def scalar(name, scalar, collections=None, new_style=False):
"""Outputs a `Summary` protocol buffer containing a single scalar value.
The generated Summary has a Tensor.proto containing the input Tensor.
Args:
name: A name for the generated node. Will also serve as the series name in
TensorBoard.
tensor: A real numeric Tensor containing a single value.
collections: Optional list of graph collections keys. The new summary op is
added to these collections. Defaults to `[GraphKeys.SUMMARIES]`.
new_style: Whether to use new style (tensor field) or old style (simple_value
field). New style could lead to faster data loading.
Returns:
A scalar `Tensor` of type `string`. Which contains a `Summary` protobuf.
Raises:
ValueError: If tensor has the wrong shape or type.
"""
scalar = make_np(scalar)
assert scalar.squeeze().ndim == 0, "scalar should be 0D"
scalar = float(scalar)
if new_style:
plugin_data = SummaryMetadata.PluginData(plugin_name="scalars")
smd = SummaryMetadata(plugin_data=plugin_data)
return Summary(value=[
Summary.Value(
tag=name,
tensor=TensorProto(float_val=[scalar], dtype="DT_FLOAT"),
metadata=smd,
)
])
else:
return Summary(value=[Summary.Value(tag=name, simple_value=scalar)])
def event(step, values):
s = Summary()
scalar = [
Summary.Value(tag="{}/{}".format(name, field), simple_value=v)
for name, value in zip(names, values)
for field, v in value._asdict().items()
]
hist = [
Summary.Value(tag="{}/inferred_normal_hist".format(name),
histo=inferred_histo(value))
for name, value in zip(names, values)
]
s.value.extend(scalar + hist)
return Event(wall_time=int(step), step=step, summary=s)
def _ImageSummary(tag, height, width, colorspace, encoded_image):
from tensorboard.compat.proto.summary_pb2 import Summary
image = Summary.Image(height=height,
width=width,
colorspace=colorspace,
encoded_image_string=encoded_image)
return Summary(value=[Summary.Value(tag=tag, image=image)])
def make_tensor_summary(name, nparray):
tensor_pb = TensorProto(
dtype='DT_FLOAT',
float_val=nparray.reshape(-1).tolist(),
tensor_shape=TensorShapeProto(
dim=[TensorShapeProto.Dim(size=s) for s in nparray.shape]))
return Summary(value=[Summary.Value(tag=name, tensor=tensor_pb)])
def _add_3d_torch(self,
tag,
data,
step,
logdir=None,
max_outputs=1,
label_to_names=None,
description=None):
walltime = None
if step is None:
raise ValueError("Step is not provided or set.")
mdata = {} if label_to_names is None else {'label_to_names': label_to_names}
summary_metadata = metadata.create_summary_metadata(description=description,
metadata=mdata)
writer = self._get_file_writer()
if logdir is None:
logdir = writer.get_logdir()
write_dir = PluginDirectory(logdir, metadata.PLUGIN_NAME)
geometry_metadata_string = _write_geometry_data(write_dir, tag, step, data,
max_outputs)
tensor_proto = TensorProto(dtype='DT_STRING',
string_val=[geometry_metadata_string],
tensor_shape=TensorShapeProto())
writer.add_summary(
Summary(value=[
Summary.Value(
tag=tag, tensor=tensor_proto, metadata=summary_metadata)
]), step, walltime)
def add_scalar(self, tag, scalar_value, global_step=None, walltime=None):
r"""Add a scalar value."""
scalar_value = float(scalar_value)
self._get_file_writer().add_summary(
Summary(value=[Summary.Value(tag=tag, simple_value=scalar_value)]),
global_step, walltime
)
def audio(tag, tensor, sample_rate=44100):
tensor = make_np(tensor)
tensor = tensor.squeeze()
if abs(tensor).max() > 1:
print('warning: audio amplitude out of range, auto clipped.')
tensor = tensor.clip(-1, 1)
assert (tensor.ndim == 1), 'input tensor should be 1 dimensional.'
tensor_list = [int(32767.0 * x) for x in tensor]
import io
import wave
import struct
fio = io.BytesIO()
wave_write = wave.open(fio, 'wb')
wave_write.setnchannels(1)
wave_write.setsampwidth(2)
wave_write.setframerate(sample_rate)
tensor_enc = b''
for v in tensor_list:
tensor_enc += struct.pack('<h', v)
wave_write.writeframes(tensor_enc)
wave_write.close()
audio_string = fio.getvalue()
fio.close()
audio = Summary.Audio(sample_rate=sample_rate,
num_channels=1,
length_frames=len(tensor_list),
encoded_audio_string=audio_string,
content_type='audio/wav')
return Summary(value=[Summary.Value(tag=tag, audio=audio)])
def image(tag, tensor, rescale=1, dataformats="NCHW"):
"""Outputs a `Summary` protocol buffer with images.
