ArrowM
|
6cf6bf81b4
|
Merge remote-tracking branch 'ExponentialML/feat/dynamic-image-norm' into dev
|
2023-05-05 14:27:03 -05:00 |
ArrowM
|
fc6793a096
|
Merge branch '1208-rm-tensorflow' into dev
# Conflicts:
# requirements.txt
|
2023-05-05 11:55:18 -05:00 |
shenzehong
|
2da74941ce
|
rm tensorflow, use tensorboard for log parsing
|
2023-05-02 12:08:50 +08:00 |
ExponentialML
|
913ad8e5de
|
Add dynamic image normalization, change prior loss method
|
2023-04-30 14:21:59 -07:00 |
d8ahazard
|
d2e7908d03
|
More new UI fixes
|
2023-04-28 08:55:12 -05:00 |
d8ahazard
|
6058f2182c
|
New UI fun
|
2023-04-25 17:50:38 -05:00 |
d8ahazard
|
bfc918d031
|
New UI Updates
|
2023-04-18 10:37:08 -05:00 |
saunderez
|
42b7304c12
|
Revert log_parser
|
2023-04-02 14:50:44 +10:00 |
saunderez
|
9f292ecc6c
|
All the stuff I commited earlier that vanished
|
2023-04-01 19:10:28 +10:00 |
ArrowM
|
1b6d68c821
|
Merge pull request #1142 from gammagoat/shareddiff
Allow LoRA model training to use shared diffusers models
|
2023-03-31 09:29:32 -05:00 |
ArrowM
|
7380dcc8b3
|
simplify
|
2023-03-30 09:22:33 -05:00 |
ArrowM
|
98883f131d
|
capitalize Lora
|
2023-03-30 09:21:42 -05:00 |
gammagoat
|
f37339e20f
|
Merge remote-tracking branch 'upstream/dev' into shareddiff
|
2023-03-29 15:32:17 -04:00 |
gammagoat
|
d18ab42384
|
Shared src models for LoRA
Implement a new experimental option for using shared
base models for LoRA training so we don't need to copy 4GB every time we
train a LoRA model.
|
2023-03-29 15:09:24 -04:00 |
ArrowM
|
753c3f6a3e
|
Xformers installation updates & js fixes
|
2023-03-29 10:08:40 -05:00 |
ArrowM
|
0fb9428ac9
|
Fix class prompt processing
|
2023-03-23 19:22:06 -05:00 |
ArrowM
|
7d2d047962
|
separate training and saving progress bars
|
2023-03-22 15:45:03 -05:00 |
ArrowM
|
21affbbf43
|
Make the install smarter
|
2023-03-19 16:24:20 -05:00 |
ArrowM
|
87e9d85681
|
Rework postinstall.py
|
2023-03-17 23:06:55 -05:00 |
ArrowM
|
f8657b1a27
|
Merge branch 'optimizers' into dev
# Conflicts:
# dreambooth/shared.py
# dreambooth/train_dreambooth.py
|
2023-03-15 14:25:09 -05:00 |
ArrowM
|
dd2c058802
|
change post-scale cropping to center crop instead of bottom left crop. Also change default divider to 32 and default sampler to euler a
|
2023-03-12 12:58:09 -05:00 |
ArrowM
|
46359e0719
|
Gradio and new schedulers
|
2023-03-09 17:44:44 -06:00 |
d8ahazard
|
c53b3ede2b
|
Fix error importing log parser, cleanup API logging.
|
2023-03-09 09:09:39 -06:00 |
ArrowM
|
1c98f8c85f
|
more fixes
|
2023-03-08 21:25:33 -06:00 |
d8ahazard
|
3e8ddce4dd
|
One set of imports to rule them all.
|
2023-03-08 12:51:24 -06:00 |
d8ahazard
|
2bcc837fc1
|
Log parsing improvements
Cache parsed log data, so we only need to parse changed/new logfiles.
