Commits

Breadcrumbs

History for

onmain

User selector

All users

All time

Commit history

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
  

  3 peopleauthored
  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

Commits on Aug 2, 2025

• Update autoencoder_kl_cosmos.py (#12045)

  Show description for 359b605
  

  3 peopleauthored
  359b605

  Copy full SHA for 359b605

• Fix type of force_upcast to bool (#12046)

  BerndDoser

  authored
  6febc08

  Copy full SHA for 6febc08

Commits on Aug 1, 2025

• [wan2.2] fix vae patches (#12041)

  Show description for 58d2b10
  yiyixuxu

  authored
  58d2b10

  Copy full SHA for 58d2b10

Commits on Jul 30, 2025

• [core] support attention backends for LTX (#12021)

  Show description for c052791
  

  sayakpaul

  and

  a-r-r-o-w

  authored
  c052791

  Copy full SHA for c052791

Commits on Jul 29, 2025

• Fix huggingface-hub failing tests (#11994)

  Show description for edcbe80
  asomoza

  authored
  edcbe80

  Copy full SHA for edcbe80

• [refactor] Wan single file implementation (#11918)

  Show description for c02c4a6
  

  a-r-r-o-w

  and

  DN6

  authored
  c02c4a6

  Copy full SHA for c02c4a6

Commits on Jul 28, 2025

• [WIP] Wan2.2 (#12004)

  Show description for a6d9f6a
  

  yiyixuxu

  and

  bghira

  authored
  a6d9f6a

  Copy full SHA for a6d9f6a

Commits on Jul 22, 2025

• Remove logger warnings for attention backends and hard error during runtime instead (#11967)

  Show description for e46e139
  a-r-r-o-w

  authored
  e46e139

  Copy full SHA for e46e139

• fix "Expected all tensors to be on the same device, but found at least two devices" error (#11690)

  Show description for 1472516
  

  yao-matrix

  and

  sayakpaul

  authored
  1472516

  Copy full SHA for 1472516

Commits on Jul 17, 2025

• [refactor] Flux/Chroma single file implementation + Attention Dispatcher (#11916)

  Show description for 18c8f10
  

  a-r-r-o-w

  and

  DN6

  authored
  18c8f10

  Copy full SHA for 18c8f10

Commits on Jul 16, 2025

• Add SkyReels V2: Infinite-Length Film Generative Model (#11518)

  Show description for 7298bdd
  

  3 peopleauthored
  7298bdd

  Copy full SHA for 7298bdd

• Fixed bug: Uncontrolled recursive calls that caused an infinite loop when loading certain pipelines containing Transformer2DModel (#11923)

  Show description for c5d6e0b
  

  lengmo1996

  and

  asomoza

  authored
  c5d6e0b

  Copy full SHA for c5d6e0b

Commits on Jul 15, 2025

• Remove device synchronization when loading weights (#11927)

  Show description for b73c738
  a-r-r-o-w

  authored
  b73c738

  Copy full SHA for b73c738

Commits on Jul 11, 2025

• Flux: pass joint_attention_kwargs when using gradient_checkpointing (#11814)

  Show description for 3c8b67b
  piercus

  authored
  3c8b67b

  Copy full SHA for 3c8b67b

• Speedup model loading by 4-5x ⚡ (#11904)

  Show description for c903527
  

  a-r-r-o-w

  and

  SunMarc

  authored
  c903527

  Copy full SHA for c903527

Commits on Jul 10, 2025

• Avoid creating tensor in CosmosAttnProcessor2_0 (#11761) (#11763)

  Show description for 941b7fc
  

  chenxiao111222

  and

  yiyixuxu

  authored
  941b7fc

  Copy full SHA for 941b7fc

• [ControlnetUnion] Multiple Fixes (#11888)

  Show description for 76a62ac
  

  asomoza

  and

  hlky

  authored
  76a62ac

  Copy full SHA for 76a62ac

Commits on Jul 8, 2025

• First Block Cache (#11180)

  Show description for 0454fbb
  

  a-r-r-o-w

  and

  DN6

  authored
  0454fbb

  Copy full SHA for 0454fbb

Commits on Jul 1, 2025

• Use real-valued instead of complex tensors in Wan2.1 RoPE (#11649)

  Show description for 62e847d
  

  3 peopleauthored
  62e847d

  Copy full SHA for 62e847d