Commits

Breadcrumbs

History for

onmain

User selector

All users

All time

Commit History

Commits on Nov 4, 2024

• Refac training utils.py (#9815)

  Show description for a3cc641
  

  RogerSinghChugh

  and

  sayakpaul

  authored
  a3cc641

  Copy full SHA for a3cc641

Commits on Nov 1, 2024

• Fixes EMAModel "from_pretrained" method (#9779)

  Show description for f55f1f7
  

  SahilCarterr

  and

  sayakpaul

  authored
  f55f1f7

  Copy full SHA for f55f1f7

Commits on Oct 25, 2024

• [bugfix] bugfix for npu free memory (#9640)

  Show description for 94643fa
  

  leisuzz

  and

  sayakpaul

  authored
  94643fa

  Copy full SHA for 94643fa

Commits on Oct 16, 2024

• Resolves [BUG] 'GatheredParameters' object is not callable (#9614)

  Show description for 29a2c5d
  

  charchit7

  and

  a-r-r-o-w

  authored
  29a2c5d

  Copy full SHA for 29a2c5d

Commits on Oct 15, 2024

• [doc] Fix some docstrings in `src/diffusers/training_utils.py` (#9606)

  Show description for 355bb64
  

  mreraser

  and

  sayakpaul

  authored
  355bb64

  Copy full SHA for 355bb64

Commits on Sep 28, 2024

• [chore] fix: retain memory utility. (#9543)

  Show description for 8e7d6c0
  sayakpaul

  authored
  8e7d6c0

  Copy full SHA for 8e7d6c0

Commits on Sep 3, 2024

• chore: add a cleaning utility to be useful during training. (#9240)

  sayakpaul

  authored
  8ba90aa

  Copy full SHA for 8ba90aa

Commits on Aug 27, 2024

• Replace transformers.deepspeed with transformers.integrations.deepspeed (#9281)

  Show description for 9aca79f
  

  salcc

  and

  sayakpaul

  authored
  9aca79f

  Copy full SHA for 9aca79f

Commits on Jun 24, 2024

• Add extra performance features for EMAModel, torch._foreach operations and better support for non-blocking CPU offloading (#7685)

  Show description for 2ada094
  

  3 peopleauthored
  2ada094

  Copy full SHA for 2ada094

Commits on Jun 18, 2024

• [SD3 training] refactor the density and weighting utilities. (#8591)

  Show description for 4edde13
  sayakpaul

  authored
  4edde13

  Copy full SHA for 4edde13

Commits on Jun 1, 2024

• Fix DREAM training (#8302)

  Show description for bc108e1
  

  4 peopleauthored
  bc108e1

  Copy full SHA for bc108e1

Commits on Apr 27, 2024

• Add DREAM training (#6381)

  Show description for 9d16daa
  

  AmericanPresidentJimmyCarter

  and

  Jimmy

  authored
  9d16daa

  Copy full SHA for 9d16daa

Commits on Mar 9, 2024

• Add npu support (#7144)

  Show description for e6ff752
  MengqingCao

  authored
  e6ff752

  Copy full SHA for e6ff752

Commits on Feb 1, 2024

• add is_torchvision_available (#6800)

  Show description for 093a03a
  

  yiyixuxu

  and

  yiyixuxu

  authored
  093a03a

  Copy full SHA for 093a03a

Commits on Jan 31, 2024

• fix torchvision import (#6796)

  patrickvonplaten

  authored
  c3369f5

  Copy full SHA for c3369f5

Commits on Jan 15, 2024

• [Training Utils] create a utility for casting the lora params during training. (#6553)

  Show description for a080f0d
  sayakpaul

  authored
  a080f0d

  Copy full SHA for a080f0d

Commits on Jan 12, 2024

• [Training] fix training resuming problem when using FP16 (SDXL LoRA DreamBooth) (#6514)

  Show description for 79df503
  sayakpaul

  authored
  79df503

  Copy full SHA for 79df503

Commits on Jan 5, 2024

• Improve LCM(-LoRA) Distillation Scripts (#6420)

  Show description for f3d1333
  dg845

  authored
  f3d1333

  Copy full SHA for f3d1333

Commits on Dec 2, 2023

• [LoRA serialization] fix: duplicate unet prefix problem. (#5991)

  Show description for d486f0e
  

  sayakpaul

  and

  patrickvonplaten

  authored
  d486f0e

  Copy full SHA for d486f0e

Commits on Oct 11, 2023

• [Examples] use loralinear instead of depecrecated lora attn procs. (#5331)

  Show description for 0fa32bd
  sayakpaul

  authored
  0fa32bd

  Copy full SHA for 0fa32bd

Commits on Sep 27, 2023

• [Examples] add `compute_snr()` to training utils. (#5188)

  Show description for cdcc01b
  sayakpaul

  authored
  cdcc01b

  Copy full SHA for cdcc01b

Commits on Jun 22, 2023

• fix: random module seeding (#3846)

  sayakpaul

  authored
  13e781f

  Copy full SHA for 13e781f

Commits on May 22, 2023

• Refactor full determinism (#3485)

  Show description for 51843fd
  patrickvonplaten

  authored
  51843fd

  Copy full SHA for 51843fd

Commits on May 11, 2023

• [deepspeed] partial ZeRO-3 support (#3076)

  Show description for af2a237
  

  stas00

  and

  patrickvonplaten

  authored
  af2a237

  Copy full SHA for af2a237

Commits on Mar 6, 2023

• ema step, don't empty cuda cache (#2563)

  williamberman

  authored
  f727863

  Copy full SHA for f727863

Commits on Feb 16, 2023

• [Utils] Adds `store()` and `restore()` methods to EMAModel (#2302)

  Show description for 6eaebe8
  

  4 peopleauthored
  6eaebe8

  Copy full SHA for 6eaebe8

Commits on Feb 8, 2023

• EMA: fix `state_dict()` and `load_state_dict()` & add `cur_decay_value` (#2146)

  Show description for 9d0d070
  chenguolin

  authored
  9d0d070

  Copy full SHA for 9d0d070

Commits on Feb 7, 2023

• Use `accelerate` save & loading hooks to have better checkpoint structure (#2048)

  Show description for f5ccffe
  

  patrickvonplaten

  and

  pcuenca

  authored
  f5ccffe

  Copy full SHA for f5ccffe

Commits on Jan 29, 2023

• fix typo in EMAModel's load_state_dict() (#2151)

  Show description for 125d783
  dasayan05

  authored
  125d783

  Copy full SHA for 125d783

Commits on Jan 19, 2023

• Fix EMA for multi-gpu training in the unconditional example (#1930)

  Show description for 7c82a16
  

  4 peopleauthored
  7c82a16

  Copy full SHA for 7c82a16

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 Jul 28, 2022

• [Vae and AutoencoderKL] Final clean of LDM checkpoints (#137)

  Show description for 3100bc9
  patrickvonplaten

  authored
  3100bc9

  Copy full SHA for 3100bc9

Commits on Jul 27, 2022

• Add training and batched inference test for DDPM vs DDIM (#140)

  Show description for 6c15636
  anton-l

  authored
  6c15636

  Copy full SHA for 6c15636

Commits on Jul 4, 2022

• Fix mutable proj_out weight in the Attention layer (#73)

  Show description for d9316bf
  anton-l

  authored
  d9316bf

  Copy full SHA for d9316bf

• EMA model stepping updated to keep track of current step (#64)

  Show description for 3abf4bc
  tmabraham

  authored
  3abf4bc

  Copy full SHA for 3abf4bc