The summary has up to `max_images` summary values containing images. The
images are built from `tensor` which must be 3-D with shape `[height, width,
channels]` and where `channels` can be:
* 1: `tensor` is interpreted as Grayscale.
* 3: `tensor` is interpreted as RGB.
* 4: `tensor` is interpreted as RGBA.
The `name` in the outputted Summary.Value protobufs is generated based on the
name, with a suffix depending on the max_outputs setting:
* If `max_outputs` is 1, the summary value tag is '*name*/image'.
* If `max_outputs` is greater than 1, the summary value tags are
generated sequentially as '*name*/image/0', '*name*/image/1', etc.
Args:
tag: A name for the generated node. Will also serve as a series name in
TensorBoard.
tensor: A 3-D `uint8` or `float32` `Tensor` of shape `[height, width,
channels]` where `channels` is 1, 3, or 4.
'tensor' can either have values in [0, 1] (float32) or [0, 255] (uint8).
The image() function will scale the image values to [0, 255] by applying
a scale factor of either 1 (uint8) or 255 (float32).
Returns:
A scalar `Tensor` of type `string`. The serialized `Summary` protocol
buffer.
"""
tensor = make_np(tensor)
tensor = convert_to_HWC(tensor, dataformats)
# Do not assume that user passes in values in [0, 255], use data type to detect
scale_factor = _calc_scale_factor(tensor)
tensor = tensor.astype(np.float32)
tensor = (tensor * scale_factor).astype(np.uint8)
image = make_image(tensor, rescale=rescale)
return Summary(value=[Summary.Value(tag=tag, image=image)])
def _make_session_end_summary(status: str,
end_time_secs: Optional[int] = None):
"""
Args:
status: outcome of this run, one of of 'UNKNOWN', 'SUCCESS', 'FAILURE', 'RUNNING'
end_time_secs: optional ending time in seconds
Returns:
"""
status = Status.DESCRIPTOR.values_by_name[
f"STATUS_{status.upper()}"].number
if end_time_secs is None:
end_time_secs = int(time.time())
session_end_summary = SessionEndInfo(status=status,
end_time_secs=end_time_secs)
session_end_content = HParamsPluginData(
session_end_info=session_end_summary, version=PLUGIN_DATA_VERSION)
session_end_summary_metadata = SummaryMetadata(
plugin_data=SummaryMetadata.PluginData(
plugin_name=PLUGIN_NAME,
content=session_end_content.SerializeToString()))
session_end_summary = Summary(value=[
Summary.Value(tag=SESSION_END_INFO_TAG,
metadata=session_end_summary_metadata)
])
return session_end_summary
def audio(tag, tensor, sample_rate=44100):
tensor = make_np(tensor)
tensor = tensor.squeeze()
if abs(tensor).max() > 1:
print('warning: audio amplitude out of range, auto clipped.')
tensor = tensor.clip(-1, 1)
assert (tensor.ndim == 1), 'input tensor should be 1 dimensional.'
tensor = (tensor * np.iinfo(np.int16).max).astype('<i2')
import io
import wave
fio = io.BytesIO()
wave_write = wave.open(fio, 'wb')
wave_write.setnchannels(1)
wave_write.setsampwidth(2)
wave_write.setframerate(sample_rate)
wave_write.writeframes(tensor.data)
wave_write.close()
audio_string = fio.getvalue()
fio.close()
audio = Summary.Audio(sample_rate=sample_rate,
num_channels=1,
length_frames=tensor.shape[-1],
encoded_audio_string=audio_string,
content_type='audio/wav')
return Summary(value=[Summary.Value(tag=tag, audio=audio)])
def audio(tag, tensor, sample_rate=44100):
array = make_np(tensor)
array = array.squeeze()
if abs(array).max() > 1:
print("warning: audio amplitude out of range, auto clipped.")
array = array.clip(-1, 1)
assert array.ndim == 1, "input tensor should be 1 dimensional."
