Commit Graph

164 Commits

Author SHA1 Message Date
chavinlo a2772fc668 fixes 2022-11-16 10:55:38 -05:00
chavinlo fed3431f03 Revert "sync trainer with main branch"
This reverts commit 80e2422967.
2022-11-16 10:44:39 -05:00
Carlos Chavez 80e2422967
sync trainer with main branch 2022-11-16 10:39:20 -05:00
Carlos Chavez d600078008
Add options and local inference
Added options to:
- Disable Inference (it consumes about 2gb of VRAM even when not active)
- Disable wandb

and:
- if no hftoken is provided it just fills it with nothing so it doesn't argues
- if wandb is not enabled, save the inference outputs to a local folder along with information about it
2022-11-14 22:08:16 -05:00
Anthony Mercurio ae561d19f7
Merge pull request #40 from lopho/patch-1
Parse booleans in argument parser
2022-11-14 12:29:31 -05:00
lopho cd0910e82d
Parse booleans in argument parser
true, yes or 1 correspond to True, else False.
2022-11-12 11:14:48 +01:00
Anthony Mercurio 624f0f14af
correctly set args 2022-11-11 08:25:27 -07:00
Anthony Mercurio 033c1e75d0
Merge pull request #36 from harubaru/relicense
Code overhaul & relicense to AGPL 3.0
2022-11-11 07:45:08 -07:00
harubaru 15e70690de Add aesthetic 2022-11-10 13:01:11 -07:00
harubaru 1f5b671b67 relicense 2022-11-10 12:59:53 -07:00
Anthony Mercurio bc626e80e1
Merge pull request #34 from chavinlo/patch-1
typo
2022-11-10 09:23:10 -07:00
Anthony Mercurio 548ebea881
Merge pull request #32 from Maw-Fox/main
Add resize argument, various fixes
2022-11-10 09:22:48 -07:00
Anthony Mercurio 4e89309d6b
Merge branch 'main' into main 2022-11-10 09:22:37 -07:00
Anthony Mercurio 2f8d71c589
Merge pull request #33 from laksjdjf/sub
Apply xformers in diffusers_trainer.py
2022-11-10 09:21:30 -07:00
Carlos Chavez c4b1860282
typo
lol
2022-11-10 10:13:14 -05:00
Maw-Fox 46e8d98d2b Revert 2022-11-10 06:11:09 -07:00
Maw-Fox 0be39a4887 Fix depreciated enum 2022-11-10 05:42:38 -07:00
laksjdjf 353200b039
Update diffusers_trainer.py 2022-11-09 23:56:49 +00:00
Maw-Fox 3fe9df1450 Cleanup. 2022-11-08 20:50:57 -07:00
laksjdjf 0c29d1e84d
Update diffusers_trainer.py 2022-11-09 10:34:49 +09:00
Maw-Fox fe30e8942e Add resize and optional resize arg 2022-11-08 18:17:31 -07:00
Anthony Mercurio 1ea31eb71e
Merge pull request #31 from john-sungjin/fix-ema-checkpoint-saving
Restore non-EMA weights after saving checkpoint
2022-11-06 16:44:11 -08:00
John Kim 94603d9578 Added store/restore to EMAModel, restore non-EMA weights after saving checkpoint 2022-11-07 00:28:55 +00:00
Anthony Mercurio 517f3e154f
Merge pull request #30 from harubaru/scheduler-options
Updates to Scheduler Options and Resume Code
2022-11-06 07:51:48 -08:00
cafeai 2048ce67ef Remove Unnecessary Line (Scheduler) 2022-11-06 19:34:33 +09:00
cafeai b303fdc293 Scheduler Options & Resume Updates 2022-11-06 19:08:38 +09:00
Anthony Mercurio 56923359a3
Add checkpoint resuming, exception handling, and learning rate schedulers 2022-11-06 01:05:40 -07:00
Anthony Mercurio 2321e22fc1
Merge pull request #28 from AstraliteHeart/main
Misc cleanups from pony-diffusion test run
2022-11-05 08:55:48 -07:00
Anthony Mercurio 7be7fc3b98
Merge branch 'main' into main 2022-11-05 08:55:29 -07:00
Anthony Mercurio 5b793feb52
Merge pull request #29 from harubaru/bucket-tests
Bucket tests
2022-11-05 08:54:26 -07:00
cafeai ed35788059 Cleanup 2022-11-06 00:53:30 +09:00
cafeai b0b1337bc5 ImageStore Filename Fix 2022-11-06 00:32:15 +09:00
cafeai 305efa20f2 Bucket Helpers 2022-11-05 22:55:56 +09:00
Astralite Heart 0e0acf31b1 diffusers no longer use tensor_format in DDPMScheduler 2022-11-05 11:05:45 +00:00
Astralite Heart 74ef98667c Add bitsandbytes instructions 2022-11-05 10:38:35 +00:00
Astralite Heart bb153c1ec0 Add missing reqs / fix versions 2022-11-04 17:31:08 -07:00
Anthony Mercurio 6b73cfe448
Add CLIP penultimate layer for diffusers trainer 2022-11-02 16:20:46 -07:00
Anthony Mercurio 1a94c70736
Implement distributed training 2022-11-01 13:28:12 -07:00
Anthony Mercurio ef25ccac36
add vae config 2022-11-01 12:40:39 -07:00
Anthony Mercurio 6bf942eb63
Merge pull request #26 from GripenANM/Update-Weights-Readme.md-to-v1.3-model
Update Weights Readme.md to v1.3 model
2022-10-29 18:58:56 -07:00
Anthony Mercurio 8247915345
add attribution 2022-10-29 16:11:48 -07:00
Anthony Mercurio be702cce2e
Aspect ratio bucketing, Training overhaul, more new stuff 2022-10-29 09:26:03 -07:00
Akumi Chan e1e416a5b9
Removed Huggingface metadata table 2022-10-25 13:31:22 -07:00
Akumi Chan 1947519397
Update Weights Readme.md to v1.3 model
cloned huggingface model card
2022-10-25 13:23:18 -07:00
harubaru 76efa00de3 fix deprecation warnings in dataloader 2022-10-15 18:43:28 -07:00
harubaru 67140231e7 fix inference config 2022-10-15 11:41:39 -07:00
harubaru 61ed683d1c ensure wandb run doesnt get overwritten 2022-10-15 11:40:53 -07:00
harubaru 18ad704fd6 fix wandb logging 2022-10-15 11:20:32 -07:00
harubaru f980137430 add fp16 training 2022-10-15 10:57:00 -07:00
harubaru 9581fbc226
Merge pull request #24 from hasuwoof/feature/extended_clip
Implement CLIP extensions (training + inference)
2022-10-14 16:27:21 -07:00