[Mlir-commits] [mlir] [mlir][tosa] Optimize block scaled cast sequences (PR #188018)

[Hocky Yudhiono]

hockyy wrote:

I see, basically the 4bit floating point is has lower precision

and downcasting (to_block_scaled) is basically pruning some extra data

https://github.com/llvm/llvm-project/pull/188018