array = (array * np.iinfo(np.int16).max).astype("<i2")
import io
import wave
fio = io.BytesIO()
wave_write = wave.open(fio, "wb")
wave_write.setnchannels(1)
wave_write.setsampwidth(2)
wave_write.setframerate(sample_rate)
wave_write.writeframes(array.data)
wave_write.close()
audio_string = fio.getvalue()
fio.close()
audio = Summary.Audio(
sample_rate=sample_rate,
num_channels=1,
length_frames=array.shape[-1],
encoded_audio_string=audio_string,
content_type="audio/wav",
)
return Summary(value=[Summary.Value(tag=tag, audio=audio)])
def custom_scalars(layout):
categories = []
for k, v in layout.items():
charts = []
for chart_name, chart_meatadata in v.items():
tags = chart_meatadata[1]
if chart_meatadata[0] == "Margin":
assert len(tags) == 3
mgcc = layout_pb2.MarginChartContent(series=[
layout_pb2.MarginChartContent.Series(
value=tags[0], lower=tags[1], upper=tags[2])
])
chart = layout_pb2.Chart(title=chart_name, margin=mgcc)
else:
mlcc = layout_pb2.MultilineChartContent(tag=tags)
chart = layout_pb2.Chart(title=chart_name, multiline=mlcc)
charts.append(chart)
categories.append(layout_pb2.Category(title=k, chart=charts))
layout = layout_pb2.Layout(category=categories)
plugin_data = SummaryMetadata.PluginData(plugin_name="custom_scalars")
smd = SummaryMetadata(plugin_data=plugin_data)
tensor = TensorProto(
dtype="DT_STRING",
string_val=[layout.SerializeToString()],
tensor_shape=TensorShapeProto(),
)
return Summary(value=[
Summary.Value(
tag="custom_scalars__config__", tensor=tensor, metadata=smd)
])
def pr_curve_raw(tag,
tp,
fp,
tn,
fn,
precision,
recall,
num_thresholds=127,
weights=None):
if num_thresholds > 127: # weird, value > 127 breaks protobuf
num_thresholds = 127
data = np.stack((tp, fp, tn, fn, precision, recall))
pr_curve_plugin_data = PrCurvePluginData(
version=0, num_thresholds=num_thresholds).SerializeToString()
plugin_data = SummaryMetadata.PluginData(plugin_name="pr_curves",
content=pr_curve_plugin_data)
smd = SummaryMetadata(plugin_data=plugin_data)
tensor = TensorProto(
dtype="DT_FLOAT",
float_val=data.reshape(-1).tolist(),
tensor_shape=TensorShapeProto(dim=[
TensorShapeProto.Dim(size=data.shape[0]),
TensorShapeProto.Dim(size=data.shape[1]),
]),
)
return Summary(value=[Summary.Value(tag=tag, metadata=smd, tensor=tensor)])
def _image3_animated_gif(tag: str,
image: Union[np.ndarray, torch.Tensor],
scale_factor: float = 1.0) -> Summary:
"""Function to actually create the animated gif.
Args:
tag: Data identifier
image: 3D image tensors expected to be in `HWD` format
scale_factor: amount to multiply values by. if the image data is between 0 and 1, using 255 for this value will
scale it to displayable range
"""
assert len(
image.shape
) == 3, "3D image tensors expected to be in `HWD` format, len(image.shape) != 3"
ims = [(np.asarray((image[:, :, i])) * scale_factor).astype(np.uint8)
for i in range(image.shape[2])]
ims = [GifImage.fromarray(im) for im in ims]
img_str = b""
for b_data in PIL.GifImagePlugin.getheader(ims[0])[0]:
img_str += b_data
img_str += b"\x21\xFF\x0B\x4E\x45\x54\x53\x43\x41\x50" b"\x45\x32\x2E\x30\x03\x01\x00\x00\x00"
for i in ims:
for b_data in PIL.GifImagePlugin.getdata(i):
img_str += b_data
img_str += b"\x3B"
summary_image_str = Summary.Image(height=10,
width=10,
colorspace=1,
encoded_image_string=img_str)
image_summary = Summary.Value(tag=tag, image=summary_image_str)
return Summary(value=[image_summary])
def histogram_raw(name, min, max, num, sum, sum_squares, bucket_limits,
bucket_counts):
# pylint: disable=line-too-long
"""Outputs a `Summary` protocol buffer with a histogram.
The generated
[`Summary`](https://www.tensorflow.org/code/tensorflow/core/framework/summary.proto)
has one summary value containing a histogram for `values`.
Args:
name: A name for the generated node. Will also serve as a series name in
TensorBoard.
min: A float or int min value
max: A float or int max value
num: Int number of values
sum: Float or int sum of all values
sum_squares: Float or int sum of squares for all values
bucket_limits: A numeric `Tensor` with upper value per bucket
bucket_counts: A numeric `Tensor` with number of values per bucket
Returns:
A scalar `Tensor` of type `string`. The serialized `Summary` protocol
buffer.
