CiaraStrawberry/svd-temporal-controlnet

Issues

• bug in encode_image(pixel_values) in train_svd.py?

  #24 opened 2 months ago by Joseph-LIN-Tech
  0

• Is there any way to enhance temporal consistency?

  #13 opened 2 months ago by DINGYANB
  0

• What is purpose of the motion values?

  #14 opened a year ago by phil329
  3

• "exp_vml_cpu" not implemented for 'Half'

  #7 opened a year ago by weijiawu
  7

• Discussion with CFG

  #19 opened 9 months ago by jiangzhengkai
  1

• some other conditions

  #8 opened a year ago by pokameng
  3

• diffusers==0.28.0 is not suitable for controlnet?

  #22 opened 7 months ago by Yww12
  1

• sv3d

  #23 opened 5 months ago by YuzhiChen001
  0

• {'controlnet'} was not found in config. Values will be initialized to default values.

  #10 opened a year ago by pokameng
  2

• why scheduling_euler_discrete_karras_fix.py is needed instead of the official verison?

  #21 opened 8 months ago by yijicheng
  0

• What is motion value?

  #20 opened 8 months ago by guanjunwu
  1

• Sampling of sigma

  #15 opened a year ago by pixeli99
  1

• How to call this control node?

  #18 opened 9 months ago by ALONEONE123
  0

• Great Work. Curious about more details for training.

  #17 opened 10 months ago by jacobswan1
  0

• Inquiry on VRAM and training time requirement

  #16 opened 10 months ago by Junyi42
  0

• More explanations for Motion value

  #6 opened a year ago by weijiawu
  3

• noise_scheduler

  #11 opened a year ago by pokameng
  0

• Torch not compiled with CUDA enabled

  #9 opened a year ago by curlysasha
  2

• Size of output

  #5 opened a year ago by deeplearn-art
  2

• Do you have any plans to implement lora for Temporal Attention?

  #4 opened a year ago by guoqincode
  0

• Code in repo not working, outdated?

  #2 opened a year ago by BlakeOne
  4

• reference materials

  #1 opened a year ago by LiuShiyu95
  2