Repository navigation

stable diffusion fine-tuning - #356

Merged

patil-suraj merged 59 commits into

mainfrom

finetune-txt2img

Oct 11, 2022

Merged

stable diffusion fine-tuning#356
patil-suraj merged 59 commits into
mainfrom
finetune-txt2img

Commits on Sep 5, 2022
- begin text2image script
  patil-suraj
  committed
  66a51ed
- loading the datasets, preprocessing & transforms
  anton-l
  committed
  d062da6
- handle input features correctly
  anton-l
  committed
  3ed3a34
Commits on Sep 22, 2022
- Merge branch 'main' of https://github.com/huggingface/diffusers into finetune-txt2img
  patil-suraj
  committed
  066af65
- add gradient checkpointing support
  patil-suraj
  committed
  ce569a1
- fix output names
  patil-suraj
  committed
  837a586
- run unet in train mode not text encoder
  patil-suraj
  committed
  3893029
Commits on Sep 23, 2022
- use no_grad instead of freezing params
  patil-suraj
  committed
  61513b0
- default max steps None
  patil-suraj
  committed
  ed8f4dd
- pad to longest
  patil-suraj
  committed
  e4fb478
- don't pad when tokenizing
  patil-suraj
  committed
  7414de1
- fix encode on multi gpu
  patil-suraj
  committed
  ce4a7a2
- fix stupid bug
  patil-suraj
  committed
  95d7836
- add random flip
  patil-suraj
  committed
  54b700d
- add ema
  patil-suraj
  committed
  725fb96
- fix ema
  patil-suraj
  committed
  584b3f7
- put ema on cpu
  patil-suraj
  committed
  0f0b098
- improve EMA model
  patil-suraj
  committed
  56a9fd0
- contiguous_format
  patil-suraj
  committed
  5c05401
- don't warp vae and text encode in accelerate
  patil-suraj
  committed
  ad42acb
- remove no_grad
  patil-suraj
  committed
  4e54ae2
- use randn_like
  patil-suraj
  committed
  9cf8d2b
Commits on Sep 24, 2022
- fix resize
  patil-suraj
  committed
  2feec19
- improve few things
  patil-suraj
  committed
  7044b2d
Commits on Oct 1, 2022
- log epoch loss
  patil-suraj
  committed
  4809770
- set log level
  patil-suraj
  committed
  fdfbad3
- don't log each step
  patil-suraj
  committed
  03d124b
- remove max_length from collate
  patil-suraj
  committed
  abebd23
- style
  patil-suraj
  committed
  4779819
- add report_to option
  patil-suraj
  committed
  c6ad723
- make scale_lr false by default
  patil-suraj
  committed
  f4cd6ff
- add grad clipping
  patil-suraj
  committed
  4cc238d
Commits on Oct 2, 2022
- Merge branch 'main' of https://github.com/huggingface/diffusers into finetune-txt2img
  patil-suraj
  committed
  438514c
- add an option to use 8bit adam
  patil-suraj
  committed
  c643e94
Commits on Oct 3, 2022
- fix logging in multi-gpu, log every step
  patil-suraj
  committed
  3caf7c6
Commits on Oct 4, 2022
- more comments
  patil-suraj
  committed
  518448c
- remove eval for now
  patil-suraj
  committed
  926c20e
- adress review comments
  patil-suraj
  committed
  12d19df
- add requirements file
  patil-suraj
  committed
  ac0b09e
- begin readme
  patil-suraj
  committed
  48930b7
- begin readme
  patil-suraj
  committed
  7964342
Commits on Oct 6, 2022
- fix typo
  patil-suraj
  committed
  1c8387c
- fix push to hub
  patil-suraj
  committed
  3eea0db
- populate readme
  patil-suraj
  committed
  eb8e6c3
- update readme
  patil-suraj
  committed
  2cb7c43
- remove use_auth_token from the script
  patil-suraj
  committed
  7228818
- Merge branch 'main' of https://github.com/huggingface/diffusers into finetune-txt2img
  patil-suraj
  committed
  4ffbf57
Commits on Oct 11, 2022
- address some review comments
  patil-suraj
  committed
  b08d85d
- Merge branch 'main' of https://github.com/huggingface/diffusers into finetune-txt2img
  patil-suraj
  committed
  64338d2
- better mixed precision support
  patil-suraj
  committed
  db8e31a
- remove redundant to
  patil-suraj
  committed
  17cb6e7
- create ema model early
  patil-suraj
  committed
  25625ce
- Apply suggestions from code review
  
  patil-suraj
  and
  pcuenca
  authored
  5d71880
- better description for train_data_dir
  patil-suraj
  committed
  3a6e4f2
- Merge branch 'finetune-txt2img' of https://github.com/huggingface/diffusers into finetune-txt2img
  patil-suraj
  committed
  b05a860
- add diffusers in requirements
  patil-suraj
  committed
  1c8b026
- update dataset_name_mapping
  patil-suraj
  committed
  5b22178
- update readme
  patil-suraj
  committed
  f0b4357
- add inference example
  patil-suraj
  committed
  f9a4025

Commits

Commits on Sep 5, 2022

Commits on Sep 22, 2022

Commits on Sep 23, 2022

Commits on Sep 24, 2022

Commits on Oct 1, 2022

Commits on Oct 2, 2022

Commits on Oct 3, 2022

Commits on Oct 4, 2022

Commits on Oct 6, 2022

Commits on Oct 11, 2022