TorchLean: Formalizing Neural Networks in Lean

2 comments

• measurablefunc 2 hours ago ago

  I guess the next step would be adding support for quantized arithmetic.

  [-]
  • pstoll an hour ago ago

    And the lower precision float variants.