def _next_internal(self):
autograph_status = autograph_ctx.control_status_ctx().status
autograph_disabled = autograph_status == autograph_ctx.Status.DISABLED
if not context.executing_eagerly() and autograph_disabled:
self._get_next_call_count += 1
if self._get_next_call_count > GET_NEXT_CALL_ERROR_THRESHOLD:
raise ValueError(GET_NEXT_CALL_ERROR_MESSAGE)
if not context.executing_eagerly():
# TODO(b/169442955): Investigate the need for this colocation constraint.
with ops.colocate_with(self._iterator_resource):
ret = gen_dataset_ops.iterator_get_next(
self._iterator_resource,
output_types=self._flat_output_types,
output_shapes=self._flat_output_shapes)
return structure.from_compatible_tensor_list(
self._element_spec, ret)
# TODO(b/77291417): This runs in sync mode as iterators use an error status
# to communicate that there is no more data to iterate over.
with context.execution_mode(context.SYNC):
ret = gen_dataset_ops.iterator_get_next(
self._iterator_resource,
output_types=self._flat_output_types,
output_shapes=self._flat_output_shapes)
try:
# Fast path for the case `self._structure` is not a nested structure.
return self._element_spec._from_compatible_tensor_list(ret) # pylint: disable=protected-access
except AttributeError:
return structure.from_compatible_tensor_list(
self._element_spec, ret)
def _next_internal(self):
if not context.executing_eagerly():
# TODO(b/169442955): Investigate the need for this colocation constraint.
with ops.colocate_with(self._iterator_resource):
ret = gen_dataset_ops.iterator_get_next(
self._iterator_resource,
output_types=self._flat_output_types,
output_shapes=self._flat_output_shapes)
return structure.from_compatible_tensor_list(
self._element_spec, ret)
# TODO(b/77291417): This runs in sync mode as iterators use an error status
# to communicate that there is no more data to iterate over.
with context.execution_mode(context.SYNC):
ret = gen_dataset_ops.iterator_get_next(
self._iterator_resource,
output_types=self._flat_output_types,
output_shapes=self._flat_output_shapes)
try:
# Fast path for the case `self._structure` is not a nested structure.
return self._element_spec._from_compatible_tensor_list(ret) # pylint: disable=protected-access
except AttributeError:
return structure.from_compatible_tensor_list(
self._element_spec, ret)
def _next_internal(self):
"""Returns a nested structure of `tf.Tensor`s containing the next element.
"""
if not context.executing_eagerly():
with ops.device(self._device):
ret = gen_dataset_ops.iterator_get_next(
self._iterator_resource,
output_types=self._flat_output_types,
output_shapes=self._flat_output_shapes)
return structure.from_compatible_tensor_list(
self._element_spec, ret)
# This runs in sync mode as iterators use an error status to communicate
# that there is no more data to iterate over.
# TODO(b/77291417): Fix
with context.execution_mode(context.SYNC):
with ops.device(self._device):
# TODO(ashankar): Consider removing this ops.device() contextmanager
# and instead mimic ops placement in graphs: Operations on resource
# handles execute on the same device as where the resource is placed.
ret = gen_dataset_ops.iterator_get_next(
self._iterator_resource,
output_types=self._flat_output_types,
output_shapes=self._flat_output_shapes)
try:
# Fast path for the case `self._structure` is not a nested structure.
return self._element_spec._from_compatible_tensor_list(ret) # pylint: disable=protected-access
except AttributeError:
return structure.from_compatible_tensor_list(
self._element_spec, ret)
def _next_internal(self):
"""Returns a nested structure of `tf.Tensor`s containing the next element.
