feffy380
6b3148fd3f
Fix min-snr-gamma for v-prediction and ZSNR.
...
This fixes min-snr for vpred+zsnr by dividing directly by SNR+1.
The old implementation did it in two steps: (min-snr/snr) * (snr/(snr+1)), which causes division by zero when combined with --zero_terminal_snr
2023-11-07 23:02:25 +01:00
Kohya S
6231aa91e2
common lr logging, set default None to ddp_timeout
2023-11-05 19:09:17 +09:00
Kohya S
96d877be90
support separate LR for Text Encoder for SD1/2
2023-10-29 21:30:32 +09:00
Kohya S
9d6a5a0c79
Merge pull request #899 from shirayu/use_moving_average
...
Show moving average loss in the progress bar
2023-10-29 14:37:58 +09:00
Yuta Hayashibe
63992b81c8
Fix initialize place of loss_recorder
2023-10-27 21:13:29 +09:00
Yuta Hayashibe
0d21925bdf
Use @property
2023-10-27 18:14:27 +09:00
Yuta Hayashibe
efef5c8ead
Show "avr_loss" instead of "loss" because it is moving average
2023-10-27 17:59:58 +09:00
Yuta Hayashibe
3d2bb1a8f1
Add LossRecorder and use moving average in all places
2023-10-27 17:49:49 +09:00
青龍聖者@bdsqlsz
202f2c3292
Debias Estimation loss ( #889 )
...
* update for bnb 0.41.1
* fixed generate_controlnet_subsets_config for training
* Revert "update for bnb 0.41.1"
This reverts commit 70bd3612d86539363c5c
2023-10-23 22:59:14 +09:00
Yuta Hayashibe
27f9b6ffeb
updated typos to v1.16.15 and fix typos
2023-10-01 21:51:24 +09:00
Disty0
b64389c8a9
Intel ARC support with IPEX
2023-09-19 18:05:05 +03:00
Kohya S
73a08c0be0
Merge pull request #630 from ddPn08/sdxl
...
make tracker init_kwargs configurable
2023-07-20 22:05:55 +09:00
Kohya S
6d2d8dfd2f
add zero_terminal_snr option
2023-07-18 23:17:23 +09:00
ddPn08
b841dd78fe
make tracker init_kwargs configurable
2023-07-11 10:21:45 +09:00
Kohya S
ea182461d3
add min/max_timestep
2023-07-03 20:44:42 +09:00
Kohya S
bfd909ab79
Merge branch 'main' into original-u-net
2023-06-24 08:49:07 +09:00
Kohya S
0cfcb5a49c
fix lr/d*lr is not logged with prodigy in finetune
2023-06-24 08:36:09 +09:00
Kohya S
5114e8daf1
fix training scripts except controlnet not working
2023-06-22 08:46:53 +09:00
Kohya S
92e50133f8
Merge branch 'original-u-net' into dev
2023-06-17 21:57:08 +09:00
Kohya S
19dfa24abb
Merge branch 'main' into original-u-net
2023-06-16 20:59:34 +09:00
青龍聖者@bdsqlsz
e97d67a681
Support for Prodigy(Dadapt variety for Dylora) ( #585 )
...
* Update train_util.py for DAdaptLion
* Update train_README-zh.md for dadaptlion
* Update train_README-ja.md for DAdaptLion
* add DAdatpt V3
* Alignment
* Update train_util.py for experimental
* Update train_util.py V3
* Update train_README-zh.md
* Update train_README-ja.md
* Update train_util.py fix
* Update train_util.py
* support Prodigy
* add lower
2023-06-15 21:12:53 +09:00
Kohya S
9806b00f74
add arbitrary dataset feature to each script
2023-06-15 20:39:39 +09:00
ykume
9e1683cf2b
support sdpa
2023-06-11 21:26:15 +09:00
ykume
0315611b11
remove workaround for accelerator=0.15, fix XTI
2023-06-11 18:32:14 +09:00
Kohya S
ec2efe52e4
scale v-pred loss like noise pred
2023-06-03 10:52:22 +09:00
ddPn08
4f8ce00477
update diffusers to 1.16 | finetune
2023-06-01 20:47:54 +09:00
Kohya S
968bbd2f47
Merge pull request #480 from yanhuifair/main
...
