlier patches. The transformers team also hardened device pinning in GDN mixer layers [4], preventing unexpected cuda:0 allocation when target devices were specified elsewhere. Swift-transformers added Sendable conformance to Hub.Repo and Hub.RepoType [5], unblocking Swift 6 strict concurrency checks for anyone caching or passing these types across task boundaries.

Action items

References

  1. [1] Fix Kohya UNet LoRA key conversion for conv_in/conv_out/time_embedding (#14006) huggingface/diffusers
  2. [2] Fix Kohya UNet LoRA key conversion for conv_in/conv_out/time_embedding ↗ huggingface/diffusers
  3. [3] Fix auto-mappings registration for remote code & fixes a few custom code issues ↗ huggingface/transformers
  4. [4] Don't pin the gated delta net norm to `cuda:0` with a hardcoded device (#46817) huggingface/transformers
  5. [5] Add Sendable conformance to Hub.Repo and Hub.RepoType (#374) huggingface/swift-transformers

FAQ

What changed in Hugging Face on June 29, 2026?
Kohya-format LoRA weights for conv_in, conv_out, and time_embedding layers were silently failing to load in diffusers, now fixed.
What should Hugging Face teams do about it?
If using Kohya LoRA fine-tunes with diffusers, pull the latest version to restore conv_in/conv_out/time_embedding loading • Review transformers remote code loading paths if you're using custom model implementations • Swift developers: update swift-transformers to use Repo in concurrent contexts without warnings
Which Hugging Face repositories shipped on June 29, 2026?
huggingface/diffusers, huggingface/transformers, huggingface/swift-transformers

Related across the cluster

For your repos

The showcase is a teaser.
Your wire is the product.

Same engine. Different stack. Below: what changes when the wire is yours.

Showcase wire

  • •14 famous open source orgs
  • •One wire per day
  • •Public, generic
  • •Read on the web, when you remember

Your wire

  • →Up to 1,500 of your repos - orgs, deps, vendors
  • →Morning and evening briefs
  • →Action items routed to your team
  • →Slack delivery, email, breaking-news CVE alerts

Want a hands-on demo first? Ask a current user for an invite link.