Class MultiResponseMessage
Defined in File multi_response.hpp
Base Type
public morpheus::DerivedMultiMessage< MultiResponseMessage, MultiTensorMessage >
(Template Class DerivedMultiMessage)
-
class MultiResponseMessage : public morpheus::DerivedMultiMessage<MultiResponseMessage, MultiTensorMessage>
This class is used to get or set the inference output from message containers derived from ResponseMemory.
Public Functions
- MultiResponseMessage(const MultiResponseMessage &other) = default
Default copy constructor.
Construct a new Multi Response Message object.
- Parameters
meta – Holds a data table, in practice a cudf DataFrame, with the ability to return both Python and C++ representations of the table
mess_offset – Offset into the metadata batch
mess_count – Messages count
memory – Shared pointer of a tensor memory
offset – Message offset in inference memory instance
count – Message count in inference memory instance
id_tensor_name – Name of the tensor that correlates tensor rows to message IDs
probs_tensor_name – Name of the tensor that holds output probabilities
-
const TensorObject get_output(const std::string &name) const
Returns the output tensor with the given name.
- Parameters
- Throws
- Returns
name –
std::runtime_error – If no tensor matching
name
existsconst TensorObject
-
TensorObject get_output(const std::string &name)
Returns the output tensor with the given name.
- Parameters
- Throws
- Returns
name –
std::runtime_error – If no tensor matching
name
exists
-
void set_output(const std::string &name, const TensorObject &value)
Update the value of a given output tensor. The tensor must already exist, otherwise this will halt on a fatal error.
- Parameters
name –
value –
-
TensorObject get_probs_tensor() const
Get the tensor that holds output probabilities. Equivalent to
get_tensor(probs_tensor_name)
- Returns
const TensorObject
Public Members
- std::string probs_tensor_name