Is the transformation type one-to-one or not. Returns: bool: ``True`` denotes injective. ``False`` denotes non-injective.

Load the stored model, parameters and optimizer states. Args: path (str): The prefix of files storing the model states a

Apply `func` to each entry in `structure` and return a new structure.

Generate reparameterized samples of the specified shape. Args: shape (Sequence[int], optional): Shape of the generated samples.

Remove nodes safely since links connected to these removed nodes are also removed. Args: remove_nodes(set): the

Set graph configuration to the IpuStrategy instance. Args: num_ipus (int, optional): Number of IPU devices. Default 1, w

Declare a Variable as input with random initialization.

For parse argruments from args and kwargs Get item from PyObject* args or PyObject* kwargs

getter

Call _print_tensor_in_gpu and capture its C-level stdout output. CUDA device-side printf writes to fd 1, so we redirect fd 1 to a temporary f

x_shape: original shape new_shape: new shape format: "coo" or "csr" Transform a sparse tensor with shape "x_shape" to

x_shape: original shape new_shape: new shape static graph only supports coo format. Transform a sparse tensor with sh

Full name for this layer, composed by name_scope + "/" + MyLayer.__class__.__name__ Returns: str, full name of this lay

Get the context held by this operation state.

The number of :ref:`api_guide_Block_en` in this Program. .. note:: This API has no effect in Dygraph mode. Ret

Trace the value of the tracked variables from the frame. It used for generating the guard. Returns: The value of the tra

Get current worker id. Returns: int: node id

Create a new variable from a given value, or return None if the value cannot be converted to a variable. Args: value (Any