text stringlengths 81 112k |
|---|
Add a new item to the Reservoir with the given tag.
If the reservoir has not yet reached full size, the new item is guaranteed
to be added. If the reservoir is full, then behavior depends on the
always_keep_last boolean.
If always_keep_last was set to true, the new item is guaranteed to be added
t... |
Filter items within a Reservoir, using a filtering function.
Args:
filterFn: A function that returns True for the items to be kept.
key: An optional bucket key to filter. If not specified, will filter all
all buckets.
Returns:
The number of items removed.
def FilterItems(self, filte... |
Add an item to the ReservoirBucket, replacing an old item if necessary.
The new item is guaranteed to be added to the bucket, and to be the last
element in the bucket. If the bucket has reached capacity, then an old item
will be replaced. With probability (_max_size/_num_items_seen) a random item
in th... |
Filter items in a ReservoirBucket, using a filtering function.
Filtering items from the reservoir bucket must update the
internal state variable self._num_items_seen, which is used for determining
the rate of replacement in reservoir sampling. Ideally, self._num_items_seen
would contain the exact numbe... |
Returns the inferred dense dimensions of a list of lists.
def _GetDenseDimensions(list_of_lists):
"""Returns the inferred dense dimensions of a list of lists."""
if not isinstance(list_of_lists, (list, tuple)):
return []
elif not list_of_lists:
return [0]
else:
return [len(list_... |
Create a TensorProto.
Args:
values: Values to put in the TensorProto.
dtype: Optional tensor_pb2 DataType value.
shape: List of integers representing the dimensions of tensor.
verify_shape: Boolean that enables verification of a shape of values.
Returns:
A `TensorPr... |
Create a numpy ndarray from a tensor.
Create a numpy ndarray with the same shape and data as the tensor.
Args:
tensor: A TensorProto.
Returns:
A numpy array with the tensor contents.
Raises:
TypeError: if tensor has unsupported type.
def make_ndarray(tensor):
"""Create a numpy ndarray from ... |
Creates a summary that contains a layout.
When users navigate to the custom scalars dashboard, they will see a layout
based on the proto provided to this function.
Args:
scalars_layout: The scalars_layout_pb2.Layout proto that specifies the
layout.
collections: Optional list of graph collections... |
Creates a summary that contains a layout.
When users navigate to the custom scalars dashboard, they will see a layout
based on the proto provided to this function.
Args:
scalars_layout: The scalars_layout_pb2.Layout proto that specifies the
layout.
Returns:
A summary proto containing the layo... |
Returns true if `other` is convertible with this Dimension.
Two known Dimensions are convertible if they have the same value.
An unknown Dimension is convertible with all other Dimensions.
Args:
other: Another Dimension.
Returns:
True if this Dimension and `other` ... |
Returns a Dimension that combines the information in `self` and `other`.
Dimensions are combined as follows:
```python
tf.Dimension(n) .merge_with(tf.Dimension(n)) == tf.Dimension(n)
tf.Dimension(n) .merge_with(tf.Dimension(None)) == tf.Dimension(n)
tf.Dimension(None).me... |
Returns the rank of this shape, or None if it is unspecified.
def ndims(self):
"""Returns the rank of this shape, or None if it is unspecified."""
if self._dims is None:
return None
else:
if self._ndims is None:
self._ndims = len(self._dims)
r... |
Returns the total number of elements, or none for incomplete shapes.
def num_elements(self):
"""Returns the total number of elements, or none for incomplete shapes."""
if self.is_fully_defined():
size = 1
for dim in self._dims:
size *= dim.value
retur... |
Returns a `TensorShape` combining the information in `self` and `other`.
The dimensions in `self` and `other` are merged elementwise,
according to the rules defined for `Dimension.merge_with()`.
Args:
other: Another `TensorShape`.
Returns:
A `TensorShape` containin... |
Returns the concatenation of the dimension in `self` and `other`.
