Function quantize_blockwise

Quantize tensor A in blocks of size 4096 values. Quantizes tensor A by dividing it into blocks of 4096 values. Then the absolute maximum value within these blocks is calculated for the non-linear quantization. Parameters ---------- A : torch.Tensor The input te

(
    A: Tensor,
    code: Optional[torch.Tensor] = None,
    absmax: Optional[torch.Tensor] = None,
    out: Optional[torch.Tensor] = None,
    blocksize=4096,
    nested=False,
)