Return pointer to IDimensionExp that represents the given operation applied to first and second. Returns nullptr if op is not a valid DimensionOperati

\brief get the logger with which the builder was created \return the logger

\brief Get the number of inputs of a layer.

Marks a safe string as being inline when used with other Script APIs. Non-inlined strings will remain strings when used with any of these API

\brief Set the dimensions of a tensor. For a network input, the dimensions are assigned by the application. For a network output, the dimensions are

\brief Get the dimensions of a tensor. \return The dimensions of the tensor. \warning getDimensions() returns a -1 for dimensions that are derived f

Copy constructor. \tparam T type of concrete half expression \param rhs half expression to copy from

Locate path to file, given its filename or filepath suffix and possible dirs it might lie in. Function will also walk back MAX_DEPTH dirs from CWD to

Logs a message to stdout with START severity. Args: message (Union[str, Callable() -> str]): A strin

Empty the errorStack.

Creates a generator that yields runner names (str).

\brief Set the name of the loop. The name is used in error diagnostics. This method copies the name string. \warning The string name must be null-te

Convenience function for adding entries. Args: name (str): The name of the input. dtype (numpy.dtype): The d

Compute the amax from the collected histogram Args: method: A string. One of ['entropy', 'mse', 'percentile'] Keyword Ar

Creates a TensorRT BuilderConfig according to arguments provided on the command-line. Args: builder (trt.Builder):

Returns a Pytorch "dynamic_axes" encoding for onnx.export. Returns: dict: Returns a 'dynamic' index with corresponding n

Append a line to the suffix of the script Args: line (str): The line to append.

\brief Query whether the network was created with an implicit batch dimension. \return True if tensors have implicit batch dimension, false otherwise

Returns a context manager that temporarily changes the severity of the logger for its duration. Args: severity (Logger.S

A convenience function to add shapes for a single binding. Args: name (str): The name of the binding. min (T

\brief Return the type of a layer. \see LayerType

Supplies calibration data to TensorRT to calibrate the network for INT8 inference. Args: data_loader (Sequence[OrderedDict[str, Unio

Absolute value. \param arg operand \return absolute value of \a arg template<typename T> typename enable<T,T>::type abs(T arg) { return unary_speciali

Copies from the provided host buffer into this device buffer. Args: host_buffer (numpy.ndarray): The

\brief Deserialize an engine from a stream. If an error recorder has been set for the runtime, it will also be passed to the engine. \param blob The

\brief Get the dimensions of a binding. \param bindingIndex The binding index. \return The dimensions of the binding if the index is in range, otherw

Helper function for __init__ of quantized module Create input and weight quantizer based on quant_desc passed by kwargs, or default of the cl

Skips the batches

Convert Keras saved model into ONNX format. Works directly with CreateAssetsFolder object path. Args: saved_model_dir (str): Path to

Returns a context manager that indents all strings logged by the specified amount. Args: level (int): The indentation le

Logs a message to stdout. Args: message (Union[str, Callable() -> str]): A string or callable which

Parse the ONNX graph and create the corresponding TensorRT network definition. :param onnx_path: The path to the ONNX graph to load.

\brief Returns the format of an Algorithm input or output. Algorithm inputs are incrementally numbered first, followed by algorithm outputs. \param in

\brief Get the layer specified by the given index. \param index The index of the layer. \return The layer, or nullptr if the index is out of range.

\brief Get the number of layers in the network. The number of layers in the network is not necessarily the number in the original network definition,

Runs inference using the provided feed_dict. Must be called only after ``activate()`` and before ``deactivate()``. NOTE: So

Returns a new sequence dimension. Return: str: Returns a sequence dimension which Dims.SEQUENCE appended by dim_type.

Interleave input tensor with `num_beams`, similar to HuggingFace's _expand_inputs_for_generation() in generation_utils.py

Returns a context manager that temporarily changes the severity of the logger for its duration. Args: severity (Union[in

Tracks the absolute max of all tensors Args: x: A tensor Raises: RuntimeError: If amax shape changes

\brief Create an execution context. The execution context created will call setOptimizationProfile(0) implicitly if there are no other execution cont

\brief Retrieve the binding index for a named tensor. IExecutionContext::enqueueV2() and IExecutionContext::executeV2() require an array of buffers.

\brief Get the operation for the layer. \see setOperation(), TopKOperation

Get the global TensorRT logger created by Polygraphy. Returns: trt.Logger: The TensorRT logger.

Insert Q/DQ nodes in Keras model and return a copy. Weights are preserved unlike native keras clone. Args: model(tf.keras.Model): Ke

Encode an object as JSON and save it to a file. NOTE: For Polygraphy objects, you should use the ``save()`` method instead. Args:

Square root implementation. \param arg function argument \return function value stored in single-preicision

Creates a wrapper to emulate quantization for the Add keras layer. Args: layer (tf.keras.layers.Layer): The keras layer to

\brief Add an activation layer to the network. \param input The input tensor to the layer. \param type The type of activation function to apply. Not

\brief Add an If-conditional input. \param input An input to the conditional that can be used by either or both of the conditional’s subgraphs. \see

Loads a file and decodes the JSON contents. NOTE: For Polygraphy objects, you should use the ``load()`` method instead. Args: s

Args: value: An instance of :class:`QuantDescriptor <pytorch_quantization.tensor_quant.QuantDescriptor>`