"""
if not context.executing_eagerly():
with ops.device(self._device):
ret = gen_dataset_ops.iterator_get_next(
self._iterator_resource,
output_types=self._flat_output_types,
output_shapes=self._flat_output_shapes)
return structure.from_compatible_tensor_list(self._element_spec, ret)
# This runs in sync mode as iterators use an error status to communicate
# that there is no more data to iterate over.
# TODO(b/77291417): Fix
with context.execution_mode(context.SYNC):
with ops.device(self._device):
# TODO(ashankar): Consider removing this ops.device() contextmanager
# and instead mimic ops placement in graphs: Operations on resource
# handles execute on the same device as where the resource is placed.
# NOTE(mrry): Here we use the "_sync" variant of `iterator_get_next`
# because in eager mode this code will run synchronously on the calling
# thread. Therefore we do not need to make a defensive context switch
# to a background thread, and can achieve a small constant performance
# boost by invoking the iterator synchronously.
ret = gen_dataset_ops.iterator_get_next_sync(
self._iterator_resource,
output_types=self._flat_output_types,
output_shapes=self._flat_output_shapes)
try:
# Fast path for the case `self._structure` is not a nested structure.
return self._element_spec._from_compatible_tensor_list(ret) # pylint: disable=protected-access
except AttributeError:
return structure.from_compatible_tensor_list(self._element_spec, ret)
def get_next(self, name=None):
"""Returns a nested structure of `tf.Tensor`s containing the next element.
Args:
name: (Optional.) A name for the created operation.
Returns:
A nested structure of `tf.Tensor` objects.
"""
self._get_next_call_count += 1
if self._get_next_call_count > GET_NEXT_CALL_WARNING_THRESHOLD:
warnings.warn(GET_NEXT_CALL_WARNING_MESSAGE)
return sparse.deserialize_sparse_tensors(
nest.pack_sequence_as(
self._output_types,
gen_dataset_ops.iterator_get_next(
self._iterator_resource,
output_types=nest.flatten(
sparse.as_dense_types(self._output_types,
self._output_classes)),
output_shapes=nest.flatten(
sparse.as_dense_shapes(self._output_shapes,
self._output_classes)),
name=name)), self._output_types, self._output_shapes,
self._output_classes)
def get_next(self, name=None):
"""Returns a nested structure of `tf.Tensor`s containing the next element.
Args:
name: (Optional.) A name for the created operation.
Returns:
A nested structure of `tf.Tensor` objects.
"""
self._get_next_call_count += 1
if self._get_next_call_count > GET_NEXT_CALL_WARNING_THRESHOLD:
warnings.warn(GET_NEXT_CALL_WARNING_MESSAGE)
return sparse.deserialize_sparse_tensors(
nest.pack_sequence_as(self._output_types,
gen_dataset_ops.iterator_get_next(
self._iterator_resource,
output_types=nest.flatten(
sparse.as_dense_types(
self._output_types,
self._output_classes)),
output_shapes=nest.flatten(
sparse.as_dense_shapes(
self._output_shapes,
self._output_classes)),
name=name)), self._output_types,
self._output_shapes, self._output_classes)
def _next_internal(self):
"""Returns a nested structure of `tf.Tensor`s containing the next element.
"""
if not context.executing_eagerly():
with ops.device(self._device):
ret = gen_dataset_ops.iterator_get_next(
self._iterator_resource,
output_types=self._flat_output_types,
output_shapes=self._flat_output_shapes)
return self._structure._from_compatible_tensor_list(ret) # pylint: disable=protected-access
# This runs in sync mode as iterators use an error status to communicate
# that there is no more data to iterate over.
# TODO(b/77291417): Fix
with context.execution_mode(context.SYNC):
with ops.device(self._device):
# TODO(ashankar): Consider removing this ops.device() contextmanager
# and instead mimic ops placement in graphs: Operations on resource
# handles execute on the same device as where the resource is placed.
# NOTE(mrry): Here we use the "_sync" variant of `iterator_get_next`
# because in eager mode this code will run synchronously on the calling
# thread. Therefore we do not need to make a defensive context switch
# to a background thread, and can achieve a small constant performance
# boost by invoking the iterator synchronously.
ret = gen_dataset_ops.iterator_get_next_sync(
self._iterator_resource,
output_types=self._flat_output_types,
output_shapes=self._flat_output_shapes)
return self._structure._from_compatible_tensor_list(ret) # pylint: disable=protected-access
def get_next(self, name=None):
"""Returns the next element.
