Skip to content

Issues: kohya-ss/sd-scripts

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

675 Open 541 Closed
675 Open 541 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

SDXL LoRA Training Stuck at Step 0
#1939 opened Feb 16, 2025 by hanpi2233
1
Bug❓: --cache_latents prevents regular checkpoint saving in train_network.py 😥
#1937 opened Feb 15, 2025 by Jackiiiii
2
BUG: Validation breaks cached image sizes
#1934 opened Feb 14, 2025 by feffy380
3
WARNING Unsuppored keys for sd-scripts ！
#1933 opened Feb 14, 2025 by Shmtu-Herven
Train flux api
#1932 opened Feb 14, 2025 by iemskl
use multigpu(8*A800 80G) train flux_train.py OOM problem
#1930 opened Feb 13, 2025 by godwenbin
2
The lora+ setting and prodigy cannot be used together.
#1929 opened Feb 12, 2025 by ytoaa
1
Can the algorithm for automatic scaling of ARB buckets be specified?
#1928 opened Feb 12, 2025 by TLFZ1
1
Separate Learning Rates for UNET and TE with Prodigy
#1926 opened Feb 9, 2025 by MarvyD
1
Multi GPU need a fix : to perform barrier as devices used by this process are currently unknown. This can potentially cause a hang if this rank to GPU mapping is incorrect.Specify device_ids in barrier() to force use of a particular device,or call
#1925 opened Feb 7, 2025 by FurkanGozukara
Support Lumina 2.0
#1924 opened Feb 7, 2025 by sdbds
11
Why have the options to define the single and double blocks_to_swap been removed?
#1923 opened Feb 7, 2025 by Deathawaits4
1
--save_state doesn't produce anything?
#1921 opened Feb 5, 2025 by alexgilseg
3
FLUX-Fill-LoRa-Training and Clip-L training when doing Fine Tuning / DreamBooth with FLUX Dev
#1920 opened Feb 5, 2025 by FurkanGozukara
flux_minimal_inference: CUDA Out of Memory with LORA and negative_prompt
#1919 opened Feb 3, 2025 by mamawr
Can not generate full body with face lora model.
#1916 opened Feb 3, 2025 by 1179021477
Hardcoded bucket_step overrides configurable parameter bug Something isn't working
#1915 opened Feb 1, 2025 by iqddd
4
Problems with deliverables when using the full_fp16 option in lora learning
#1914 opened Jan 31, 2025 by ytoaa
6
Voluptuous is down
#1912 opened Jan 30, 2025 by xingqiu774
how to set timestep shift when train flux-dev lora?
#1907 opened Jan 28, 2025 by SlZeroth
1
Shouldn't fix_noise_scheduler_betas_for_zero_terminal_snr be applied BEFORE prepare_scheduler_for_custom_training? help wanted Extra attention is needed
#1905 opened Jan 27, 2025 by 67372a
2
Flux training, no matter the settings I get: returned non-zero exit status 3221225477
#1902 opened Jan 27, 2025 by alexgilseg
2
Implement Test and Validation Set Loss
#1897 opened Jan 24, 2025 by stepfunction83
2
Does the full fine-tuning in SD3.5 have a setting similar to --guidance_scale?
#1896 opened Jan 24, 2025 by raindrop313
lora extracted from dreambooth trained model has poor compatibility with controlnet
#1892 opened Jan 23, 2025 by Nomination-NRB
ProTip! Adding no:label will show everything without a label.