Skip to main content

Mojo function

gpu_qint4_repack_GPTQ

gpu_qint4_repack_GPTQ[group_size: Int, target: StringSlice[StaticConstantOrigin]](b: LayoutTensor[DType.uint8, layout, origin, address_space=address_space, element_layout=element_layout, layout_int_type=layout_int_type, linear_idx_type=linear_idx_type, masked=masked, alignment=alignment], b_packed: LayoutTensor[DType.uint8, layout, origin, address_space=address_space, element_layout=element_layout, layout_int_type=layout_int_type, linear_idx_type=linear_idx_type, masked=masked, alignment=alignment], perm_idx: OptionalReg[LayoutTensor[DType.int32, Layout.row_major(-1), MutableAnyOrigin]] = None, ctx: DeviceContextPtr = DeviceContextPtr())

Was this page helpful?