In graph mode, you should typically call this method *once* and use its
result as the input to another computation. A typical loop will then call
`tf.Session.run` on the result of that computation. The loop will terminate
when the `Iterator.get_next()` operation raises
`tf.errors.OutOfRangeError`. The following skeleton shows how to use
this method when building a training loop:
```python
dataset = ... # A `tf.data.Dataset` object.
iterator = dataset.make_initializable_iterator()
next_element = iterator.get_next()
# Build a TensorFlow graph that does something with each element.
loss = model_function(next_element)
optimizer = ... # A `tf.compat.v1.train.Optimizer` object.
train_op = optimizer.minimize(loss)
with tf.compat.v1.Session() as sess:
try:
while True:
sess.run(train_op)
except tf.errors.OutOfRangeError:
pass
```
NOTE: It is legitimate to call `Iterator.get_next()` multiple times, e.g.
when you are distributing different elements to multiple devices in a single
step. However, a common pitfall arises when users call `Iterator.get_next()`
in each iteration of their training loop. `Iterator.get_next()` adds ops to
the graph, and executing each op allocates resources (including threads); as
a consequence, invoking it in every iteration of a training loop causes
slowdown and eventual resource exhaustion. To guard against this outcome, we
log a warning when the number of uses crosses a fixed threshold of
suspiciousness.
Args:
name: (Optional.) A name for the created operation.
Returns:
A (nested) structure of values matching `tf.data.Iterator.element_spec`.
"""
self._get_next_call_count += 1
if self._get_next_call_count > GET_NEXT_CALL_WARNING_THRESHOLD:
warnings.warn(GET_NEXT_CALL_WARNING_MESSAGE)
# TODO(b/169442955): Investigate the need for this colocation constraint.
with ops.colocate_with(self._iterator_resource):
# pylint: disable=protected-access
flat_ret = gen_dataset_ops.iterator_get_next(
self._iterator_resource,
output_types=self._flat_tensor_types,
output_shapes=self._flat_tensor_shapes,
name=name)
return structure.from_tensor_list(self._element_spec, flat_ret)
def next(self):
"""Return the next tf.Tensor from the dataset."""
try:
ret = gen_dataset_ops.iterator_get_next(
self._resource,
output_types=self._flat_output_types,
output_shapes=self._flat_output_shapes)
return nest.pack_sequence_as(self._output_types, ret)
except errors.OutOfRangeError:
raise StopIteration
def next(self):
"""Return the next tf.Tensor from the dataset."""
try:
ret = gen_dataset_ops.iterator_get_next(
self._resource,
output_types=self._flat_output_types,
output_shapes=self._flat_output_shapes)
return nest.pack_sequence_as(self._output_types, ret)
except errors.OutOfRangeError:
raise StopIteration
def get_next(self, name=None):
"""Returns a nested structure of `tf.Tensor`s representing the next element.
In graph mode, you should typically call this method *once* and use its
result as the input to another computation. A typical loop will then call
`tf.Session.run` on the result of that computation. The loop will terminate
when the `Iterator.get_next()` operation raises
`tf.errors.OutOfRangeError`. The following skeleton shows how to use
this method when building a training loop:
```python
dataset = ... # A `tf.data.Dataset` object.
iterator = dataset.make_initializable_iterator()
next_element = iterator.get_next()
# Build a TensorFlow graph that does something with each element.
loss = model_function(next_element)
optimizer = ... # A `tf.train.Optimizer` object.
train_op = optimizer.minimize(loss)
with tf.Session() as sess:
try:
while True:
sess.run(train_op)
except tf.errors.OutOfRangeError:
pass
```
NOTE: It is legitimate to call `Iterator.get_next()` multiple times, e.g.
when you are distributing different elements to multiple devices in a single
step. However, a common pitfall arises when users call `Iterator.get_next()`
in each iteration of their training loop. `Iterator.get_next()` adds ops to
the graph, and executing each op allocates resources (including threads); as
a consequence, invoking it in every iteration of a training loop causes
slowdown and eventual resource exhaustion. To guard against this outcome, we
log a warning when the number of uses crosses a fixed threshold of
suspiciousness.
Args:
name: (Optional.) A name for the created operation.
Returns:
A nested structure of `tf.Tensor` objects.