*N.B.* If either `self` or `other` is completely unknown,
concatenation will discard information about the other shape. In
future, we might support concatenation that preserves this
information for use with slicing.
... |
Raises an exception if `self` and `other` do not have convertible ranks.
Args:
other: Another `TensorShape`.
Raises:
ValueError: If `self` and `other` do not represent shapes with the
same rank.
def assert_same_rank(self, other):
"""Raises an exception if `self... |
Returns a shape based on `self` with the given rank.
This method promotes a completely unknown shape to one with a
known rank.
Args:
rank: An integer.
Returns:
A shape that is at least as specific as `self` with the given rank.
Raises:
ValueError... |
Returns a shape based on `self` with at least the given rank.
Args:
rank: An integer.
Returns:
A shape that is at least as specific as `self` with at least the given
rank.
Raises:
ValueError: If `self` does not represent a shape with at least the given
... |
Returns a shape based on `self` with at most the given rank.
Args:
rank: An integer.
Returns:
A shape that is at least as specific as `self` with at most the given
rank.
Raises:
ValueError: If `self` does not represent a shape with at most the given
... |
Returns True iff `self` is convertible with `other`.
Two possibly-partially-defined shapes are convertible if there
exists a fully-defined shape that both shapes can represent. Thus,
convertibility allows the shape inference code to reason about
partially-defined shapes. For example:
... |
Returns the most specific TensorShape convertible with `self` and `other`.
* TensorShape([None, 1]) is the most specific TensorShape convertible with
both TensorShape([2, 1]) and TensorShape([5, 1]). Note that
TensorShape(None) is also convertible with above mentioned TensorShapes.
... |
Returns True iff `self` is fully defined in every dimension.
def is_fully_defined(self):
"""Returns True iff `self` is fully defined in every dimension."""
return self._dims is not None and all(
dim.value is not None for dim in self._dims
) |
Returns a list of integers or `None` for each dimension.
Returns:
A list of integers or `None` for each dimension.
Raises:
ValueError: If `self` is an unknown shape with an unknown rank.
def as_list(self):
"""Returns a list of integers or `None` for each dimension.
... |
Returns this shape as a `TensorShapeProto`.
def as_proto(self):
"""Returns this shape as a `TensorShapeProto`."""
if self._dims is None:
return tensor_shape_pb2.TensorShapeProto(unknown_rank=True)
else:
return tensor_shape_pb2.TensorShapeProto(
dim=[
... |
Converts a PredictResponse to ClassificationResponse or RegressionResponse.
Args:
pred: PredictResponse to convert.
serving_bundle: A `ServingBundle` object that contains the information about
the serving request that the response was generated by.
Returns:
A ClassificationResponse or Regression... |
Converts tensor values into ClassificationResponse or RegressionResponse.
Args:
values: For classification, a 2D list of numbers. The first dimension is for
each example being predicted. The second dimension are the probabilities
for each class ID in the prediction. For regression, a 1D list of numbe... |
Return the string message associated with TensorBoard purges.
def _GetPurgeMessage(most_recent_step, most_recent_wall_time, event_step,
event_wall_time, num_expired):
"""Return the string message associated with TensorBoard purges."""
return ('Detected out of order event.step likely caused by ... |
Returns a dict mapping tags to content specific to that plugin.
Args:
plugin_name: The name of the plugin for which to fetch plugin-specific
content.
Raises:
KeyError: if the plugin name is not found.
Returns:
A dict mapping tags to plugin-specific content (which are always stri... |
Called whenever an event is loaded.
def _ProcessEvent(self, event):
"""Called whenever an event is loaded."""
if self._first_event_timestamp is None:
self._first_event_timestamp = event.wall_time
if event.HasField('file_version'):
new_file_version = _ParseFileVersion(event.file_version)
... |
Return all tags found in the value stream.
Returns:
A `{tagType: ['list', 'of', 'tags']}` dictionary.
def Tags(self):
"""Return all tags found in the value stream.
Returns:
A `{tagType: ['list', 'of', 'tags']}` dictionary.
