Commits

Breadcrumbs

History for

onv0.27.2-patch

User selector

All users

All time

Commit History

Commits on Mar 20, 2024

• [LoRA] fix `cross_attention_kwargs` problems and tighten tests (#7388)

  Show description for 84bc0e4
  sayakpaul

  committed
  84bc0e4

  Copy full SHA for 84bc0e4

Commits on Mar 19, 2024

• [LoRA] pop the LoRA scale so that it doesn't get propagated to the weeds (#7338)

  Show description for 1ab57b6
  

  sayakpaul

  and

  yiyixuxu

  committed
  1ab57b6

  Copy full SHA for 1ab57b6

Commits on Mar 13, 2024

• Fix typos in `UNet2DConditionModel` documentation (#7291)

  Show description for ee6a3a9
  alexanderbonnet

  authored
  ee6a3a9

  Copy full SHA for ee6a3a9

• Add single file support for Stable Cascade (#7274)

  Show description for ed224f9
  

  DN6

  and

  sayakpaul

  authored
  ed224f9

  Copy full SHA for ed224f9

• [LoRA] use the PyTorch classes wherever needed and start depcrecation cycles (#7204)

  Show description for 531e719
  

  sayakpaul

  and

  yiyixuxu

  authored
  531e719

  Copy full SHA for 531e719

• [Chore] switch to `logger.warning` (#7289)

  Show description for 4fbd310
  sayakpaul

  authored
  4fbd310

  Copy full SHA for 4fbd310

Commits on Mar 9, 2024

• [Chore] remove tf mention (#7245)

  Show description for ec64f37
  sayakpaul

  authored
  ec64f37

  Copy full SHA for ec64f37

• [UNet_Spatio_Temporal_Condition] fix default num_attention_heads in unet_spatio_temporal_condition (#7205)

  Show description for 6f2b310
  

  Wang-Xiaodong1899

  and

  sayakpaul

  authored
  6f2b310

  Copy full SHA for 6f2b310

Commits on Mar 8, 2024

• Make mid block optional for flax UNet (#7083)

  Show description for b33b64f
  mar-muel

  authored
  b33b64f

  Copy full SHA for b33b64f

• [Core] throw error when patch inputs and layernorm are provided for Transformers2D (#7200)

  Show description for 8e19c07
  sayakpaul

  authored
  8e19c07

  Copy full SHA for 8e19c07

Commits on Mar 7, 2024

• fix: remove duplicated code in TemporalBasicTransformerBlock. (#7212)

  Show description for ada3bb9
  

  SanftMonster

  and

  sayakpaul

  authored
  ada3bb9

  Copy full SHA for ada3bb9

Commits on Mar 6, 2024

• [Pipiline] Wuerstchen v3 aka Stable Cascasde pipeline (#6487)

  Show description for 40aa47b
  

  8 peopleauthored
  40aa47b

  Copy full SHA for 40aa47b

Commits on Mar 4, 2024

• Fix: UNet2DModel::__init__ type hints; fixes issue #4806 (#7175)

  Show description for df8559a
  

  fpgaminer

  and

  sayakpaul

  authored
  df8559a

  Copy full SHA for df8559a

Commits on Mar 3, 2024

• Fix PixArt 256px inference (#6789)

  Show description for f55873b
  

  3 peopleauthored
  f55873b

  Copy full SHA for f55873b

Commits on Feb 28, 2024

• [docs] unet type hints (#7134)

  Show description for abd922b
  a-r-r-o-w

  authored
  abd922b

  Copy full SHA for abd922b

• refactor: move model helper function in pipeline to a mixin class (#6571)

  Show description for fa633ed
  

  3 peopleauthored
  fa633ed

  Copy full SHA for fa633ed

Commits on Feb 27, 2024

• denormalize latents with the mean and std if available (#7111)

  Show description for d71ecad
  patil-suraj

  authored
  d71ecad

  Copy full SHA for d71ecad

• Make LoRACompatibleConv padding_mode work. (#6031)

  Show description for 88aa7f6
  

  3 peopleauthored
  88aa7f6

  Copy full SHA for 88aa7f6

Commits on Feb 26, 2024

• Pass use_linear_projection parameter to mid block in UNetMotionModel (#7035)

  Show description for ae84e40
  stephen-iezzi

  authored
  ae84e40

  Copy full SHA for ae84e40

Commits on Feb 25, 2024

• Fix head_to_batch_dim for IPAdapterAttnProcessor (#7077)

  Show description for 1c47d1f
  fabiorigano

  authored
  1c47d1f

  Copy full SHA for 1c47d1f

Commits on Feb 22, 2024

• re-add unet refactor PR (#7044)

  Show description for e4b8f17
  

  3 peopleauthored
  e4b8f17

  Copy full SHA for e4b8f17

Commits on Feb 19, 2024

• IP-Adapter attention masking (#6847)

  Show description for eba7e7a
  

  3 peopleauthored
  eba7e7a

  Copy full SHA for eba7e7a

Commits on Feb 18, 2024

• Fixed typos in dosctrings of __init__() and in forward() of Unet3DConditionModel (#6663)

  Show description for c18058b
  

  MK-2012

  and

  yiyixuxu

  authored
  c18058b

  Copy full SHA for c18058b

• start depcrecation cycle for lora_attention_proc 👋 (#7007)

  sayakpaul

  authored
  d4ade82

  Copy full SHA for d4ade82

Commits on Feb 10, 2024

• allow attention processors to have different signatures (#6915)

  Show description for 0071478
  

  yiyixuxu

  and

  yiyixuxu

  authored
  0071478

  Copy full SHA for 0071478

Commits on Feb 9, 2024

• [LoRA] deprecate certain lora methods from the old backend. (#6889)

  Show description for ca9ed5e
  sayakpaul

  authored
  ca9ed5e

  Copy full SHA for ca9ed5e

Commits on Feb 8, 2024

• change to 2024 in the license (#6902)

  Show description for 30e5e81
  sayakpaul

  authored
  30e5e81

  Copy full SHA for 30e5e81

• [I2VGenXL] `attention_head_dim` in the UNet (#6872)

  Show description for 491a933
  sayakpaul

  authored
  491a933

  Copy full SHA for 491a933

Commits on Feb 6, 2024

• Revert "add attention_head_dim"

  Show description for 4f1df69
  sayakpaul

  committed
  4f1df69

  Copy full SHA for 4f1df69

• add attention_head_dim

  sayakpaul

  committed
  15f6b22

  Copy full SHA for 15f6b22

• [I2vGenXL] clean up things (#6845)

  Show description for e6fd9ad
  sayakpaul

  authored
  e6fd9ad

  Copy full SHA for e6fd9ad

Commits on Feb 5, 2024

• Fix `AutoencoderTiny` with `use_slicing` (#6850)

  Show description for 493228a
  edwardwli

  authored
  493228a

  Copy full SHA for 493228a

• Fix forward pass in UNetMotionModel when gradient checkpoint is enabled (#6744)

  Show description for c6f8c31
  

  BaaBaaGoat

  and

  DN6

  authored
  c6f8c31

  Copy full SHA for c6f8c31

• update IP-adapter code in UNetMotionModel (#6828)

  Show description for 64909f1
  

  yiyixuxu

  and

  yiyixuxu

  authored
  64909f1

  Copy full SHA for 64909f1

Commits on Feb 4, 2024

• add `self.use_ada_layer_norm_*` params back to `BasicTransformerBlock` (#6841)

  Show description for a5fc62f
  

  yiyixuxu

  and

  yiyixuxu

  authored
  a5fc62f

  Copy full SHA for a5fc62f