Python tensorflow.Dtype() Examples
The following are 12
code examples of tensorflow.Dtype().
You can vote up the ones you like or vote down the ones you don't like,
and go to the original project or source file by following the links above each example.
You may also want to check out all available functions/classes of the module
tensorflow
, or try the search function
.
Example #1
Source File: nn.py From THUMT with BSD 3-Clause "New" or "Revised" License | 6 votes |
def maxout(inputs, output_size, maxpart=2, use_bias=True, concat=True, dtype=None, scope=None): """ Maxout layer :param inputs: see the corresponding description of ``linear'' :param output_size: see the corresponding description of ``linear'' :param maxpart: an integer, the default value is 2 :param use_bias: a boolean value indicate whether to use bias term :param concat: concat all tensors if inputs is a list of tensors :param dtype: an optional instance of tf.Dtype :param scope: the scope of this layer, the default value is ``maxout'' :returns: a Tensor with shape [batch, output_size] :raises RuntimeError: see the corresponding description of ``linear'' """ candidate = linear(inputs, output_size * maxpart, use_bias, concat, dtype=dtype, scope=scope or "maxout") shape = tf.concat([tf.shape(candidate)[:-1], [output_size, maxpart]], axis=0) value = tf.reshape(candidate, shape) output = tf.reduce_max(value, -1) return output
Example #2
Source File: nn.py From Document-Transformer with BSD 3-Clause "New" or "Revised" License | 6 votes |
def maxout(inputs, output_size, maxpart=2, use_bias=True, concat=True, dtype=None, scope=None): """ Maxout layer :param inputs: see the corresponding description of ``linear'' :param output_size: see the corresponding description of ``linear'' :param maxpart: an integer, the default value is 2 :param use_bias: a boolean value indicate whether to use bias term :param concat: concat all tensors if inputs is a list of tensors :param dtype: an optional instance of tf.Dtype :param scope: the scope of this layer, the default value is ``maxout'' :returns: a Tensor with shape [batch, output_size] :raises RuntimeError: see the corresponding description of ``linear'' """ candidate = linear(inputs, output_size * maxpart, use_bias, concat, dtype=dtype, scope=scope or "maxout") shape = tf.concat([tf.shape(candidate)[:-1], [output_size, maxpart]], axis=0) value = tf.reshape(candidate, shape) output = tf.reduce_max(value, -1) return output
Example #3
Source File: nn.py From transformer-aan with BSD 3-Clause "New" or "Revised" License | 6 votes |
def maxout(inputs, output_size, maxpart=2, use_bias=True, concat=True, dtype=None, scope=None): """ Maxout layer :param inputs: see the corresponding description of ``linear'' :param output_size: see the corresponding description of ``linear'' :param maxpart: an integer, the default value is 2 :param use_bias: a boolean value indicate whether to use bias term :param concat: concat all tensors if inputs is a list of tensors :param dtype: an optional instance of tf.Dtype :param scope: the scope of this layer, the default value is ``maxout'' :returns: a Tensor with shape [batch, output_size] :raises RuntimeError: see the corresponding description of ``linear'' """ candidate = linear(inputs, output_size * maxpart, use_bias, concat, dtype=dtype, scope=scope or "maxout") shape = tf.concat([tf.shape(candidate)[:-1], [output_size, maxpart]], axis=0) value = tf.reshape(candidate, shape) output = tf.reduce_max(value, -1) return output
Example #4
Source File: lrp.py From THUMT with BSD 3-Clause "New" or "Revised" License | 5 votes |
def maxout_v2n(inputs, output_size, maxpart, w, params, use_bias=True, concat=True, dtype=None, scope=None): """ Maxout layer :param inputs: see the corresponding description of ``linear'' :param output_size: see the corresponding description of ``linear'' :param maxpart: an integer, the default value is 2 :param use_bias: a boolean value indicate whether to use bias term :param concat: concat all tensors if inputs is a list of tensors :param dtype: an optional instance of tf.Dtype :param scope: the scope of this layer, the default value is ``maxout'' :returns: a Tensor with shape [batch, output_size] :raises RuntimeError: see the corresponding description of ``linear'' """ w_x_dec, w_x_ctx = w w_x_dec = tf.transpose(w_x_dec, [1, 2, 0, 3]) w_x_ctx = tf.transpose(w_x_ctx, [1, 2, 0, 3]) w_x_y = tf.zeros(tf.shape(w_x_dec), dtype=tf.float32) candidate_linear = linear_v2n(inputs, output_size * maxpart, use_bias, [w_x_y, w_x_dec, w_x_ctx], params, concat, dtype=dtype, scope=scope or "maxout") candidate = candidate_linear["output"] _, w_x_dec_readout, w_x_ctx_readout = candidate_linear["weight_ratios"] w_x_readout = w_x_dec_readout + w_x_ctx_readout w_x_readout = tf.transpose(w_x_readout, [0, 2, 1, 3]) output_maxout = maxpool(candidate, output_size, params) output = output_maxout["output"] # direct w_readout_maxout = output_maxout["weight_ratio"] #propagate propagater = tf.matmul w_x_maxout = propagater(w_x_readout, w_readout_maxout) weight_ratios = [w_x_maxout] return {"output": output, "weight_ratios": weight_ratios}
