CaraJ7/CoMat

Issues

• Where is Ltxt? How does the Mixed Latent Strategy involved in training?

  #13 opened 4 months ago by lrzjason
  3

• checkpoints when?

  #12 opened 4 months ago by foureyednymph
  0

• How can I get train_data/gan_abc5k_t2comp_hrs_20k_sd15.jsonl data?

  #10 opened 5 months ago by ugiugi0823
  1

• Why not directly use resolution 1024 for SDXL?

  #9 opened 5 months ago by kjzju
  1

• Training Error: ModuleNotFoundError: No module named 'TrainableSDPipeline'

  #11 opened 5 months ago by rdcoder33
  2

• reward_norm == 0 and no weights update for lora?

  #8 opened 6 months ago by imyangs
  2

• 训练的最低配置

  #7 opened 6 months ago by ysm2000
  1

• 可以开源训练好的模型么

  #6 opened 7 months ago by sifa868
  1

• about citation 55 in the paper.

  #5 opened 7 months ago by icoz69
  2

• Can you release a CoMat-SDXL-turbo model ?

  #4 opened 7 months ago by ApolloRay
  0

• 大佬们有开源权重的计划么？

  #3 opened 8 months ago by wangqixun
  1

• Excuse me, does the author have a demo?

  #1 opened 8 months ago by shams2023
  2

• how to write L_cap in pytorch?

  #2 opened 8 months ago by tian-2024
  1