lucidrains/linear-attention-transformer

Issues

• [Feature request] Self-attention with Persistent Memory

  #21 opened 5 months ago by MarcusLoppe
  1

• Image linear attention reference

  #19 opened 9 months ago by pravn
  0

• Why dim != dim_head * heads?

  #18 opened 9 months ago by zzczzc20
  0

• How to perform training?

  #17 opened 2 years ago by pangshengwei
  0

• Is the causal attention really works here?

  #16 opened 2 years ago by charlesxu90
  2

• Scaling factors

  #14 opened 2 years ago by radandreicristian
  1

• Tooooo many functions added, but no annotations

  #15 opened 2 years ago by charlesxu90
  0

• Questions on the implementation of a linear variant and reference

  #8 opened 4 years ago by scaomath
  1

• ImageLinearAttention showcase

  #11 opened 3 years ago by monajalal
  0

• Challenge in replacing SelfAttention with ImageLinearAttention in Vision Transformer

  #13 opened 3 years ago by monajalal
  0

• .

  #12 opened 3 years ago by monajalal
  0

• Causal linear attention from which paper ? please tell me thx

  #10 opened 3 years ago by hquzhuguofeng
  0

• dalle

  #9 opened 3 years ago by adamonkey
  1

• Loss returns Nan

  #6 opened 4 years ago by terencenwz
  3

• Where does this constant come from?

  #7 opened 4 years ago by aluo-x
  1

• causal = True

  #5 opened 4 years ago by wajihullahbaig
  2

• Autopadder doesn't work with LinearAttentionTransformer

  #4 opened 4 years ago by jamarju
  1

• Positional encoding?

  #3 opened 4 years ago by matthew-jurewicz
  2

• [Question] Merging with Trans-XL?

  #2 opened 4 years ago by gaceladri
  40

• seq2seq decoder ids

  #1 opened 4 years ago
  4