"""
return {
TENSORS: list(self.tensors_by_tag.keys())... |
Maybe purge orphaned data due to a TensorFlow crash.
When TensorFlow crashes at step T+O and restarts at step T, any events
written after step T are now "orphaned" and will be at best misleading if
they are included in TensorBoard.
This logic attempts to determine if there is orphaned data, and purge ... |
Check for out-of-order event.step and discard expired events for tags.
Check if the event is out of order relative to the global most recent step.
If it is, purge outdated summaries for tags that the event contains.
Args:
event: The event to use as reference. If the event is out-of-order, all
... |
Purge all events that have occurred after the given event.step.
If by_tags is True, purge all events that occurred after the given
event.step, but only for the tags that the event has. Non-sequential
event.steps suggest that a TensorFlow restart occurred, and we discard
the out-of-order events to displ... |
Returns the plugin, if possible.
Args:
context: The TBContext flags.
Returns:
A BeholderPlugin instance or None if it couldn't be loaded.
def load(self, context):
"""Returns the plugin, if possible.
Args:
context: The TBContext flags.
Returns:
A BeholderPlugin instance o... |
Walks the nested keras layer configuration in preorder.
Args:
keras_layer: Keras configuration from model.to_json.
Yields:
A tuple of (name_scope, layer_config).
name_scope: a string representing a scope name, similar to that of tf.name_scope.
layer_config: a dict representing a Keras layer configu... |
Updates input_to_in_layer, model_name_to_output, and prev_node_name
based on the model_layer.
Args:
name_scope: a string representing a scope name, similar to that of tf.name_scope.
model_layer: a dict representing a Keras model configuration.
input_to_in_layer: a dict mapping Keras.layers.Input to inb... |
Returns a GraphDef representation of the Keras model in a dict form.
Note that it only supports models that implemented to_json().
Args:
keras_layer: A dict from Keras model.to_json().
Returns:
A GraphDef representation of the layers in the model.
def keras_model_to_graph_def(keras_layer):
"""Return... |
Returns the plugin, if possible.
Args:
context: The TBContext flags.
Returns:
A HParamsPlugin instance or None if it couldn't be loaded.
def load(self, context):
"""Returns the plugin, if possible.
Args:
context: The TBContext flags.
Returns:
A HParamsPlugin instance or ... |
Returns True if the hparams plugin is active.
The hparams plugin is active iff there is a tag with
the hparams plugin name as its plugin name and the scalars plugin is
registered and active.
def is_active(self):
"""Returns True if the hparams plugin is active.
The hparams plugin is active iff the... |
Convert Markdown to HTML that's safe to splice into the DOM.
Arguments:
markdown_string: A Unicode string or UTF-8--encoded bytestring
containing Markdown source. Markdown tables are supported.
Returns:
A string containing safe HTML.
def markdown_to_safe_html(markdown_string):
"""Convert Markdown... |
Converts the given `type_value` to a `DType`.
Args:
type_value: A value that can be converted to a `tf.DType` object. This may
currently be a `tf.DType` object, a [`DataType`
enum](https://www.tensorflow.org/code/tensorflow/core/framework/types.proto),
a string type name, or a `numpy.... |
Returns the dtype correspond to this dtype's real part.
def real_dtype(self):
"""Returns the dtype correspond to this dtype's real part."""
base = self.base_dtype
if base == complex64:
return float32
elif base == complex128:
return float64
else:
... |
Returns whether this is a (non-quantized) integer type.
def is_integer(self):
"""Returns whether this is a (non-quantized) integer type."""
return (
self.is_numpy_compatible
and not self.is_quantized
and np.issubdtype(self.as_numpy_dtype, np.integer)
) |
Returns whether this is a (non-quantized, real) floating point type.
def is_floating(self):
"""Returns whether this is a (non-quantized, real) floating point type."""
return (
self.is_numpy_compatible and np.issubdtype(self.as_numpy_dtype, np.floating)
) or self.base_dtype == bfloat... |
Returns the minimum representable value in this data type.
