Commits

Breadcrumbs

History for

onmain

User selector

All users

All time

Commit History

Commits on Sep 30, 2022

• Fix slow tests (#689)

  Show description for b2cfc7a
  NouamaneTazi

  authored
  b2cfc7a

  Copy full SHA for b2cfc7a

• Allow resolutions that are not multiples of 64 (#505)

  Show description for a784be2
  

  jachiam

  and

  patrickvonplaten

  authored
  a784be2

  Copy full SHA for a784be2

• Optimize Stable Diffusion (#371)

  Show description for 9ebaea5
  NouamaneTazi

  authored
  9ebaea5

  Copy full SHA for 9ebaea5

Commits on Sep 29, 2022

• Renamed x -> hidden_states in resnet.py (#676)

  Show description for a7058f4
  daspartho

  authored
  a7058f4

  Copy full SHA for a7058f4

Commits on Sep 27, 2022

• Fix `SpatialTransformer` (#578)

  Show description for d886e49
  ydshieh

  authored
  d886e49

  Copy full SHA for d886e49

Commits on Sep 23, 2022

• Flax documentation (#589)

  Show description for 8b0be93
  

  3 peopleauthored
  8b0be93

  Copy full SHA for 8b0be93

Commits on Sep 22, 2022

• [UNet2DConditionModel] add gradient checkpointing (#461)

  Show description for e7120ba
  patil-suraj

  authored
  e7120ba

  Copy full SHA for e7120ba

Commits on Sep 21, 2022

• Replace `dropout_prob` by `dropout` in `vae` (#595)

  Show description for 3fc8ef7
  younesbelkada

  authored
  3fc8ef7

  Copy full SHA for 3fc8ef7

Commits on Sep 20, 2022

• [Flax] Fix unet and ddim scheduler (#594)

  Show description for 2345481
  patrickvonplaten

  authored
  2345481

  Copy full SHA for 2345481

• FlaxDiffusionPipeline & FlaxStableDiffusionPipeline (#559)

  Show description for d934d3d
  

  3 peopleauthored
  d934d3d

  Copy full SHA for d934d3d

• [FlaxAutoencoderKL] rename weights to align with PT (#584)

  Show description for c01ec2d
  patil-suraj

  authored
  c01ec2d

  Copy full SHA for c01ec2d

• Add `from_pt` argument in `.from_pretrained` (#527)

  Show description for 0902449
  younesbelkada

  authored
  0902449

  Copy full SHA for 0902449

Commits on Sep 19, 2022

• Fix `CrossAttention._sliced_attention` (#563)

  Show description for 84616b5
  ydshieh

  authored
  84616b5

  Copy full SHA for 84616b5

• [Flax] Solve problem with VAE (#574)

  patrickvonplaten

  authored
  2a8477d

  Copy full SHA for 2a8477d

• [Flax] Add Vae for Stable Diffusion (#555)

  Show description for bf5ca03
  

  patrickvonplaten

  and

  patil-suraj

  authored
  bf5ca03

  Copy full SHA for bf5ca03

• Fix `_upsample_2d` (#535)

  Show description for b17d49f
  ydshieh

  authored
  b17d49f

  Copy full SHA for b17d49f

• revert the accidental commit

  ydshieh

  committed
  0424615

  Copy full SHA for 0424615

• Fix CrossAttention._sliced_attention

  ydshieh

  committed
  8187865

  Copy full SHA for 8187865

Commits on Sep 18, 2022

• FlaxUNet2DConditionOutput @flax.struct.dataclass (#550)

  

  Mishig Davaadorj

  authored
  0c0c222

  Copy full SHA for 0c0c222

Commits on Sep 16, 2022

• Fix typos and add Typo check GitHub Action (#483)

  Show description for 76d492e
  

  shirayu

  and

  anton-l

  authored
  76d492e

  Copy full SHA for 76d492e

• Remove the usage of numpy in up/down sample_2d (#503)

  Show description for c049372
  ydshieh

  authored
  c049372

  Copy full SHA for c049372

Commits on Sep 15, 2022

• UNet Flax with FlaxModelMixin (#502)

  Show description for d8b0e4f
  

  5 peopleauthored
  d8b0e4f

  Copy full SHA for d8b0e4f

• [UNet2DConditionModel, UNet2DModel] pass norm_num_groups to all the blocks (#442)

  Show description for d144c46
  patil-suraj

  authored
  d144c46

  Copy full SHA for d144c46

Commits on Sep 14, 2022

• [CrossAttention] add different method for sliced attention (#446)

  Show description for 8b45096
  

  patil-suraj

  and

  patrickvonplaten

  authored
  8b45096

  Copy full SHA for 8b45096

• Removing `.float()` (`autocast` in fp16 will discard this (I think)). (#495)

  Narsil

  authored
  7c4b38b

  Copy full SHA for 7c4b38b

Commits on Sep 12, 2022

• update expected results of slow tests (#268)

  Show description for f4781a0
  

  kashif

  and

  patrickvonplaten

  authored
  f4781a0

  Copy full SHA for f4781a0

Commits on Sep 9, 2022

• Renamed variables from single letter to better naming (#449)

  Show description for 5843487
  daspartho

  authored
  5843487

  Copy full SHA for 5843487

• use torch.matmul instead of einsum in attnetion. (#445)

  Show description for 5adb0a7
  patil-suraj

  authored
  5adb0a7

  Copy full SHA for 5adb0a7

Commits on Sep 8, 2022

• [Black] Update black (#433)

  Show description for b2b3b1a
  patrickvonplaten

  authored
  b2b3b1a

  Copy full SHA for b2b3b1a

• [Docs] Models (#416)

  Show description for 5e6417e
  

  3 peopleauthored
  5e6417e

  Copy full SHA for 5e6417e

• Inference support for `mps` device (#355)

  Show description for 5dda173
  

  3 peopleauthored
  5dda173

  Copy full SHA for 5dda173

Commits on Sep 7, 2022

• Rename variables from single letter to meaningful name fix (#395)

  Show description for 1a431ae
  

  rashmimarganiatgithub

  and

  Rashmi S

  authored
  1a431ae

  Copy full SHA for 1a431ae

Commits on Sep 6, 2022

• Efficient Attention (#366)

  Show description for 5c4ea00
  

  patrickvonplaten

  and

  pcuenca

  authored
  5c4ea00

  Copy full SHA for 5c4ea00

• Use `expand` instead of ones to broadcast tensor (#373)

  Show description for 56c0037
  pcuenca

  authored
  56c0037

  Copy full SHA for 56c0037

Commits on Sep 5, 2022

• [Type Hint] VAE models (#365)

  Show description for f085d2f
  

  daspartho

  and

  anton-l

  authored
  f085d2f

  Copy full SHA for f085d2f