"""
hist = HistogramProto(
min=min,
max=max,
num=num,
sum=sum,
sum_squares=sum_squares,
bucket_limit=bucket_limits,
bucket=bucket_counts,
)
return Summary(value=[Summary.Value(tag=name, histo=hist)])
def text(tag, text):
plugin_data = SummaryMetadata.PluginData(
plugin_name='text', content=TextPluginData(version=0).SerializeToString())
smd = SummaryMetadata(plugin_data=plugin_data)
tensor = TensorProto(dtype='DT_STRING',
string_val=[text.encode(encoding='utf_8')],
tensor_shape=TensorShapeProto(dim=[TensorShapeProto.Dim(size=1)]))
return Summary(value=[Summary.Value(tag=tag + '/text_summary', metadata=smd, tensor=tensor)])
def video(tag, tensor, fps=4):
tensor = make_np(tensor)
tensor = _prepare_video(tensor)
# If user passes in uint8, then we don't need to rescale by 255
scale_factor = _calc_scale_factor(tensor)
tensor = tensor.astype(np.float32)
tensor = (tensor * scale_factor).astype(np.uint8)
video = make_video(tensor, fps)
return Summary(value=[Summary.Value(tag=tag, image=video)])
def _make_session_start_summary(
hparam_values,
group_name: Optional[str] = None,
start_time_secs: Optional[int] = None,
):
"""Assign values to the hyperparameters in the context of this session.
Args:
hparam_values: a dict of ``hp_name`` -> ``hp_value`` mappings
group_name: optional group name for this session
start_time_secs: optional starting time in seconds
Returns:
"""
if start_time_secs is None:
start_time_secs = int(time.time())
session_start_info = SessionStartInfo(group_name=group_name,
start_time_secs=start_time_secs)
for hp_name, hp_value in hparam_values.items():
# Logging a None would raise an exception when setting session_start_info.hparams[hp_name].number_value = None.
# Logging a float.nan instead would work, but that run would not show at all in the tensorboard hparam plugin.
# The best thing to do here is to skip that value, it will show as a blank cell in the table view of the
# tensorboard plugin. However, that run would not be shown in the parallel coord or in the scatter plot view.
if hp_value is None:
logger.warning(
f"Hyper parameter {hp_name} is `None`: the tensorboard hp plugin "
f"will show this run in table view, but not in parallel coordinates "
f"view or in scatter plot matrix view")
continue
if isinstance(hp_value, (str, list, tuple)):
session_start_info.hparams[hp_name].string_value = str(hp_value)
continue
if isinstance(hp_value, bool):
session_start_info.hparams[hp_name].bool_value = hp_value
continue
if not isinstance(hp_value, (int, float)):
hp_value = make_np(hp_value)[0]
session_start_info.hparams[hp_name].number_value = hp_value
session_start_content = HParamsPluginData(
session_start_info=session_start_info, version=PLUGIN_DATA_VERSION)
session_start_summary_metadata = SummaryMetadata(
plugin_data=SummaryMetadata.PluginData(
plugin_name=PLUGIN_NAME,
content=session_start_content.SerializeToString()))
session_start_summary = Summary(value=[
Summary.Value(tag=SESSION_START_INFO_TAG,
metadata=session_start_summary_metadata)
])
return session_start_summary
def _get_tensor_summary(
name, display_name, description, tensor, content_type, components, json_config
):
"""Creates a tensor summary with summary metadata.
Args:
name: Uniquely identifiable name of the summary op. Could be replaced by
combination of name and type to make it unique even outside of this
summary.
display_name: Will be used as the display name in TensorBoard.
Defaults to `name`.
description: A longform readable description of the summary data. Markdown
is supported.
tensor: Tensor to display in summary.
content_type: Type of content inside the Tensor.
components: Bitmask representing present parts (vertices, colors, etc.) that
belong to the summary.
json_config: A string, JSON-serialized dictionary of ThreeJS classes
configuration.
Returns:
Tensor summary with metadata.
"""
import torch
from tensorboard.plugins.mesh import metadata
tensor = torch.as_tensor(tensor)
tensor_metadata = metadata.create_summary_metadata(
name,
display_name,
content_type,
components,
tensor.shape,
description,
json_config=json_config,
)
tensor = TensorProto(
dtype="DT_FLOAT",
float_val=tensor.reshape(-1).tolist(),
tensor_shape=TensorShapeProto(
dim=[
TensorShapeProto.Dim(size=tensor.shape[0]),
TensorShapeProto.Dim(size=tensor.shape[1]),
TensorShapeProto.Dim(size=tensor.shape[2]),
]
),
)
tensor_summary = Summary.Value(
tag=metadata.get_instance_name(name, content_type),
tensor=tensor,
metadata=tensor_metadata,
)
return tensor_summary
def image_boxes(tag, tensor_image, tensor_boxes, rescale=1, dataformats='CHW'):
'''Outputs a `Summary` protocol buffer with images.'''
