lean-dojo / FloatLib

git clone https://github.com/lean-dojo/FloatLib.git

Arbitrary verified fast precision floating point arithmetic in Lean, with proofs, optimized backends, and support for IEEE binary and decimal, posits, and custom formats.