Creates a new variable from arguments. Args: initial_value: A `Tensor`, or Python object convertible to a `Tensor`, which is the initial value for the Variable. The initial value must have a shape specified unless `validate_shape` is set to False. Can also be a cal
(self,
initial_value=None,
trainable=None,
collections=None,
validate_shape=True,
caching_device=None,
name=None,
dtype=None,
expected_shape=None,
constraint=None,
synchronization=None,
aggregation=None,
shape=None)
| 1904 | self.name, self.get_shape(), self.dtype.name) |
| 1905 | |
| 1906 | def _init_from_args(self, |
| 1907 | initial_value=None, |
| 1908 | trainable=None, |
| 1909 | collections=None, |
| 1910 | validate_shape=True, |
| 1911 | caching_device=None, |
| 1912 | name=None, |
| 1913 | dtype=None, |
| 1914 | expected_shape=None, |
| 1915 | constraint=None, |
| 1916 | synchronization=None, |
| 1917 | aggregation=None, |
| 1918 | shape=None): |
| 1919 | """Creates a new variable from arguments. |
| 1920 | |
| 1921 | Args: |
| 1922 | initial_value: A `Tensor`, or Python object convertible to a `Tensor`, |
| 1923 | which is the initial value for the Variable. The initial value must have |
| 1924 | a shape specified unless `validate_shape` is set to False. Can also be a |
| 1925 | callable with no argument that returns the initial value when called. |
| 1926 | (Note that initializer functions from init_ops.py must first be bound to |
| 1927 | a shape before being used here.) |
| 1928 | trainable: If `True`, also adds the variable to the graph collection |
| 1929 | `GraphKeys.TRAINABLE_VARIABLES`. This collection is used as the default |
| 1930 | list of variables to use by the `Optimizer` classes. Defaults to `True`, |
| 1931 | unless `synchronization` is set to `ON_READ`, in which case it defaults |
| 1932 | to `False`. |
| 1933 | collections: List of graph collections keys. The new variable is added to |
| 1934 | these collections. Defaults to `[GraphKeys.GLOBAL_VARIABLES]`. |
| 1935 | validate_shape: If `False`, allows the variable to be initialized with a |
| 1936 | value of unknown shape. If `True`, the default, the shape of |
| 1937 | `initial_value` must be known. |
| 1938 | caching_device: Optional device string or function describing where the |
| 1939 | Variable should be cached for reading. Defaults to the Variable's |
| 1940 | device. If not `None`, caches on another device. Typical use is to |
| 1941 | cache on the device where the Ops using the Variable reside, to |
| 1942 | deduplicate copying through `Switch` and other conditional statements. |
| 1943 | name: Optional name for the variable. Defaults to `'Variable'` and gets |
| 1944 | uniquified automatically. |
| 1945 | dtype: If set, initial_value will be converted to the given type. If None, |
| 1946 | either the datatype will be kept (if initial_value is a Tensor) or |
| 1947 | float32 will be used (if it is a Python object convertible to a Tensor). |
| 1948 | expected_shape: Deprecated. Ignored. |
| 1949 | constraint: An optional projection function to be applied to the variable |
| 1950 | after being updated by an `Optimizer` (e.g. used to implement norm |
| 1951 | constraints or value constraints for layer weights). The function must |
| 1952 | take as input the unprojected Tensor representing the value of the |
| 1953 | variable and return the Tensor for the projected value (which must have |
| 1954 | the same shape). Constraints are not safe to use when doing asynchronous |
| 1955 | distributed training. |
| 1956 | synchronization: Indicates when a distributed a variable will be |
| 1957 | aggregated. Accepted values are constants defined in the class |
| 1958 | `tf.VariableSynchronization`. By default the synchronization is set to |
| 1959 | `AUTO` and the current `DistributionStrategy` chooses when to |
| 1960 | synchronize. |
| 1961 | aggregation: Indicates how a distributed variable will be aggregated. |
| 1962 | Accepted values are constants defined in the class |
| 1963 | `tf.VariableAggregation`. |
no test coverage detected