tensor_image = make_np(tensor_image)
tensor_image = convert_to_HWC(tensor_image, dataformats)
tensor_boxes = make_np(tensor_boxes)
tensor_image = tensor_image.astype(
np.float32) * _calc_scale_factor(tensor_image)
image = make_image(tensor_image.astype(np.uint8),
rescale=rescale,
rois=tensor_boxes)
return Summary(value=[Summary.Value(tag=tag, image=image)])
def scalar(name, tensor, collections=None, new_style=False, double_precision=False):
"""Outputs a `Summary` protocol buffer containing a single scalar value.
The generated Summary has a Tensor.proto containing the input Tensor.
Args:
name: A name for the generated node. Will also serve as the series name in
TensorBoard.
tensor: A real numeric Tensor containing a single value.
collections: Optional list of graph collections keys. The new summary op is
added to these collections. Defaults to `[GraphKeys.SUMMARIES]`.
new_style: Whether to use new style (tensor field) or old style (simple_value
field). New style could lead to faster data loading.
Returns:
A scalar `Tensor` of type `string`. Which contains a `Summary` protobuf.
Raises:
ValueError: If tensor has the wrong shape or type.
"""
tensor = make_np(tensor).squeeze()
assert (
tensor.ndim == 0
), f"Tensor should contain one element (0 dimensions). Was given size: {tensor.size} and {tensor.ndim} dimensions."
# python float is double precision in numpy
scalar = float(tensor)
if new_style:
tensor_proto = TensorProto(float_val=[scalar], dtype="DT_FLOAT")
if double_precision:
tensor_proto = TensorProto(double_val=[scalar], dtype="DT_DOUBLE")
plugin_data = SummaryMetadata.PluginData(plugin_name="scalars")
smd = SummaryMetadata(plugin_data=plugin_data)
return Summary(
value=[
Summary.Value(
tag=name,
tensor=tensor_proto,
metadata=smd,
)
]
)
else:
return Summary(value=[Summary.Value(tag=name, simple_value=scalar)])
def test_event_file_writer_roundtrip(self): _TAGNAME = 'dummy' _DUMMY_VALUE = 42 logdir = self.get_temp_dir() w = EventFileWriter(logdir) summary = Summary(value=[Summary.Value(tag=_TAGNAME, simple_value=_DUMMY_VALUE)]) fakeevent = event_pb2.Event(summary=summary) w.add_event(fakeevent) w.close() event_files = sorted(glob.glob(os.path.join(logdir, '*'))) self.assertEqual(len(event_files), 1) r = PyRecordReader_New(event_files[0]) r.GetNext() # meta data, so skip r.GetNext() self.assertEqual(fakeevent.SerializeToString(), r.record())
def pr_curve(tag, labels, predictions, num_thresholds=127, weights=None):
# weird, value > 127 breaks protobuf
num_thresholds = min(num_thresholds, 127)
data = compute_curve(labels, predictions,
num_thresholds=num_thresholds, weights=weights)
pr_curve_plugin_data = PrCurvePluginData(
version=0, num_thresholds=num_thresholds).SerializeToString()
plugin_data = SummaryMetadata.PluginData(
plugin_name='pr_curves', content=pr_curve_plugin_data)
smd = SummaryMetadata(plugin_data=plugin_data)
tensor = TensorProto(dtype='DT_FLOAT',
float_val=data.reshape(-1).tolist(),
tensor_shape=TensorShapeProto(
dim=[TensorShapeProto.Dim(size=data.shape[0]), TensorShapeProto.Dim(size=data.shape[1])]))
return Summary(value=[Summary.Value(tag=tag, metadata=smd, tensor=tensor)])
def scalar(name, scalar, collections=None):
"""Outputs a `Summary` protocol buffer containing a single scalar value.
The generated Summary has a Tensor.proto containing the input Tensor.
Args:
name: A name for the generated node. Will also serve as the series name in
TensorBoard.
tensor: A real numeric Tensor containing a single value.
collections: Optional list of graph collections keys. The new summary op is
added to these collections. Defaults to `[GraphKeys.SUMMARIES]`.
Returns:
A scalar `Tensor` of type `string`. Which contains a `Summary` protobuf.
Raises:
ValueError: If tensor has the wrong shape or type.
