Commits

Breadcrumbs

History for

onmain

User selector

All users

All time

Commit History

Commits on Sep 24, 2025

• Context Parallel w/ Ring & Ulysses & Unified Attention (#11941)

  Show description for dcb6dd9
  

  3 peopleauthored
  dcb6dd9

  Copy full SHA for dcb6dd9

• Fix Custom Code loading (#12378)

  Show description for 7c54a7b
  DN6

  authored
  7c54a7b

  Copy full SHA for 7c54a7b

Commits on Sep 23, 2025

• Allow Automodel to support custom model code (#12353)

  Show description for 80de641
  DN6

  authored
  80de641

  Copy full SHA for 80de641

Commits on Sep 22, 2025

• [Fix] enable_xformers_memory_efficient_attention() in Flux Pipeline (#12337)

  Show description for 78031c2
  SahilCarterr

  authored
  78031c2

  Copy full SHA for 78031c2

• Fix bug with VAE slicing in autoencoder_dc.py (#12343)

  mingyi456

  authored
  d83d35c

  Copy full SHA for d83d35c

Commits on Sep 17, 2025

• Fix many type hint errors (#12289)

  Show description for efb7a29
  DefTruth

  authored
  efb7a29

  Copy full SHA for efb7a29

Commits on Sep 16, 2025

• Fix autoencoder_kl_wan.py bugs for Wan2.2 VAE (#12335)

  Show description for d06750a
  franciszzj

  authored
  d06750a

  Copy full SHA for d06750a

• fixed bug in defining embed dim for UNet1D (#12111)

  Show description for 751e250
  

  SammyAgrawal

  and

  DN6

  authored
  751e250

  Copy full SHA for 751e250

Commits on Sep 3, 2025

• Add AttentionMixin to WanVACETransformer3DModel (#12268)

  Show description for 6682956
  J4BEZ

  authored
  6682956

  Copy full SHA for 6682956

• [core] use `kernels` to support `_flash_3_hub` attention backend (#12236)

  Show description for 130fd8d
  sayakpaul

  authored
  130fd8d

  Copy full SHA for 130fd8d

Commits on Aug 30, 2025

• Improve performance of NPU FA (#12260)

  Show description for 827fad6
  

  leisuzz

  and

  a-r-r-o-w

  authored
  827fad6

  Copy full SHA for 827fad6

Commits on Aug 26, 2025

• Deprecate Flax support (#12151)

  Show description for 532f41c
  

  sayakpaul

  and

  DN6

  authored
  532f41c

  Copy full SHA for 532f41c

• Propose to update & upgrade SkyReels-V2 (#12167)

  Show description for 5fcd5f5
  tolgacangoz

  authored
  5fcd5f5

  Copy full SHA for 5fcd5f5

• NPU attention refactor for FLUX (#12209)

  Show description for 0fd7ee7
  

  3 peopleauthored
  0fd7ee7

  Copy full SHA for 0fd7ee7

Commits on Aug 23, 2025

• Fix PyTorch 2.3.1 compatibility: add version guard for torch.library.… (#12206)

  Show description for 9a7ae77
  

  3 peopleauthored
  9a7ae77

  Copy full SHA for 9a7ae77

Commits on Aug 22, 2025

• add attentionmixin to qwen image (#12219)

  sayakpaul

  authored
  673d435

  Copy full SHA for 673d435

• Support ControlNet for Qwen-Image (#12215)

  Show description for 561ab54
  

  3 peopleauthored
  561ab54

  Copy full SHA for 561ab54

Commits on Aug 20, 2025

• [chore] remove extra validation check in determine_device_map (#12176)

  Show description for 4fcd0bc
  sayakpaul

  authored
  4fcd0bc

  Copy full SHA for 4fcd0bc

• Bria 3 2 pipeline (#12010)

  Show description for 7993be9
  

  3 peopleauthored
  7993be9

  Copy full SHA for 7993be9

Commits on Aug 18, 2025

• fix: caching allocator behaviour for quantization. (#12172)

  Show description for e824660
  

  sayakpaul

  and

  a-r-r-o-w

  authored
  e824660

  Copy full SHA for e824660

• Minor modification to support DC-AE-turbo (#12169)

  Show description for 85cbe58
  chenjy2003

  authored
  85cbe58

  Copy full SHA for 85cbe58

• [qwen] Qwen image edit followups (#12166)

  Show description for 4d9b822
  sayakpaul

  authored
  4d9b822

  Copy full SHA for 4d9b822

• remove silu for CogView4 (#12150)

  Show description for 76c809e
  

  4 peopleauthored
  76c809e

  Copy full SHA for 76c809e

Commits on Aug 17, 2025

• Qwen Image Edit Support (#12164)

  Show description for e682af2
  naykun

  authored
  e682af2

  Copy full SHA for e682af2

Commits on Aug 14, 2025

• support `hf_quantizer` in cache warmup. (#12043)

  Show description for 58bf268
  sayakpaul

  authored
  58bf268

  Copy full SHA for 58bf268

• make parallel loading flag a part of constants. (#12137)

  sayakpaul

  authored
  123506e

  Copy full SHA for 123506e

Commits on Aug 13, 2025

• [core] parallel loading of shards (#12028)

  Show description for baa9b58
  

  sayakpaul

  and

  DN6

  authored
  baa9b58

  Copy full SHA for baa9b58

Commits on Aug 12, 2025

• [Bugfix] typo fix in NPU FA (#12129)

  Show description for 480fb35
  

  leisuzz

  and

  a-r-r-o-w

  authored
  480fb35

  Copy full SHA for 480fb35

Commits on Aug 11, 2025

• enable compilation in qwen image. (#12061)

  Show description for 4a9dbd5
  

  sayakpaul

  and

  a-r-r-o-w

  authored
  4a9dbd5

  Copy full SHA for 4a9dbd5

Commits on Aug 7, 2025

• fix flux type hint (#12089)

  Show description for 4b17fa2
  DefTruth

  authored
  4b17fa2

  Copy full SHA for 4b17fa2

Commits on Aug 4, 2025

• fix(qwen-image): update vae license (#12063)

  Show description for 69a9828
  

  3 peopleauthored
  69a9828

  Copy full SHA for 69a9828

• Cross attention module to Wan Attention (#12058)

  Show description for 11d22e0
  

  3 peopleauthored
  11d22e0

  Copy full SHA for 11d22e0

• tests + minor refactor for QwenImage (#12057)

  Show description for 9a38fab
  a-r-r-o-w

  authored
  9a38fab

  Copy full SHA for 9a38fab

• [wan2.2] follow-up (#12024)

  Show description for cb8e61e
  

  yiyixuxu

  and

  github-actions[bot]

  authored
  cb8e61e

  Copy full SHA for cb8e61e

Commits on Aug 3, 2025

• Qwen-Image (#12055)

  Show description for 8e53cd9
  

  naykun

  and

  yiyixuxu

  authored
  8e53cd9

  Copy full SHA for 8e53cd9