com.microsoft - DequantizeWithOrder#
DequantizeWithOrder - 1 (com.microsoft)#
Version
domain: com.microsoft
since_version: 1
function:
support_level:
shape inference:
This version of the operator has been available since version 1 of domain com.microsoft.
Summary
Dequantize input matrix to specific layout used in cublaslt. attr to specify output type, float16 or float32
Attributes
order_input (required): cublasLt order of input matrix. See the schema of QuantizeWithOrder for order definition. Default value is
?.order_output (required): cublasLt order of output matrix Default value is
?.to (required): The output data type, only support TensorProto_DataType_FLOAT (1) and TensorProto_DataType_FLOAT16 (10) Default value is
?.
Inputs
input (heterogeneous) - Q: TODO: input tensor of (ROWS, COLS). if less than 2d, will broadcast to (1, X). If 3d, it is treated as (B, ROWS, COS)
scale_input (heterogeneous) - S: scale of the input
Outputs
output (heterogeneous) - F: output tensor
Examples