Raises:
TypeError: if this is a non-numeric, unordered, or quantized type.
def min(self):
"""Returns the minimum representable value in this data type.
Raises:
TypeError: if this is a non-numeric, unordered, or qua... |
Return intensity limits, i.e. (min, max) tuple, of the dtype.
Args:
clip_negative : bool, optional
If True, clip the negative range (i.e. return 0 for min intensity)
even if the image dtype allows negative values.
Returns
min, max : tuple
Lower... |
Returns True if the `other` DType will be converted to this DType.
The conversion rules are as follows:
```python
DType(T) .is_compatible_with(DType(T)) == True
DType(T) .is_compatible_with(DType(T).as_ref) == True
DType(T).as_ref.is_compatible_with(DType(T))... |
Start listening on the given gRPC port.
This method of an instance of InteractiveDebuggerPlugin can be invoked at
most once. This method is not thread safe.
Args:
grpc_port: port number to listen at.
Raises:
ValueError: If this instance is already listening at a gRPC port.
def listen(sel... |
Obtains a mapping between routes and handlers.
This function also starts a debugger data server on separate thread if the
plugin has not started one yet.
Returns:
A mapping between routes and handlers (functions that respond to
requests).
def get_plugin_apps(self):
"""Obtains a mapping be... |
The audio plugin is active iff any run has at least one relevant tag.
def is_active(self):
"""The audio plugin is active iff any run has at least one relevant tag."""
if not self._multiplexer:
return False
return bool(self._multiplexer.PluginRunToTagToContent(metadata.PLUGIN_NAME)) |
Return information about the tags in each run.
Result is a dictionary of the form
{
"runName1": {
"tagName1": {
"displayName": "The first tag",
"description": "<p>Long ago there was just one tag...</p>",
"samples": 3
},
... |
Given a tag and list of runs, serve a list of metadata for audio.
Note that the actual audio data are not sent; instead, we respond
with URLs to the audio. The frontend should treat these URLs as
opaque and should not try to parse information about them or
generate them itself, as the format may change... |
Builds a JSON-serializable object with information about audio.
Args:
tensor_events: A list of image event_accumulator.TensorEvent objects.
run: The name of the run.
tag: The name of the tag the audio entries all belong to.
sample: The zero-indexed sample of the audio sample for which to
... |
Builds a URL for accessing the specified audio.
This should be kept in sync with _serve_audio_metadata. Note that the URL is
*not* guaranteed to always return the same audio, since audio may be
unloaded from the reservoir as new audio entries come in.
Args:
run: The name of the run.
tag: T... |
Serve encoded audio data.
def _serve_individual_audio(self, request):
"""Serve encoded audio data."""
tag = request.args.get('tag')
run = request.args.get('run')
index = int(request.args.get('index'))
sample = int(request.args.get('sample', 0))
events = self._filter_by_sample(self._multiplexer.... |
Writes __main__'s docstring to stdout with some help text.
Args:
shorthelp: bool, if True, prints only flags from the main module,
rather than all flags.
def _usage(shorthelp):
"""Writes __main__'s docstring to stdout with some help text.
Args:
shorthelp: bool, if True, prints only ... |
Runs the program with an optional 'main' function and 'argv' list.
def run(main=None, argv=None):
"""Runs the program with an optional 'main' function and 'argv' list."""
# Define help flags.
_define_help_flags()
# Parse known flags.
argv = flags.FLAGS(_sys.argv if argv is None else argv, known_o... |
Create a legacy image summary op for use in a TensorFlow graph.
Arguments:
name: A unique name for the generated summary node.
images: A `Tensor` representing pixel data with shape `[k, h, w, c]`,
where `k` is the number of images, `h` and `w` are the height and
width of the images, and `c` is th... |
Create a legacy image summary protobuf.
This behaves as if you were to create an `op` with the same arguments
(wrapped with constant tensors where appropriate) and then execute
that summary op in a TensorFlow session.