"""
scalar = make_np(scalar)
assert (scalar.squeeze().ndim == 0), 'scalar should be 0D'
scalar = float(scalar)
return Summary(value=[Summary.Value(tag=name, simple_value=scalar)])
def _write_logs(self, logs, index):
"""Write log values to the log files
:param logs: holds the loss and metric values computed at the most
recent interval (batch or epoch)
:type logs: dict
:param index: if update_freq='batch', the total number of samples that
have been seen, else if update_freq='epoch', the epoch index
:type index: int
"""
for name, value in logs.items():
if name in ['batch', 'size']:
continue
summary = Summary(
value=[Summary.Value(tag=name, simple_value=value)])
self.writer.add_summary(summary, index)
self.writer.flush()
def histogram(name, values, bins, max_bins=None):
# pylint: disable=line-too-long
"""Outputs a `Summary` protocol buffer with a histogram.
The generated
[`Summary`](https://www.tensorflow.org/code/tensorflow/core/framework/summary.proto)
has one summary value containing a histogram for `values`.
This op reports an `InvalidArgument` error if any value is not finite.
Args:
name: A name for the generated node. Will also serve as a series name in
TensorBoard.
values: A real numeric `Tensor`. Any shape. Values to use to
build the histogram.
Returns:
A scalar `Tensor` of type `string`. The serialized `Summary` protocol
buffer.
"""
values = make_np(values)
hist = make_histogram(values.astype(float), bins, max_bins)
return Summary(value=[Summary.Value(tag=name, histo=hist)])
def hparams(hparam_dict=None, metric_dict=None):
"""Outputs three `Summary` protocol buffers needed by hparams plugin.
`Experiment` keeps the metadata of an experiment, such as the name of the
hyperparameters and the name of the metrics.
`SessionStartInfo` keeps key-value pairs of the hyperparameters
`SessionEndInfo` describes status of the experiment e.g. STATUS_SUCCESS
Args:
hparam_dict: A dictionary that contains names of the hyperparameters
and their values.
metric_dict: A dictionary that contains names of the metrics
and their values.
Returns:
The `Summary` protobufs for Experiment, SessionStartInfo and
SessionEndInfo
"""
import torch
from six import string_types
from tensorboard.plugins.hparams.api_pb2 import (Experiment, HParamInfo,
MetricInfo, MetricName,
Status)
from tensorboard.plugins.hparams.metadata import (PLUGIN_NAME,
PLUGIN_DATA_VERSION,
EXPERIMENT_TAG,
SESSION_START_INFO_TAG,
SESSION_END_INFO_TAG)
from tensorboard.plugins.hparams.plugin_data_pb2 import (HParamsPluginData,
SessionEndInfo,
SessionStartInfo)
# TODO: expose other parameters in the future.
# hp = HParamInfo(name='lr',display_name='learning rate',
# type=DataType.DATA_TYPE_FLOAT64, domain_interval=Interval(min_value=10,
# max_value=100))
# mt = MetricInfo(name=MetricName(tag='accuracy'), display_name='accuracy',
# description='', dataset_type=DatasetType.DATASET_VALIDATION)
# exp = Experiment(name='123', description='456', time_created_secs=100.0,
# hparam_infos=[hp], metric_infos=[mt], user='******')
if not isinstance(hparam_dict, dict):
logging.warning(
'parameter: hparam_dict should be a dictionary, nothing logged.')
raise TypeError(
'parameter: hparam_dict should be a dictionary, nothing logged.')
if not isinstance(metric_dict, dict):
logging.warning(
'parameter: metric_dict should be a dictionary, nothing logged.')
raise TypeError(
'parameter: metric_dict should be a dictionary, nothing logged.')
hps = [HParamInfo(name=k) for k in hparam_dict.keys()]
mts = [MetricInfo(name=MetricName(tag=k)) for k in metric_dict.keys()]
exp = Experiment(hparam_infos=hps, metric_infos=mts)
content = HParamsPluginData(experiment=exp, version=PLUGIN_DATA_VERSION)
smd = SummaryMetadata(plugin_data=SummaryMetadata.PluginData(
plugin_name=PLUGIN_NAME, content=content.SerializeToString()))
exp = Summary(value=[Summary.Value(tag=EXPERIMENT_TAG, metadata=smd)])
ssi = SessionStartInfo()
for k, v in hparam_dict.items():
if v is None:
continue
if isinstance(v, int) or isinstance(v, float):
ssi.hparams[k].number_value = v
continue
if isinstance(v, string_types):
ssi.hparams[k].string_value = v
continue
if isinstance(v, bool):
ssi.hparams[k].bool_value = v
continue
if isinstance(v, torch.Tensor):
v = make_np(v)[0]
ssi.hparams[k].number_value = v
continue
raise ValueError(
'value should be one of int, float, str, bool, or torch.Tensor')
content = HParamsPluginData(session_start_info=ssi,
version=PLUGIN_DATA_VERSION)
smd = SummaryMetadata(plugin_data=SummaryMetadata.PluginData(
plugin_name=PLUGIN_NAME, content=content.SerializeToString()))
ssi = Summary(
value=[Summary.Value(tag=SESSION_START_INFO_TAG, metadata=smd)])
sei = SessionEndInfo(status=Status.Value('STATUS_SUCCESS'))
content = HParamsPluginData(session_end_info=sei,
version=PLUGIN_DATA_VERSION)
smd = SummaryMetadata(plugin_data=SummaryMetadata.PluginData(
plugin_name=PLUGIN_NAME, content=content.SerializeToString()))
sei = Summary(
value=[Summary.Value(tag=SESSION_END_INFO_TAG, metadata=smd)])
return exp, ssi, sei
def hparams(hparam_dict=None, metric_dict=None, hparam_domain_discrete=None):
"""Outputs three `Summary` protocol buffers needed by hparams plugin.