Example #5
Source File: swaption.py From tf-quant-finance with Apache License 2.0 | 5 votes |
def __init__(self, swap, expiry_date=None, dtype=None, name=None): """Initialize a batch of European swaptions. Args: swap: An instance of `InterestRateSwap` specifying the interest rate swaps underlying the swaptions. The batch size of the swaptions being created would be the same as the batch size of the `swap`. expiry_date: An optional rank 1 `DateTensor` specifying the expiry dates for each swaption. The shape of the input should be the same as the batch size of the `swap` input. Default value: None in which case the option expity date is the same as the start date of each underlying swap. dtype: `tf.Dtype`. If supplied the dtype for the real variables or ops either supplied to the Swaption object or created by the Swaption object. Default value: None which maps to the default dtype inferred by TensorFlow. name: Python str. The name to give to the ops created by this class. Default value: `None` which maps to 'swaption'. """ self._name = name or 'swaption' with tf.name_scope(self._name): self._dtype = dtype self._expiry_date = dates.convert_to_date_tensor(expiry_date) self._swap = swap
Example #6
Source File: cms_swap.py From tf-quant-finance with Apache License 2.0 | 5 votes |
def __init__(self, start_date, end_date, coupon_spec, dtype=None, name=None): """Initialize a batch of CMS cashflow streams. Args: start_date: A rank 1 `DateTensor` specifying the starting dates of the accrual of the first coupon of the cashflow stream. The shape of the input correspond to the numbercof streams being created. end_date: A rank 1 `DateTensor` specifying the end dates for accrual of the last coupon in each cashflow stream. The shape of the input should be the same as that of `start_date`. coupon_spec: A list of `CMSCouponSpecs` specifying the details of the coupon payment for the cashflow stream. The length of the list should be the same as the number of streams being created. Each coupon within the list must have the same daycount_convention and businessday_rule. dtype: `tf.Dtype`. If supplied the dtype for the real variables or ops either supplied to the FloatingCashflowStream object or created by the object. Default value: None which maps to the default dtype inferred by TensorFlow. name: Python str. The name to give to the ops created by this class. Default value: `None` which maps to 'floating_cashflow_stream'. """ super(CMSCashflowStream, self).__init__() self._name = name or 'cms_cashflow_stream' with tf.name_scope(self._name): self._start_date = dates.convert_to_date_tensor(start_date) self._end_date = dates.convert_to_date_tensor(end_date) self._batch_size = self._start_date.shape[0] self._first_coupon_date = None self._penultimate_coupon_date = None self._dtype = dtype self._setup(coupon_spec)
Example #7
Source File: test_model.py From TGAN with MIT License | 5 votes |
def check_operation_nodes(graph, name, node_type, dtype, shape, consumers): """Test a graph node parameters. Args: graph(tf): Graph object the node belongs to. name(str): Name of the node. node_type(str): Operation type of the node. dtype(tf.Dtype): Dtype of the output tensor. shape(tuple[int]): Shape of the output tensor. consumers(list[str]): List of names of nodes consuming the node's output. Returns: None. Raises: AssertionError: If any check fail. """ operation = graph.get_operation_by_name(name) assert len(operation.outputs) == 1 output = operation.outputs[0] assert operation.type == node_type assert output.dtype == dtype assert output.shape.as_list() == shape assert output.consumers() == [graph.get_operation_by_name(cons) for cons in consumers]
Example #8
Source File: neighbor_features.py From neural-structured-learning with Apache License 2.0 | 4 votes |
def make_missing_neighbor_inputs(neighbor_config, inputs, weight_dtype=tf.float32): """Makes additional inputs for neighbor features if necessary. Args: neighbor_config: An instance of `configs.GraphNeighborConfig` specifying the number of neighbors and how neighbor features should be named. inputs: Dictionary of input tensors that may be missing neighbor features. The keys are the features names. See `utils.unpack_neighbor_features` for expected names of neighbor features and weights. weight_dtype: `tf.Dtype` for neighbors weights. Defaults to `tf.float32`. Returns: A dictionary of neighbor feature and weight tensors that do not already exist in `inputs`. The keys are specified according to `neighbor_config`. """ existing_feature_names = set(inputs.keys()) neighbor_inputs = {} for