Arguments:
name: A unique name for the generated summary, including any desired
na... |
Apply user per-summary size guidance overrides.
def tensor_size_guidance_from_flags(flags):
"""Apply user per-summary size guidance overrides."""
tensor_size_guidance = dict(DEFAULT_TENSOR_SIZE_GUIDANCE)
if not flags or not flags.samples_per_plugin:
return tensor_size_guidance
for token in flags.samples_... |
Construct a TensorBoardWSGIApp with standard plugins and multiplexer.
Args:
flags: An argparse.Namespace containing TensorBoard CLI flags.
plugin_loaders: A list of TBLoader instances.
assets_zip_provider: See TBContext documentation for more information.
Returns:
The new TensorBoard WSGI applicat... |
Constructs the TensorBoard application.
Args:
logdir: the logdir spec that describes where data will be loaded.
may be a directory, or comma,separated list of directories, or colons
can be used to provide named directories
plugins: A list of base_plugin.TBPlugin subclass instances.
multiplexe... |
Parses `logdir` into a map from paths to run group names.
The events files flag format is a comma-separated list of path specifications.
A path specification either looks like 'group_name:/path/to/directory' or
'/path/to/directory'; in the latter case, the group is unnamed. Group names
cannot start with a forw... |
Starts automatically reloading the given multiplexer.
If `load_interval` is positive, the thread will reload the multiplexer
by calling `ReloadMultiplexer` every `load_interval` seconds, starting
immediately. Otherwise, reloads the multiplexer once and never again.
Args:
multiplexer: The `EventMultiplexer... |
Returns TBContext fields relating to SQL database.
Args:
db_uri: A string URI expressing the DB file, e.g. "sqlite:~/tb.db".
Returns:
A tuple with the db_module and db_connection_provider TBContext fields. If
db_uri was empty, then (None, None) is returned.
Raises:
ValueError: If db_uri scheme ... |
Returns function that returns SQLite Connection objects.
Args:
db_uri: A string URI expressing the DB file, e.g. "sqlite:~/tb.db".
Returns:
A function that returns a new PEP-249 DB Connection, which must be closed,
each time it is called.
Raises:
ValueError: If db_uri is not a valid sqlite file... |
Serves an object mapping plugin name to whether it is enabled.
Args:
request: The werkzeug.Request object.
Returns:
A werkzeug.Response object.
def _serve_plugins_listing(self, request):
"""Serves an object mapping plugin name to whether it is enabled.
Args:
request: The werkzeug.R... |
Parse a string as time indices.
Args:
s: A valid slicing string for time indices. E.g., '-1', '[:]', ':', '2:10'
Returns:
A slice object.
Raises:
ValueError: If `s` does not represent valid time indices.
def parse_time_indices(s):
"""Parse a string as time indices.
Args:
s: A valid slicin... |
Process a buffer for human-readable display.
This function performs the following operation on each of the buffers in `s`.
1. Truncate input buffer if the length of the buffer is greater than
`limit`, to prevent large strings from overloading the frontend.
2. Apply `binascii.b2a_qp` on the truncated b... |
View a slice or the entirety of an ndarray.
Args:
array: The input array, as an numpy.ndarray.
slicing: Optional slicing string, e.g., "[:, 1:3, :]".
mapping: Optional mapping string. Supported mappings:
`None` or case-insensitive `'None'`: Unmapped nested list.
`'image/png'`: Image encoding ... |
Convert an array into base64-enoded PNG image.
Args:
array: A 2D np.ndarray or nested list of items.
Returns:
A base64-encoded string the image. The image is grayscale if the array is
2D. The image is RGB color if the image is 3D with lsat dimension equal to
3.
Raises:
ValueError: If the in... |
Safely merge values from `src_proto_list` into `dst_proto_list`.
Each element in `dst_proto_list` must be mapped by `get_key` to a key
value that is unique within that list; likewise for `src_proto_list`.
