Commits

Breadcrumbs

History for

oncache-location

User selector

All users

All time

Commit History

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 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 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

• 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

• [refactor] make set_attention_slice recursive (#1532)

  Show description for bce65cd
  

  patil-suraj

  and

  patrickvonplaten

  authored
  bce65cd

  Copy full SHA for bce65cd

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

• [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

• simplyfy AttentionBlock (#1492)

  patil-suraj

  authored
  2bbf8b6

  Copy full SHA for 2bbf8b6

Commits on Nov 25, 2022

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

  Show description for babfb8a
  kashif

  authored
  babfb8a

  Copy full SHA for babfb8a

Commits on Nov 24, 2022

• Adapt UNet2D for supre-resolution (#1385)

  Show description for cecdd8b
  patil-suraj

  authored
  cecdd8b

  Copy full SHA for cecdd8b

Commits on Nov 23, 2022

• [Transformer2DModel] don't norm twice (#1381)

  Show description for 1524122
  patil-suraj

  authored
  1524122

  Copy full SHA for 1524122

• update unet2d (#1376)

  Show description for f07a16e
  patil-suraj

  authored
  f07a16e

  Copy full SHA for f07a16e

• [Versatile Diffusion] Add versatile diffusion model (#1283)

  Show description for 2625fb5
  

  patrickvonplaten

  and

  anton-l

  authored
  2625fb5

  Copy full SHA for 2625fb5

Commits on Nov 22, 2022

• use memory_efficient_attention by default (#1354)

  Show description for 2d6d4ed
  

  patil-suraj

  and

  pcuenca

  authored
  2d6d4ed

  Copy full SHA for 2d6d4ed

Commits on Nov 21, 2022

• perf: prefer batched matmuls for attention (#1203)

  Show description for ad93593
  Birch-san

  authored
  ad93593

  Copy full SHA for ad93593

Commits on Nov 14, 2022

• Edited attention.py for older xformers (#1270)

  Show description for 33d7e89
  Lime-Cakes

  authored
  33d7e89

  Copy full SHA for 33d7e89

Commits on Nov 8, 2022

• handle dtype xformers attention (#1196)

  Show description for 5786b0e
  patil-suraj

  authored
  5786b0e

  Copy full SHA for 5786b0e

Commits on Nov 3, 2022

• VQ-diffusion (#658)

  Show description for ef2ea33
  

  4 peopleauthored
  ef2ea33

  Copy full SHA for ef2ea33

Commits on Nov 2, 2022

• Fix a small typo of a variable name (#1063)

  Show description for 1216a3b
  omihub777

  authored
  1216a3b

  Copy full SHA for 1216a3b

• Up to 2x speedup on GPUs using memory efficient attention (#532)

  Show description for 98c4213
  

  MatthieuToulemont

  and

  NouamaneTazi

  authored
  98c4213

  Copy full SHA for 98c4213

Commits on Oct 31, 2022

• Remove nn sequential (#1086)

  Show description for 888468d
  patrickvonplaten

  authored
  888468d

  Copy full SHA for 888468d

Commits on Oct 29, 2022

• Experimental: allow fp16 in `mps` (#961)

  Show description for 95414bd
  pcuenca

  authored
  95414bd

  Copy full SHA for 95414bd

Commits on Oct 25, 2022

• mps changes for PyTorch 1.13 (#926)

  Show description for 3d02c92
  

  pcuenca

  and

  anton-l

  authored
  3d02c92

  Copy full SHA for 3d02c92

Commits on Oct 12, 2022

• add or fix license formatting in models directory (#808)

  Show description for 5afc2b6
  

  Nathan Lambert

  authored
  5afc2b6

  Copy full SHA for 5afc2b6