fzh0917/SparseTT

Issues

• RuntimeError: CUDA error: CUBLAS_STATUS_NOT_SUPPORTED when calling `cublasSgemm( handle, opa, opb, m, n, k, &alpha, a, lda, b, ldb, &beta, c, ldc)`

  #17 opened 9 months ago by jasonwjw
  0

• mmcv

  #16 opened a year ago by wsasdsda
  1

• 精度差别太大

  #15 opened 2 years ago by ElliottZhen
  4

• 恢复训练

  #13 opened 2 years ago by C-C-Y
  6

• 请问推理部分是这么聚合两个头的结果的？

  #3 opened 3 years ago by llz-lian
  1

• Welcome update to OpenMMLab 2.0

  #14 opened 2 years ago by vansin
  0

• Did your model predicts the task’s objective directly, without any post-processing steps?

  #11 opened 2 years ago by JananiKugaa
  1

• Target Focus Network

  #12 opened 2 years ago by ChenJian7578
  2

• Training Time

  #10 opened 2 years ago by JananiKugaa
  0

• What are the FLOPs and parameters of your models?

  #9 opened 2 years ago by JananiKugaa
  0

• About Sparse MHSA

  #7 opened 2 years ago by cxq1
  2

• about mask

  #8 opened 2 years ago by INTOUCHABLE-VS
  4

• Computation of sparse self-attention within decoder

  #6 opened 2 years ago by davidyang180
  2

• about got10k test

  #4 opened 3 years ago by USCT-YQJ
  1

• backbone

  #5 opened 3 years ago by INTOUCHABLE-VS
  0

• code

  #2 opened 3 years ago by duweidai
  1

• Where to access the paper?

  #1 opened 3 years ago by zanilzanzan
  1