If an element of `src_proto_list` has the same key as an existing
element in `dst_proto_list`, then the el... |
Combines two GraphDefs by adding nodes from from_proto into to_proto.
All GraphDefs are expected to be of TensorBoard's.
It assumes node names are unique across GraphDefs if contents differ. The
names can be the same if the NodeDef content are exactly the same.
Args:
to_proto: A destination TensorBoard Gr... |
Write a scalar summary.
Arguments:
name: A name for this summary. The summary tag used for TensorBoard will
be this name prefixed by any active name scopes.
data: A real numeric scalar value, convertible to a `float32` Tensor.
step: Explicit `int64`-castable monotonic step value for this summary. I... |
Create a scalar summary_pb2.Summary protobuf.
Arguments:
tag: String tag for the summary.
data: A 0-dimensional `np.array` or a compatible python number type.
description: Optional long-form description for this summary, as a
`str`. Markdown is supported. Defaults to empty.
Raises:
ValueErro... |
Dumps plugin data to the log directory.
def dump_data(logdir):
"""Dumps plugin data to the log directory."""
# Create a tfevents file in the logdir so it is detected as a run.
write_empty_event_file(logdir)
plugin_logdir = plugin_asset_util.PluginDirectory(
logdir, profile_plugin.ProfilePlugin.plugin_na... |
Calculate health pill of a tensor.
Args:
tensor: An instance of `np.array` (for initialized tensors) or
`tensorflow.python.debug.lib.debug_data.InconvertibleTensorProto`
(for unininitialized tensors).
Returns:
If `tensor` is an initialized tensor of numeric or boolean types:
the calculat... |
Reads the config file from disk or creates a new one.
def _get_config(self):
'''Reads the config file from disk or creates a new one.'''
filename = '{}/{}'.format(self.PLUGIN_LOGDIR, CONFIG_FILENAME)
modified_time = os.path.getmtime(filename)
if modified_time != self.config_last_modified_time:
c... |
Writes the frame to disk as a tensor summary.
def _write_summary(self, session, frame):
'''Writes the frame to disk as a tensor summary.'''
summary = session.run(self.summary_op, feed_dict={
self.frame_placeholder: frame
})
path = '{}/{}'.format(self.PLUGIN_LOGDIR, SUMMARY_FILENAME)
write_f... |
For limiting how often frames are computed.
def _enough_time_has_passed(self, FPS):
'''For limiting how often frames are computed.'''
if FPS == 0:
return False
else:
earliest_time = self.last_update_time + (1.0 / FPS)
return time.time() >= earliest_time |
Adds a frame to the current video output.
def _update_recording(self, frame, config):
'''Adds a frame to the current video output.'''
# pylint: disable=redefined-variable-type
should_record = config['is_recording']
if should_record:
if not self.is_recording:
self.is_recording = True
... |
Creates a frame and writes it to disk.
Args:
arrays: a list of np arrays. Use the "custom" option in the client.
frame: a 2D np array. This way the plugin can be used for video of any
kind, not just the visualization that comes with the plugin.
frame can also be a function, w... |
A helper to get the gradients out at each step.
Args:
optimizer: the optimizer op.
loss: the op that computes your loss value.
Returns: the gradient tensors and the train_step op.
def gradient_helper(optimizer, loss, var_list=None):
'''A helper to get the gradients out at each step.
Args... |
Returns a key_func to be used in list.sort().
Returns a key_func to be used in list.sort() that sorts session groups
by the value extracted by extractor. 'None' extracted values will either
be considered largest or smallest as specified by the "none_is_largest"
boolean parameter.
Args:
extractor: An ext... |
Creates extractors to extract properties corresponding to 'col_params'.
Args:
col_params: List of ListSessionGroupsRequest.ColParam protobufs.
Returns:
A list of extractor functions. The ith element in the
returned list extracts the column corresponding to the ith element of
_request.col_params
de... |
Returns function that extracts a metric from a session group or a session.