i in range(neighbor_config.max_neighbors): # For each potential neighbor. # Weight of the neighbor. weight_name = '{}{}{}'.format(neighbor_config.prefix, i, neighbor_config.weight_suffix) if weight_name not in existing_feature_names: neighbor_inputs[weight_name] = tf.keras.Input((1,), dtype=weight_dtype, name=weight_name) # For inputs without existing neighbor features, replicate them. for feature_name, tensor in inputs.items(): if feature_name.startswith(neighbor_config.prefix): continue neighbor_feature_name = '{}{}_{}'.format(neighbor_config.prefix, i, feature_name) if neighbor_feature_name not in existing_feature_names: neighbor_inputs[neighbor_feature_name] = tf.keras.Input( tensor.shape[1:], batch_size=tensor.shape[0], dtype=tensor.dtype, name=neighbor_feature_name, ragged=isinstance(tensor, tf.RaggedTensor), sparse=isinstance(tensor, tf.sparse.SparseTensor)) return neighbor_inputs
Example #9
Source File: interest_rate_swap.py From tf-quant-finance with Apache License 2.0 | 4 votes |
def __init__(self, start_date, maturity_date, pay_leg, receive_leg, holiday_calendar=None, dtype=None, name=None): """Initialize a batch of IRS contracts. Args: start_date: A rank 1 `DateTensor` specifying the dates for the inception (start of the accrual) of the swap contracts. The shape of the input correspond to the number of instruments being created. maturity_date: A rank 1 `DateTensor` specifying the maturity dates for each contract. The shape of the input should be the same as that of `start_date`. pay_leg: A scalar or a list of either `FixedCouponSpecs` or `FloatCouponSpecs` specifying the coupon payments for the payment leg of the swap. If specified as a list then the length of the list should be the same as the number of instruments being created. If specified as a scalar, then the elements of the namedtuple must be of the same shape as (or compatible to) the shape of `start_date`. receive_leg: A scalar or a list of either `FixedCouponSpecs` or `FloatCouponSpecs` specifying the coupon payments for the receiving leg of the swap. If specified as a list then the length of the list should be the same as the number of instruments being created. If specified as a scalar, then the elements of the namedtuple must be of the same shape as (or compatible with) the shape of `start_date`. holiday_calendar: An instance of `dates.HolidayCalendar` to specify weekends and holidays. Default value: None in which case a holiday calendar would be created with Saturday and Sunday being the holidays. dtype: `tf.Dtype`. If supplied the dtype for the real variables or ops either supplied to the IRS object or created by the IRS object. Default value: None which maps to the default dtype inferred by TensorFlow. name: Python str. The name to give to the ops created by this class. Default value: `None` which maps to 'interest_rate_swap'. """ self._name = name or 'interest_rate_swap' if holiday_calendar is None: holiday_calendar = dates.create_holiday_calendar( weekend_mask=dates.WeekendMask.SATURDAY_SUNDAY) with tf.name_scope(self._name): self._dtype = dtype self._start_date = dates.convert_to_date_tensor(start_date) self._maturity_date = dates.convert_to_date_tensor(maturity_date) self._holiday_calendar = holiday_calendar self._floating_leg = None self._fixed_leg = None self._pay_leg = self._setup_leg(pay_leg) self._receive_leg = self._setup_leg(receive_leg) self._is_payer = isinstance(self._pay_leg, cs.FixedCashflowStream)
Example #10
Source File: cms_swap.py From tf-quant-finance with Apache License 2.0 | 4 votes |
def __init__(self, start_date, maturity_date, pay_leg, receive_leg, holiday_calendar=None, dtype=None, name=None): """Initialize a batch of CMS swap contracts. Args: start_date: A rank 1 `DateTensor` specifying the dates for the inception (start of the accrual) of the swap cpntracts. The shape of the input correspond to the numbercof instruments being created. maturity_date: A rank 1 `DateTensor` specifying the maturity dates for each contract. The shape of the input should be the same as that of `start_date`. pay_leg: A list of either `FixedCouponSpecs`, `FloatCouponSpecs` or `CMSCouponSpecs` specifying the coupon payments for the payment leg of the swap. The length of the list should be the same as the number of instruments being created. receive_leg: A list of either `FixedCouponSpecs` or `FloatCouponSpecs` or `CMSCouponSpecs` specifying the coupon payments for the receiving leg of the swap. The length of the list should be the same as the number of instruments being created. holiday_calendar: An instance of `dates.HolidayCalendar` to specify weekends and holidays. Default value: None in which case a holiday calendar would be created with Saturday and Sunday being the holidays. dtype: `tf.Dtype`. If supplied the dtype for the real variables or ops either supplied to the IRS object or