I found a tool[0] that helps me debug potential floating point issues when they arise. This one has modes for half-, single- and double-precision IEEE-754 float

I found a tool[0] that helps me debug potential floating point issues when they arise. This one has modes for half-, single- and double-precision IEEE-754 floats, so I can deal with the various issues when converting between 64-bit and 32-bit floats. [0] https://news.ycombinator.com/item?id=29370883 https://news.ycombinator.com/item?id=29370883