"""
self._get_next_call_count += 1
if self._get_next_call_count > GET_NEXT_CALL_WARNING_THRESHOLD:
warnings.warn(GET_NEXT_CALL_WARNING_MESSAGE)
# pylint: disable=protected-access
flat_ret = gen_dataset_ops.iterator_get_next(
self._iterator_resource,
output_types=self._structure._flat_types,
output_shapes=self._structure._flat_shapes, name=name)
return self._structure._from_tensor_list(flat_ret)
def next(self):
"""Return the next tf.Tensor from the dataset."""
try:
# TODO(ashankar): Consider removing this ops.device() contextmanager
# and instead mimic ops placement in graphs: Operations on resource
# handles execute on the same device as where the resource is placed.
with ops.device("/device:CPU:0"):
ret = gen_dataset_ops.iterator_get_next(
self._resource,
output_types=self._flat_output_types,
output_shapes=self._flat_output_shapes)
return nest.pack_sequence_as(self._output_types, ret)
except errors.OutOfRangeError:
raise StopIteration
def next(self):
"""Return the next tf.Tensor from the dataset."""
try:
# TODO(ashankar): Consider removing this ops.device() contextmanager
# and instead mimic ops placement in graphs: Operations on resource
# handles execute on the same device as where the resource is placed.
with ops.device("/device:CPU:0"):
ret = gen_dataset_ops.iterator_get_next(
self._resource,
output_types=self._flat_output_types,
output_shapes=self._flat_output_shapes)
return nest.pack_sequence_as(self._output_types, ret)
except errors.OutOfRangeError:
raise StopIteration
def get_next(self, name=None):
"""Returns a nested structure of `tf.Tensor`s containing the next element.
Args:
name: (Optional.) A name for the created operation.
Returns:
A nested structure of `tf.Tensor` objects.
"""
return nest.pack_sequence_as(
self._output_types,
gen_dataset_ops.iterator_get_next(
self._iterator_resource,
output_types=nest.flatten(self._output_types),
output_shapes=nest.flatten(self._output_shapes),
name=name))
def get_next(self, name=None):
"""Returns a nested structure of `tf.Tensor`s containing the next element.
Args:
name: (Optional.) A name for the created operation.
Returns:
A nested structure of `tf.Tensor` objects.
"""
return nest.pack_sequence_as(
self._output_types,
gen_dataset_ops.iterator_get_next(
self._iterator_resource,
output_types=nest.flatten(self._output_types),
output_shapes=nest.flatten(self._output_shapes),
name=name))
def next(self):
"""Return the next tf.Tensor from the dataset."""
with ops.device(self._device):
try:
if self._buffer_resource_handle is not None:
ret = prefetching_ops.function_buffering_resource_get_next(
function_buffer_resource=self._buffer_resource_handle,
output_types=self._flat_output_types)
else:
# TODO (ashankar): Consider removing this ops.device() contextmanager id:607 gh:608
# and instead mimic ops placement in graphs: Operations on resource
# handles execute on the same device as where the resource is placed.
ret = gen_dataset_ops.iterator_get_next(
self._resource,
output_types=self._flat_output_types,
output_shapes=self._flat_output_shapes)
except errors.OutOfRangeError:
raise StopIteration
return nest.pack_sequence_as(self._output_types, ret)
def next(self):
"""Return the next tf.Tensor from the dataset."""
try:
# TODO(ashankar): Consider removing this ops.device() contextmanager
# and instead mimic ops placement in graphs: Operations on resource
# handles execute on the same device as where the resource is placed.
with ops.device("/device:CPU:0"):
ret = gen_dataset_ops.iterator_get_next(
self._resource,
output_types=self._flat_output_types,
output_shapes=self._flat_output_shapes)
except errors.OutOfRangeError:
raise StopIteration
# Copies tensors from CPU to the current device if necessary.
# TODO(rohanj): This should be replaced by the mechanism to have the
# runtime's threads copy tensors to the destination device.
with ops.device(self._device):
ret = [array_ops.identity(x) for x in ret]
return nest.pack_sequence_as(self._output_types, ret)
def next(self):
"""Return the next tf.Tensor from the dataset."""
with ops.device(self._device):
try:
if self._buffer_resource_handle is not None:
ret = prefetching_ops.function_buffering_resource_get_next(
function_buffer_resource=self._buffer_resource_handle,
output_types=self._flat_output_types)
else:
# TODO(ashankar): Consider removing this ops.device() contextmanager
# and instead mimic ops placement in graphs: Operations on resource
# handles execute on the same device as where the resource is placed.
ret = gen_dataset_ops.iterator_get_next(
self._resource,
output_types=self._flat_output_types,
output_shapes=self._flat_output_shapes)
except errors.OutOfRangeError:
raise StopIteration
return nest.pack_sequence_as(self._output_types, ret)
def _next_internal(self):
"""Returns a nested structure of `tf.Tensor`s containing the next element.
