Commits

Breadcrumbs

History for

oncorrect_conversion

User selector

All users

All time

Commit History

Commits on Jan 4, 2023

• Improve reproduceability 2/3 (#1906)

  Show description for 9b63854
  patrickvonplaten

  authored
  9b63854

  Copy full SHA for 9b63854

Commits on Jan 3, 2023

• [Deterministic torch randn] Allow tensors to be generated on CPU (#1902)

  Show description for 8ed08e4
  

  3 peopleauthored
  8ed08e4

  Copy full SHA for 8ed08e4

Commits on Jan 2, 2023

• updated doc for stable diffusion pipelines (#1770)

  Show description for da31075
  

  3 peopleauthored
  da31075

  Copy full SHA for da31075

Commits on Jan 1, 2023

• [Attention] Finish refactor attention file (#1879)

  Show description for 21bbc63
  patrickvonplaten

  authored
  21bbc63

  Copy full SHA for 21bbc63

Commits on Dec 30, 2022

• Make repo structure consistent (#1862)

  Show description for 29b2c93
  

  patrickvonplaten

  and

  pcuenca

  authored
  29b2c93

  Copy full SHA for 29b2c93

Commits on Dec 29, 2022

• Flax: Fix img2img and align with other pipeline (#1824)

  Show description for ab0e92f
  

  skirsten

  and

  pcuenca

  authored
  ab0e92f

  Copy full SHA for ab0e92f

Commits on Dec 28, 2022

• [Versatile Diffusion] Fix cross_attention_kwargs (#1849)

  Show description for debc74f
  patrickvonplaten

  authored
  debc74f

  Copy full SHA for debc74f

Commits on Dec 27, 2022

• Width was typod as weight (#1800)

  Show description for 8a4c3e5
  Helw150

  authored
  8a4c3e5

  Copy full SHA for 8a4c3e5

Commits on Dec 20, 2022

• Refactor cross attention and allow mechanism to tweak cross attention function (#1639)

  Show description for 4125756
  

  patrickvonplaten

  and

  pcuenca

  authored
  4125756

  Copy full SHA for 4125756

• make style

  patrickvonplaten

  committed
  e29dc97

  Copy full SHA for e29dc97

• Only test for xformers when enabling them #1773 (#1776)

  Show description for 8e4733b
  kig

  authored
  8e4733b

  Copy full SHA for 8e4733b

Commits on Dec 19, 2022

• [Versatile] fix attention mask (#1763)

  patrickvonplaten

  authored
  b267d28

  Copy full SHA for b267d28

• Support attn2==None for xformers (#1759)

  anton-l

  authored
  32a5d70

  Copy full SHA for 32a5d70

• Add attention mask to uclip (#1756)

  Show description for 429e544
  patrickvonplaten

  authored
  429e544

  Copy full SHA for 429e544

Commits on Dec 18, 2022

• kakaobrain unCLIP (#1428)

  Show description for 2dcf64b
  

  williamberman

  and

  patrickvonplaten

  authored
  2dcf64b

  Copy full SHA for 2dcf64b

Commits on Dec 13, 2022

• Make sure all pipelines can run with batched input (#1669)

  Show description for b345c74
  

  patrickvonplaten

  and

  pcuenca

  authored
  b345c74

  Copy full SHA for b345c74

Commits on Dec 9, 2022

• do not automatically enable xformers (#1640)

  Show description for 6b68afd
  patrickvonplaten

  authored
  6b68afd

  Copy full SHA for 6b68afd

Commits on Dec 7, 2022

• Make cross-attention check more robust (#1560)

  Show description for 5e03692
  pcuenca

  authored
  5e03692

  Copy full SHA for 5e03692

• fix upcast in slice attention (#1591)

  Show description for ced7c96
  patil-suraj

  authored
  ced7c96

  Copy full SHA for ced7c96

• [UNet2DConditionModel] add an option to upcast attention to fp32 (#1590)

  Show description for 170ebd2
  patil-suraj

  authored
  170ebd2

  Copy full SHA for 170ebd2

• Add paint by example (#1533)

  Show description for 896c98a
  

  patrickvonplaten

  and

  pcuenca

  authored
  896c98a

  Copy full SHA for 896c98a

Commits on Dec 5, 2022

• [Docs] Correct docs (#1554)

  patrickvonplaten

  authored
  62b497c

  Copy full SHA for 62b497c

• [refactor] make set_attention_slice recursive (#1532)

  Show description for bce65cd
  

  patil-suraj

  and

  patrickvonplaten

  authored
  bce65cd

  Copy full SHA for bce65cd

• Compute embedding distances with torch.cdist (#1459)

  Show description for 720dbfc
  blefaudeux

  authored
  720dbfc

  Copy full SHA for 720dbfc

Commits on Dec 3, 2022

• Add xformers attention to VAE (#1507)

  Show description for daebee0
  

  3 peopleauthored
  daebee0

  Copy full SHA for daebee0

Commits on Dec 2, 2022

• fix tests

  patrickvonplaten

  committed
  cf4664e

  Copy full SHA for cf4664e

• Do not use torch.long in mps (#1488)

  Show description for 3ceaa28
  

  pcuenca

  and

  anton-l

  authored
  3ceaa28

  Copy full SHA for 3ceaa28

• [refactor] Making the xformers mem-efficient attention activation recursive (#1493)

  Show description for a816a87
  

  blefaudeux

  and

  Benjamin Lefaudeux

  authored
  a816a87

  Copy full SHA for a816a87

Commits on Dec 1, 2022

• Fix Flax flip_sin_to_cos (#1369)

  Show description for a6a25ce
  

  akashgokul

  and

  kashif

  authored
  a6a25ce

  Copy full SHA for a6a25ce

• simplyfy AttentionBlock (#1492)

  patil-suraj

  authored
  2bbf8b6

  Copy full SHA for 2bbf8b6

Commits on Nov 29, 2022

• StableDiffusion: Decode latents separately to run larger batches (#1150)

  Show description for c28d3c8
  

  kig

  and

  Ilmari Heikkinen

  authored
  c28d3c8

  Copy full SHA for c28d3c8

• Flax support for Stable Diffusion 2 (#1423)

  Show description for 4d1e4e2
  

  pcuenca

  and

  patrickvonplaten

  authored
  4d1e4e2

  Copy full SHA for 4d1e4e2

Commits on Nov 25, 2022

• Allow to set config params directly in init (#1419)

  Show description for 8faa822
  patrickvonplaten

  authored
  8faa822

  Copy full SHA for 8faa822

• [MPS] call contiguous after permute (#1411)

  Show description for babfb8a
  kashif

  authored
  babfb8a

  Copy full SHA for babfb8a

Commits on Nov 24, 2022

• Support SD2 attention slicing (#1397)

  Show description for d50e321
  anton-l

  authored
  d50e321

  Copy full SHA for d50e321