created by the IRS object. Default value: None which maps to the default dtype inferred by TensorFlow. name: Python str. The name to give to the ops created by this class. Default value: `None` which maps to 'cms_swap'. """ self._name = name or 'cms_swap' if holiday_calendar is None: holiday_calendar = dates.create_holiday_calendar( weekend_mask=dates.WeekendMask.SATURDAY_SUNDAY) with tf.name_scope(self._name): self._dtype = dtype self._start_date = dates.convert_to_date_tensor(start_date) self._maturity_date = dates.convert_to_date_tensor(maturity_date) self._holiday_calendar = holiday_calendar self._floating_leg = None self._fixed_leg = None self._cms_leg = None self._pay_leg = self._setup_leg(pay_leg) self._receive_leg = self._setup_leg(receive_leg)
Example #11
Source File: bond.py From tf-quant-finance with Apache License 2.0 | 4 votes |
def __init__(self, settlement_date, maturity_date, coupon_spec, start_date=None, first_coupon_date=None, penultimate_coupon_date=None, holiday_calendar=None, dtype=None, name=None): """Initialize a batch of fixed coupon bonds. Args: settlement_date: A rank 1 `DateTensor` specifying the settlement date of the bonds. maturity_date: A rank 1 `DateTensor` specifying the maturity dates of the bonds. The shape of the input should be the same as that of `settlement_date`. coupon_spec: A list of `FixedCouponSpecs` specifying the coupon payments. The length of the list should be the same as the number of bonds being created. start_date: An optional `DateTensor` specifying the dates when the interest starts to accrue for the coupons. The input can be used to specify a forward start date for the coupons. The shape of the input correspond to the numbercof instruments being created. Default value: None in which case the coupons start to accrue from the `settlement_date`. first_coupon_date: An optional rank 1 `DateTensor` specifying the dates when first coupon will be paid for bonds with irregular first coupon. penultimate_coupon_date: An optional rank 1 `DateTensor` specifying the dates when the penultimate coupon (or last regular coupon) will be paid for bonds with irregular last coupon. holiday_calendar: An instance of `dates.HolidayCalendar` to specify weekends and holidays. Default value: None in which case a holiday calendar would be created with Saturday and Sunday being the holidays. dtype: `tf.Dtype`. If supplied the dtype for the real variables or ops either supplied to the bond object or created by the bond object. Default value: None which maps to the default dtype inferred by TensorFlow. name: Python str. The name to give to the ops created by this class. Default value: `None` which maps to 'bond'. """ self._name = name or 'bond' if holiday_calendar is None: holiday_calendar = dates.create_holiday_calendar( weekend_mask=dates.WeekendMask.SATURDAY_SUNDAY) with tf.name_scope(self._name): self._dtype = dtype self._settlement_date = dates.convert_to_date_tensor(settlement_date) self._maturity_date = dates.convert_to_date_tensor(maturity_date) self._holiday_calendar = holiday_calendar self._setup(coupon_spec, start_date, first_coupon_date, penultimate_coupon_date)
Example #12
Source File: result_types.py From fold with Apache License 2.0 | 4 votes |
def convert_to_type(type_like): """Converts `type_like` to a `Type`. If `type_like` is already a `Type`, it is returned. The following conversions are performed: * Python tuples become `Tuple`s; items are recursively converted. * A `tf.TensorShape` becomes a corresponding `TensorType` with `dtype=float32`. Must be fully defined. * Lists of `shape + [dtype]` (e.g. `[3, 4, 'int32']`) become `TensorType`s, with the default `dtype=float32` if omitted. * A `tf.Dtype` or stringified version thereof (e.g. `'int64'`) becomes a corresponding scalar `TensorType((), dtype)`. * An integer `vector_len` becomes a corresponding vector `TensorType((vector_len,), dtype=float32)`. Args: type_like: Described above. Returns: A `Type`. Raises: TypeError: If `type_like` cannot be converted to a `Type`. """ if isinstance(type_like, ResultType): return type_like if isinstance(type_like, tf.TensorShape): # Check this *before* calling as_list() otherwise it throws. if not type_like.is_fully_defined(): raise TypeError('shape %s is not fully defined' % type_like) return TensorType(type_like.as_list()) if isinstance(type_like, tuple): return TupleType(convert_to_type(item) for item in type_like) if isinstance(type_like, list): if type_like and isinstance(type_like[-1], six.string_types): return TensorType(type_like[:-1], dtype=type_like[-1]) else: return TensorType(type_like) if isinstance(type_like, tf.DType) or isinstance(type_like, six.string_types): return TensorType((), dtype=type_like) if isinstance(type_like, numbers.Integral): return TensorType((type_like,)) raise TypeError('Cannot covert %s to a type.' % (type_like,))