hahnyuan/PB-LLM

Issues

• Why the importance of weights evaluated by w^2/(H_ii)^2, instead of w^2/(H_ii) like in SparseGPT?

  #6 opened 4 months ago by Xingrun-Xing
  2

• what is the practical speedup ?

  #5 opened 8 months ago by XA23i
  0

• Compression Ratio

  #1 opened a year ago by NicoNico6
  1

• Details about the code

  #4 opened 8 months ago by Hon-Chen
  2

• CUDA out of memory

  #3 opened 9 months ago by duany049
  0