fix print "saving" and "epoch" in newline
2023-05-11 21:05:37 +09:00
Kohya S
09c719c926
add adaptive noise scale
2023-05-07 18:09:08 +09:00
Fair
b08154dc36
fix print "saving" and "epoch" in newline
2023-05-07 02:51:01 +08:00
Kohya S
2127907dd3
refactor selection and logging for DAdaptation
2023-05-06 18:14:16 +09:00
青龍聖者@bdsqlsz
164a1978de
Support for more Dadaptation ( #455 )
...
* Update train_util.py for add DAdaptAdan and DAdaptSGD
* Update train_util.py for DAdaptadam
* Update train_network.py for dadapt
* Update train_README-ja.md for DAdapt
* Update train_util.py for DAdapt
* Update train_network.py for DAdaptAdaGrad
* Update train_db.py for DAdapt
* Update fine_tune.py for DAdapt
* Update train_textual_inversion.py for DAdapt
* Update train_textual_inversion_XTI.py for DAdapt
2023-05-06 17:30:09 +09:00
ykume
69579668bb
Merge branch 'dev' of https://github.com/kohya-ss/sd-scripts into dev
2023-05-03 11:17:43 +09:00
Kohya S
2e688b7cd3
Merge pull request #471 from pamparamm/multires-noise
...
Multi-Resolution Noise
2023-05-03 11:17:21 +09:00
ykume
2fcbfec178
make transform_DDP more intuitive
2023-05-03 11:07:29 +09:00
Isotr0py
e1143caf38
Fix DDP issues and Support DDP for all training scripts ( #448 )
...
* Fix DDP bugs
* Fix DDP bugs for finetune and db
* refactor model loader
* fix DDP network
* try to fix DDP network in train unet only
* remove unuse DDP import
* refactor DDP transform
* refactor DDP transform
* fix sample images bugs
* change DDP tranform location
* add autocast to train_db
* support DDP in XTI
* Clear DDP import
2023-05-03 10:37:47 +09:00
Pam
b18d099291
Multi-Resolution Noise
2023-05-02 09:42:17 +05:00
Kohya S
74008ce487
add save_every_n_steps option
2023-04-24 23:22:24 +09:00
Plat
27ffd9fe3d
feat: support wandb logging
2023-04-20 01:41:12 +09:00
Kohya S
9fc27403b2
support disk cache: same as #164 , might fix #407
2023-04-13 21:40:34 +09:00
Kohya S
2e9f7b5f91
cache latents to disk in dreambooth method
2023-04-12 23:10:39 +09:00
AI-Casanova
0d54609435
Merge branch 'kohya-ss:main' into weighted_captions
2023-04-07 14:55:40 -05:00
AI-Casanova
7527436549
Merge branch 'kohya-ss:main' into weighted_captions
2023-04-05 17:07:15 -05:00
Kohya S
541539a144
change method name, repo is private in default etc
2023-04-05 23:16:49 +09:00
Kohya S
74220bb52c
Merge pull request #348 from ddPn08/dev
...
Added a function to upload to Huggingface and resume from Huggingface.
2023-04-05 21:47:36 +09:00
Kohya S
53cc3583df
fix potential issue with dtype
2023-04-03 21:46:12 +09:00
AI-Casanova
1892c82a60
Reinstantiate weighted captions after a necessary revert to Main
2023-04-02 19:43:34 +00:00
ddPn08
b5ff4e816f
resume from huggingface repository
2023-04-02 17:39:21 +09:00
Kohya S
4f70e5dca6
fix to work with num_workers=0
2023-03-28 19:42:47 +09:00
Kohya S
066b1bb57e
fix do not mean in batch dim when min_snr_gamma
2023-03-27 20:47:11 +09:00
Kohya S
6732df93e2
Merge branch 'dev' into min-SNR
2023-03-26 17:10:53 +09:00