`Experiment` keeps the metadata of an experiment, such as the name of the
hyperparameters and the name of the metrics.
`SessionStartInfo` keeps key-value pairs of the hyperparameters
`SessionEndInfo` describes status of the experiment e.g. STATUS_SUCCESS
Args:
hparam_dict: A dictionary that contains names of the hyperparameters
and their values.
metric_dict: A dictionary that contains names of the metrics
and their values.
hparam_domain_discrete: (Optional[Dict[str, List[Any]]]) A dictionary that
contains names of the hyperparameters and all discrete values they can hold
Returns:
The `Summary` protobufs for Experiment, SessionStartInfo and
SessionEndInfo
"""
import torch
from six import string_types
from tensorboard.plugins.hparams.api_pb2 import (
Experiment,
HParamInfo,
MetricInfo,
MetricName,
Status,
DataType,
)
from tensorboard.plugins.hparams.metadata import (
PLUGIN_NAME,
PLUGIN_DATA_VERSION,
EXPERIMENT_TAG,
SESSION_START_INFO_TAG,
SESSION_END_INFO_TAG,
)
from tensorboard.plugins.hparams.plugin_data_pb2 import (
HParamsPluginData,
SessionEndInfo,
SessionStartInfo,
)
# TODO: expose other parameters in the future.
# hp = HParamInfo(name='lr',display_name='learning rate',
# type=DataType.DATA_TYPE_FLOAT64, domain_interval=Interval(min_value=10,
# max_value=100))
# mt = MetricInfo(name=MetricName(tag='accuracy'), display_name='accuracy',
# description='', dataset_type=DatasetType.DATASET_VALIDATION)
# exp = Experiment(name='123', description='456', time_created_secs=100.0,
# hparam_infos=[hp], metric_infos=[mt], user='******')
if not isinstance(hparam_dict, dict):
logger.warning(
"parameter: hparam_dict should be a dictionary, nothing logged.")
raise TypeError(
"parameter: hparam_dict should be a dictionary, nothing logged.")
if not isinstance(metric_dict, dict):
logger.warning(
"parameter: metric_dict should be a dictionary, nothing logged.")
raise TypeError(
"parameter: metric_dict should be a dictionary, nothing logged.")
hparam_domain_discrete = hparam_domain_discrete or {}
if not isinstance(hparam_domain_discrete, dict):
raise TypeError(
"parameter: hparam_domain_discrete should be a dictionary, nothing logged."