Args:
metric_name: tensorboard.hparams.MetricName protobuffer. Identifies the
metric to extract from the session group.
Returns:
A function that takes a tensorboard.hparams.SessionGroup or
tensorborad.hparams.Session protobu... |
Returns the metric_value for a given metric in a session or session group.
Args:
session_or_group: A Session protobuffer or SessionGroup protobuffer.
metric_name: A MetricName protobuffer. The metric to search for.
Returns:
A MetricValue protobuffer representing the value of the given metric or
Non... |
Returns an extractor function that extracts an hparam from a session group.
Args:
hparam_name: str. Identies the hparam to extract from the session group.
Returns:
A function that takes a tensorboard.hparams.SessionGroup protobuffer and
returns the value, as a native Python object, of the hparam identi... |
Creates filters for the given col_params.
Args:
col_params: List of ListSessionGroupsRequest.ColParam protobufs.
extractors: list of extractor functions of the same length as col_params.
Each element should extract the column described by the corresponding
element of col_params.
Returns:
A ... |
Creates a filter for the given col_param and extractor.
Args:
col_param: A tensorboard.hparams.ColParams object identifying the column
and describing the filter to apply.
extractor: A function that extract the column value identified by
'col_param' from a tensorboard.hparams.SessionGroup protobuf... |
Returns a boolean function that filters strings based on a regular exp.
Args:
regex: A string describing the regexp to use.
Returns:
A function taking a string and returns True if any of its substrings
matches regex.
def _create_regexp_filter(regex):
"""Returns a boolean function that filters string... |
Returns a function that checkes whether a number belongs to an interval.
Args:
interval: A tensorboard.hparams.Interval protobuf describing the interval.
Returns:
A function taking a number (a float or an object of a type in
six.integer_types) that returns True if the number belongs to (the closed)
... |
Converts a google.protobuf.Value to a native Python object.
def _value_to_python(value):
"""Converts a google.protobuf.Value to a native Python object."""
assert isinstance(value, struct_pb2.Value)
field = value.WhichOneof('kind')
if field == 'number_value':
return value.number_value
elif field == 'stri... |
Sets the metrics for the group to be the average of its sessions.
The resulting session group metrics consist of the union of metrics across
the group's sessions. The value of each session group metric is the average
of that metric values across the sessions in the group. The 'step' and
'wall_time_secs' fields... |
Sets the metrics for session_group to those of its "median session".
The median session is the session in session_group with the median value
of the metric given by 'aggregation_metric'. The median is taken over the
subset of sessions in the group whose 'aggregation_metric' was measured
at the largest training... |
Sets the metrics for session_group to those of its "extremum session".
The extremum session is the session in session_group with the extremum value
of the metric given by 'aggregation_metric'. The extremum is taken over the
subset of sessions in the group whose 'aggregation_metric' was measured
at the largest ... |
A generator for the values of the metric across the sessions in the group.
Args:
session_group: A SessionGroup protobuffer.
metric_name: A MetricName protobuffer.
Yields:
The next metric value wrapped in a _Measurement instance.
def _measurements(session_group, metric_name):
"""A generator for the v... |
Handles the request specified on construction.
Returns:
A ListSessionGroupsResponse object.
def run(self):
"""Handles the request specified on construction.
Returns:
A ListSessionGroupsResponse object.
"""
session_groups = self._build_session_groups()
session_groups = self._filte... |
Returns a list of SessionGroups protobuffers from the summary data.
def _build_session_groups(self):
"""Returns a list of SessionGroups protobuffers from the summary data."""
# Algorithm: We keep a dict 'groups_by_name' mapping a SessionGroup name
# (str) to a SessionGroup protobuffer. We traverse the run... |
Adds a new Session protobuffer to the 'groups_by_name' dictionary.
Called by _build_session_groups when we encounter a new session. Creates
the Session protobuffer and adds it to the relevant group in the
'groups_by_name' dict. Creates the session group if this is the first time
we encounter it.
A... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.