lucidrains/video-diffusion-pytorch

Issues

• Implementation of reconstruction-guided sampling in VDM

  #20 opened 2 years ago by wenyihong
  2

• Text and video training

  #38 opened 6 months ago by zzzbbt
  0

• can't run the code

  #33 opened a year ago by Zyj1moreT
  1

• Generating longer videos at test time

  #4 opened 3 years ago by mrkulk
  46

• Any one get any video result by training this project?

  #34 opened a year ago by double-fire-0
  6

• result rendering

  #18 opened 2 years ago by farhad-abdi
  1

• Training is not steady (exploding gradient ) when training with UCF101 datasets and base_channel_size=256.

  #37 opened 10 months ago by shim0114
  0

• Problem about cond_scale

  #35 opened 10 months ago by kanydao
  2

• Inferencing on a custom .gif after training the model

  #29 opened 10 months ago by chpk
  2

• Why SpatialLinearAttention use k mul v first?

  #30 opened a year ago by TtuHamg
  0

• Conditioning on image + text embedding

  #7 opened 3 years ago by ChintanTrivedi
  4

• video prediction

  #26 opened a year ago by XXinLiu
  3

• Images are 2-D, how to adapt this for 1-D data ?

  #28 opened a year ago by mayank64ce
  4

• #after a lot of training

  #27 opened a year ago by MehdiGhorb
  2

• Noisy output & "text_use_bert_cls" error

  #13 opened 3 years ago by GoutamKelam
  31

• What is this structure?

  #25 opened 2 years ago by jacob-kang
  0

• Class conditioned training

  #24 opened 2 years ago by Indra24710
  0

• Groupnorm is not factorised across time?

  #23 opened 2 years ago by huddyyeo
  0

• Where is the output saved if running in google collab?

  #22 opened 2 years ago by jackson-sandland
  0

• why the code has two versions of spitial transformer?

  #21 opened 2 years ago by diaodeyi
  0

• Is there any pertrained model?

  #19 opened 2 years ago by diaodeyi
  2

• the commitment: one more residual

  #14 opened 2 years ago by liangbingzhao
  26

• Gradient method for conditional sampling

  #9 opened 3 years ago by LuckyDC
  4

• Better Co-training Images and Video Method In imagen-video Paper

  #17 opened 2 years ago by lxj616
  0

• What are the length (frames) of x^a and x^b when sampling autoregressively?

  #16 opened 2 years ago by eri24816
  0

• Elucidated version

  #15 opened 2 years ago by liangbingzhao
  1

• Reason for combining rotary and relative positional embedding?

  #12 opened 3 years ago by oxjohanndiep
  1

• Training Dataset

  #5 opened 3 years ago by Mq-Zhang1
  2

• Duplicate dividing in relative positional encoding

  #11 opened 3 years ago by songweige
  1

• How to save a video？

  #2 opened 3 years ago by 980202006
  8

• How would one change the Trainer class for text conditioning?

  #10 opened 3 years ago by samlhuillier
  0

• related recent work

  #8 opened 3 years ago by rom1504
  1

• attn mask suppose be `~mask`

  #6 opened 3 years ago by CiaoHe
  2

• Unconditioned UNet

  #3 opened 3 years ago by shehan360
  1

• Just want to say this looks exciting!

  #1 opened 3 years ago by Mut1nyJD
  1