Close matplotlib figures, as directed by the softwarez.
|
2023-03-07 13:35:34 -06:00 |
ArrowM
|
d6490b47aa
|
Fix #1011 and re-enable changelog popup
|
2023-03-06 14:54:16 -06:00 |
ArrowM
|
3bd73f3cce
|
Hopeful fix bad lora_model_name and stop_text_encoder values.
Also revert diffusers to 0.13.1 (it seems like some people had some issues with it)
|
2023-03-04 10:11:16 -06:00 |
ArrowM
|
f4e4e2ae0b
|
Fix lora list issue
|
2023-03-03 16:48:04 -06:00 |
ArrowM
|
d01ef87cbb
|
cleanup
|
2023-03-03 10:53:09 -06:00 |
ArrowM
|
48161a76c5
|
I love documentation
|
2023-03-02 18:30:04 -06:00 |
Juanu
|
24193a0983
|
Merge branch 'dev' of https://github.com/d8ahazard/sd_dreambooth_extension into mac-fixes
|
2023-03-02 13:14:40 +01:00 |
Juanu
|
be25b6d569
|
ContContinue ImageBuilder if xformers not present
|
2023-03-02 12:45:09 +01:00 |
d8ahazard
|
8b4b5def4f
|
Fix scheduler for UniPC if using API
|
2023-03-01 18:17:05 -06:00 |
d8ahazard
|
adcf15c264
|
Add scheduler option to the UI/api for sample generation
|
2023-03-01 14:12:56 -06:00 |
d8ahazard
|
fe0bb0d13c
|
Code Cleanup
Basic reformatting
|
2023-02-28 15:09:22 -06:00 |
d8ahazard
|
2c80af3628
|
Fun with progress bars
|
2023-02-24 15:36:05 -06:00 |
d8ahazard
|
f2233d1414
|
Schedulers
|
2023-02-24 14:50:54 -06:00 |
d8ahazard
|
85d3e9f683
|
Merge branch 'dev' of https://github.com/d8ahazard/sd_dreambooth_extension into dev
|
2023-02-24 13:34:04 -06:00 |
d8ahazard
|
80175a5b94
|
UI WORK, bump requirements, Scheduler rework
|
2023-02-24 13:30:32 -06:00 |
ArrowM
|
a1ac25a040
|
Merge branch 'dev' into torch2
|
2023-02-23 17:04:15 -06:00 |
d8ahazard
|
01673aeae7
|
Fix accelerator logging dir warning
|
2023-02-23 16:51:13 -06:00 |
ArrowM
|
4fe951e475
|
pipeline opts
|
2023-02-23 16:10:09 -06:00 |
d8ahazard
|
1908083440
|
Fix possible commit history isssues.
|
2023-02-17 21:58:41 -06:00 |
d8ahazard
|
f8a3513b7c
|
Post-separation anxiety
I'm just kidding, this makes me nothing but giddy.
Bye Felicia!
|
2023-02-13 08:37:04 -06:00 |
d8ahazard
|
b24fe0b22c
|
I want to break free...
Add exception handling for all imports if, say, the code were launched by itself, versus as an extension...
|
2023-02-12 15:57:41 -06:00 |
d8ahazard
|
a8ae9c07e6
|
Cleanup image_builder
|
2023-02-12 13:55:34 -06:00 |
d8ahazard
|
2fe53522a8
|
DEIS For all
Add DEIS scheduler to sample generator.
Use it for actual training?
Update sample infer steps to utilize faster scheduler.
|
2023-02-12 12:14:14 -06:00 |
d8ahazard
|
12896cfcac
|
Cleanup Print Messages, Proper Xformers Check
Check xformers version again on start, remove print messages.
|
2023-02-12 11:35:15 -06:00 |
d8ahazard
|
aea519fec2
|
Add DEIS Scheduler to Image Pipe, TXT2IMG Fixes
Add DEIS Scheduler for sample/class generation with diffusers.
Fix Txt2IMG failure to load models.
Allow loading EMA Unet for sample weights.
|
2023-02-12 11:34:47 -06:00 |