archinetai/audio-diffusion-pytorch

Issues

• NaN after training for a while

  #52 opened a year ago by jameshball
  10

• Weird spikes in the loss

  #84 opened a month ago by return-nihil
  0

• RuntimeError: The size of tensor a (37) must match the size of tensor b (36) at non-singleton dimension 2

  #81 opened 5 months ago by heury
  4

• RuntimeError: The size of tensor a (91) must match the size of tensor b (90) at non-singleton dimension 2

  #83 opened 3 months ago by erikqu
  0

• Unconditional Generation generates noise

  #82 opened 3 months ago by reachomk
  0

• Model architectures from the paper

  #66 opened a year ago by AI-Guru
  4

• Unconditional model generates okay quality of fake human voice but failed on music.

  #80 opened 8 months ago by piobmx
  4

• Questions about conditional generation

  #61 opened a year ago by AI-Guru
  2

• What is the structure of encoder in diffusionAE?

  #78 opened 8 months ago by SuperiorDtj
  1

• CUDA OF Memory for 80GB A100 : follow the mousai paper setting of text condition

  #79 opened 8 months ago by SuperiorDtj
  0

• I have a few questions about 1D-UNet

  #76 opened 10 months ago by 0417keito
  4

• Class-conditional generation

  #72 opened a year ago by aqibsaeed
  1

• AssertionError: ClassiferFreeGuidancePlugin requires embedding

  #71 opened a year ago by gg4u
  0

• Can the repo be used to process MIDI data？

  #69 opened a year ago by zsy1987
  0

• Future Work - Models

  #67 opened a year ago by AI-Guru
  0

• Languages

  #64 opened a year ago by barredo
  1

• Trained models

  #63 opened a year ago by aoezis
  0

• New Try

  #60 opened a year ago by Leezp99
  0

• could provide a example recipe?

  #51 opened a year ago by gandolfxu
  12

• Add support to clip predicted samples to the desired range.

  #55 opened a year ago by Kinyugo
  2

• How to just train condition audio-diffusion without text-condition?

  #49 opened a year ago by LeonJoe13
  3

• Spectrogram-based diffusion model

  #59 opened a year ago by Tinglok
  2

• Alternative Noises: Offset, Pyramid, Pink

  #56 opened a year ago by torridgristle
  2

• What loss function is being used?

  #53 opened a year ago by kitchWWW
  2

• How to use our own background noisy dataset to generate sample?

  #48 opened a year ago by haloha123
  1

• Question: the sigma_t is not samped from 0 to 1 in v-diffusion, which is not like your thesis mentioned, will it cause any trouble?

  #50 opened a year ago by emailandxu
  1

• VRAM requirements?

  #44 opened a year ago by illtellyoulater
  3

• Typo in Paper

  #45 opened a year ago by hu-po
  1

• How to convert to wav file to listen to result?

  #46 opened a year ago by dillfrescott
  6

• Support usage with non-audio data e.g spectrograms

  #39 opened a year ago by Kinyugo
  7

• can not open music examples websit

  #47 opened a year ago by Liujingxiu23
  2

• Am I training the model correctly?

  #33 opened 2 years ago by cat-policlot
  2

• example for Unconditional Generator fails

  #43 opened a year ago by MultiTrickFox
  2

• custom dataset

  #38 opened a year ago by lxa9867
  0

• training with conditioning t5

  #34 opened 2 years ago by nikuson
  5

• Exploding loss

  #35 opened 2 years ago by alexrodi
  3

• Pre-trained Weights of AutoEncoder

  #27 opened 2 years ago by JustinYuu
  1

• Error Locating Target

  #26 opened 2 years ago by ModeratePrawn
  1

• nan outputs when the number of sampling steps is set to 1

  #15 opened 2 years ago by Kinyugo
  3

• Using the audio_975 model with colab fails

  #21 opened 2 years ago by timohear
  2

• Question: Scaling guide/suggested parameters?

  #5 opened 2 years ago by zaptrem
  10

• Add trainer

  #3 opened 2 years ago by nateraw
  4

• text conditioned infinite ASMR generator?

  #2 opened 2 years ago by lucidrains
  4