HN

Comments (5)

• insumanth
  Anyone not doing distributed training across thousands of GPU, pytorch is the best option. Others have an infra team to work on it.
• Menu

  ansk
  This kind of sounds like just giving up on making the compiler better.
• madhu_ghalame
  The idea makes sense because researchers already think in PyTorch. The tooling should adapt to them, not the other way around.
• dmzxnico
  [flagged]