bkitano/llama-from-scratch

Issues

• next level

  #3 opened 7 months ago by UmarIgan
  1

• Incorrect RMSNorm

  #4 opened 10 months ago by arunmallya
  3

• get_rotary_matrix

  #8 opened 8 months ago by nkkbr
  1

• SwiGLU

  #7 opened 8 months ago by nkkbr
  0

• RoPEMaskedAttentionHead

  #6 opened 8 months ago by nkkbr
  0

• no need to softmax before cross_entrpoy

  #5 opened 8 months ago by nkkbr
  1

• Just to thank you!

  #2 opened a year ago by Andreh1982
  1

• TypeError: MultiheadAttention.forward() got an unexpected keyword argument 'is_causal'

  #1 opened a year ago by bjpcjp
  1