adobe-research/custom-diffusion

Issues

• NameError: name 'trainer' is not defined

  #99 opened 3 months ago by wenkfjsf
  2

• retrieve.py question

  #98 opened 7 months ago by kuailexiaohunzi
  2

• Is this version trained with lora?

  #96 opened 8 months ago by Dumeowmeow
  1

• Is it possible to train multi joint training with SDXL? diffusers_training_sdxl.py seems impossible for multi joint training.

  #89 opened a year ago by taehyeonEum
  2

• knn.laion.ai link error

  #75 opened a year ago by jy-choi-git
  2

• Tensors not Updated after Training

  #97 opened 8 months ago by ChrisWang13
  1

• Does the diffusers version now work properly?

  #92 opened 10 months ago by wsfk716
  3

• 'Attention' object has no attribute 'cross_attention_norm'

  #68 opened a year ago by kopyl
  4

• KeyError in visualize attn maps

  #76 opened a year ago by OchiaiNao
  4

• Missing text prompts?

  #91 opened 9 months ago by Shentao-YANG
  2

• Unable to retrieve the image used for regularization.

  #95 opened 9 months ago by LZHLilE
  1

• why the results produced by the model training and verification are inconsistent with the results I test using the saved model weights?

  #93 opened 9 months ago by baimazhaoge
  1

• Question about loss on valid image region during training

  #90 opened a year ago by JoyHuYY1412
  1

• Evaluation：where is the “prompts.json”file？

  #94 opened 9 months ago by BlackPuuuudding
  0

• Plan for integration of SDXL? It seems to require more work than changing the MODEL_NAME

  #73 opened a year ago by asrlhhh
  6

• How to Visualize Attention Maps to Analyze Compositions?

  #70 opened a year ago by moatifbutt
  2

• why my optimization based weights merging is not work?

  #88 opened a year ago by ZZZBBBZZZ
  4

• OOM while running sample.py

  #87 opened a year ago by daeunni
  4

• if it OK with no regularization? my results don't have any consistancy with reference images.

  #85 opened a year ago by taehyeonEum
  3

• Ours (w/fine-tune all) ？

  #86 opened a year ago by wsfk716
  2

• Hi, How toMultiple-Concept Fine-tuning for DreamBooth?

  #84 opened a year ago by liushao
  0

• ImportError: cannot import name 'insecure_hashlib' from 'huggingface_hub.utils' 怎么解决？

  #83 opened a year ago by zhangyun-w
  1

• Question about the cross attention

  #82 opened a year ago by qiguming
  1

• A question for the regularization captions

  #79 opened a year ago by shuhao02
  1

• questions about crossattn in CustomDiffusionAttnProcessor

  #80 opened a year ago by qiguming
  0

• No module named 'diffusers.models.cross_attention'

  #77 opened a year ago by htrvu
  1

• Errors in using diffuser version

  #78 opened a year ago by OchiaiNao
  3

• current code is not compatible with new diffusers

  #49 opened 2 years ago by eps696
  9

• I have a bug with the diffusers

  #74 opened a year ago by ohad204
  1

• Moon Gate+ Evaluation q

  #66 opened a year ago by isamu-isozaki
  3

• Analysis of change in weights on updating all network weights during fine-tuning

  #67 opened a year ago by Yi-Qi638
  1

• CustomConcept101 dataset

  #71 opened a year ago by KaiyueSun98
  3

• PLZ HAVE A LOOK!!! Is there a use of clip-retrieval during fine-tuning of faces?

  #72 opened a year ago by SKDDJ
  3

• Is it possible for the fine-tuned model to do inpainting job rather than image generation?

  #61 opened a year ago by asrlhhh
  3

• How to train the model withot using the regularized dataset

  #69 opened a year ago by Aditi0912dec
  2

• How to show cross attention map for checking the issue of failed compositions?

  #54 opened 2 years ago by BossunWang
  3

• Can I fine-tune it using other pre-trained model other than sd-v1-4.ckpt?

  #65 opened 2 years ago by abrahamwoubie
  1

• TypeError: ModelCheckpoint.__init__() got an unexpected keyword argument 'every_n_epochs'

  #63 opened 2 years ago by sam-motamed
  2

• License

  #64 opened 2 years ago by kietbg0079
  1

• About the train data and val data!

  #62 opened 2 years ago by wangyePHD
  1

• Anyway to run it with stable diffusion 2.1?

  #60 opened 2 years ago by asrlhhh
  2

• KID range

  #59 opened 2 years ago by Yuheng-Li
  1

• Must I use regulization imgs retrieved from laion?

  #58 opened 2 years ago by CoderZhangYx
  2

• questions about the src/convert.py

  #53 opened 2 years ago by yqjiang6
  1

• problem when using diffusers for training

  #57 opened 2 years ago by yuxu915
  2

• Multiple ckpt after training. Which one should I use for sampling?

  #55 opened 2 years ago by kasim0226
  2

• Huge gap text-alignment score for textual inversion (original paper and your replication)

  #56 opened 2 years ago by studying910
  4

• About the training of input embedding.

  #52 opened 2 years ago by TousakaNagio
  5

• KeyError: 'cond_stage_model.transformer.text_model.embeddings.token_embedding.weight'

  #51 opened 2 years ago by Carlyx
  2

• Connection refused by the server

  #50 opened 2 years ago by keno-log
  2