"""
with ops.device(self._device):
if self._buffer_resource_handle is not None:
ret = prefetching_ops.function_buffering_resource_get_next(
function_buffer_resource=self._buffer_resource_handle,
output_types=self._flat_output_types)
else:
# TODO(ashankar): Consider removing this ops.device() contextmanager
# and instead mimic ops placement in graphs: Operations on resource
# handles execute on the same device as where the resource is placed.
ret = gen_dataset_ops.iterator_get_next(
self._resource,
output_types=self._flat_output_types,
output_shapes=self._flat_output_shapes)
return sparse.deserialize_sparse_tensors(
nest.pack_sequence_as(self._output_types, ret), self._output_types,
self._output_shapes, self._output_classes)
def _next_internal(self):
"""Returns a nested structure of `tf.Tensor`s containing the next element.
"""
with ops.device(self._device):
if self._buffer_resource_handle is not None:
ret = prefetching_ops.function_buffering_resource_get_next(
function_buffer_resource=self._buffer_resource_handle,
output_types=self._flat_output_types)
else:
# TODO(ashankar): Consider removing this ops.device() contextmanager
# and instead mimic ops placement in graphs: Operations on resource
# handles execute on the same device as where the resource is placed.
ret = gen_dataset_ops.iterator_get_next(
self._resource,
output_types=self._flat_output_types,
output_shapes=self._flat_output_shapes)
return sparse.deserialize_sparse_tensors(
nest.pack_sequence_as(self._output_types, ret), self._output_types,
self._output_shapes, self._output_classes)
def get_next(self, name=None):
"""Returns a nested structure of `tf.Tensor`s representing the next element.
In graph mode, you should typically call this method *once* and use its
result as the input to another computation. A typical loop will then call
@{tf.Session.run} on the result of that computation. The loop will terminate
when the `Iterator.get_next()` operation raises
@{tf.errors.OutOfRangeError}. The following skeleton shows how to use
this method when building a training loop:
```python
dataset = ... # A `tf.data.Dataset` object.
iterator = dataset.make_initializable_iterator()
next_element = iterator.get_next()
# Build a TensorFlow graph that does something with each element.
loss = model_function(next_element)
optimizer = ... # A `tf.train.Optimizer` object.
train_op = optimizer.minimize(loss)
with tf.Session() as sess:
try:
while True:
sess.run(train_op)
except tf.errors.OutOfRangeError:
pass
```
NOTE: It is legitimate to call `Iterator.get_next()` multiple times, e.g.
when you are distributing different elements to multiple devices in a single
step. However, a common pitfall arises when users call `Iterator.get_next()`
in each iteration of their training loop. `Iterator.get_next()` adds ops to
the graph, and executing each op allocates resources (including threads); as
a consequence, invoking it in every iteration of a training loop causes
slowdown and eventual resource exhaustion. To guard against this outcome, we
log a warning when the number of uses crosses a fixed threshold of
suspiciousness.
Args:
name: (Optional.) A name for the created operation.
Returns:
A nested structure of `tf.Tensor` objects.
"""
self._get_next_call_count += 1
if self._get_next_call_count > GET_NEXT_CALL_WARNING_THRESHOLD:
warnings.warn(GET_NEXT_CALL_WARNING_MESSAGE)
return sparse.deserialize_sparse_tensors(
nest.pack_sequence_as(self._output_types,
gen_dataset_ops.iterator_get_next(
self._iterator_resource,
output_types=nest.flatten(
sparse.as_dense_types(
self._output_types,
self._output_classes)),
output_shapes=nest.flatten(
sparse.as_dense_shapes(
self._output_shapes,
self._output_classes)),
name=name)), self._output_types,
self._output_shapes, self._output_classes)