)
for k, v in hparam_domain_discrete.items():
if (k not in hparam_dict or not isinstance(v, list)
or not all(isinstance(d, type(hparam_dict[k])) for d in v)):
raise TypeError(
"parameter: hparam_domain_discrete[{}] should be a list of same type as "
"hparam_dict[{}].".format(k, k))
hps = []
ssi = SessionStartInfo()
for k, v in hparam_dict.items():
if v is None:
continue
if isinstance(v, int) or isinstance(v, float):
ssi.hparams[k].number_value = v
if k in hparam_domain_discrete:
domain_discrete: Optional[
struct_pb2.ListValue] = struct_pb2.ListValue(values=[
struct_pb2.Value(number_value=d)
for d in hparam_domain_discrete[k]
])
else:
domain_discrete = None
hps.append(
HParamInfo(
name=k,
type=DataType.Value("DATA_TYPE_FLOAT64"),
domain_discrete=domain_discrete,
))
continue
if isinstance(v, string_types):
ssi.hparams[k].string_value = v
if k in hparam_domain_discrete:
domain_discrete = struct_pb2.ListValue(values=[
struct_pb2.Value(string_value=d)
for d in hparam_domain_discrete[k]
])
else:
domain_discrete = None
hps.append(
HParamInfo(
name=k,
type=DataType.Value("DATA_TYPE_STRING"),
domain_discrete=domain_discrete,
))
continue
if isinstance(v, bool):
ssi.hparams[k].bool_value = v
if k in hparam_domain_discrete:
domain_discrete = struct_pb2.ListValue(values=[
struct_pb2.Value(bool_value=d)
for d in hparam_domain_discrete[k]
])
else:
domain_discrete = None
hps.append(
HParamInfo(
name=k,
type=DataType.Value("DATA_TYPE_BOOL"),
domain_discrete=domain_discrete,
))
continue
if isinstance(v, torch.Tensor):
v = make_np(v)[0]
ssi.hparams[k].number_value = v
hps.append(
HParamInfo(name=k, type=DataType.Value("DATA_TYPE_FLOAT64")))
continue
raise ValueError(
"value should be one of int, float, str, bool, or torch.Tensor")
content = HParamsPluginData(session_start_info=ssi,
version=PLUGIN_DATA_VERSION)
smd = SummaryMetadata(plugin_data=SummaryMetadata.PluginData(
plugin_name=PLUGIN_NAME, content=content.SerializeToString()))
ssi = Summary(
value=[Summary.Value(tag=SESSION_START_INFO_TAG, metadata=smd)])
mts = [MetricInfo(name=MetricName(tag=k)) for k in metric_dict.keys()]
exp = Experiment(hparam_infos=hps, metric_infos=mts)
content = HParamsPluginData(experiment=exp, version=PLUGIN_DATA_VERSION)
smd = SummaryMetadata(plugin_data=SummaryMetadata.PluginData(
plugin_name=PLUGIN_NAME, content=content.SerializeToString()))
exp = Summary(value=[Summary.Value(tag=EXPERIMENT_TAG, metadata=smd)])
sei = SessionEndInfo(status=Status.Value("STATUS_SUCCESS"))
content = HParamsPluginData(session_end_info=sei,
version=PLUGIN_DATA_VERSION)
smd = SummaryMetadata(plugin_data=SummaryMetadata.PluginData(
plugin_name=PLUGIN_NAME, content=content.SerializeToString()))
sei = Summary(
value=[Summary.Value(tag=SESSION_END_INFO_TAG, metadata=smd)])
return exp, ssi, sei
def _make_experiment_summary(hparam_infos, metric_infos, experiment):
"""Define hyperparameters and metrics.
Args:
hparam_infos: information about all hyperparameters (name, description, type etc.),
list of dicts containing 'name' (required), 'type', 'description', 'display_name',
'domain_discrete', 'domain_interval'
metric_infos: information about all metrics (tag, description etc.),
list of dicts containing 'tag' (required), 'dataset_type', 'description', 'display_name'
experiment: dict containing 'name' (required), 'description', 'time_created_secs', 'user'
Returns:
"""
def make_hparam_info(hparam):
data_types = {
None: DataType.DATA_TYPE_UNSET,
str: DataType.DATA_TYPE_STRING,
list: DataType.DATA_TYPE_STRING,
tuple: DataType.DATA_TYPE_STRING,
bool: DataType.DATA_TYPE_BOOL,
int: DataType.DATA_TYPE_FLOAT64,
float: DataType.DATA_TYPE_FLOAT64,
}
return HParamInfo(
name=hparam["name"],
type=data_types[hparam.get("type")],
description=hparam.get("description"),
display_name=hparam.get("display_name"),
domain_discrete=hparam.get("domain_discrete"),
domain_interval=hparam.get("domain_interval"),
)
def make_metric_info(metric):
return MetricInfo(
name=MetricName(tag=metric["tag"]),
dataset_type=DatasetType.Value(
f'DATASET_{metric.get("dataset_type", "UNKNOWN").upper()}'),
description=metric.get("description"),
display_name=metric.get("display_name"),
)
def make_experiment_info(experiment, metric_infos, hparam_infos):
return Experiment(
name=experiment["name"],
description=experiment.get("description"),
time_created_secs=experiment.get("time_created_secs"),
user=experiment.get("user"),
metric_infos=metric_infos,
hparam_infos=hparam_infos,
)
metric_infos = [make_metric_info(m) for m in metric_infos]
hparam_infos = [make_hparam_info(hp) for hp in hparam_infos]
experiment = make_experiment_info(experiment, metric_infos, hparam_infos)
experiment_content = HParamsPluginData(experiment=experiment,
version=PLUGIN_DATA_VERSION)
experiment_summary_metadata = SummaryMetadata(
plugin_data=SummaryMetadata.PluginData(
plugin_name=PLUGIN_NAME,
content=experiment_content.SerializeToString()))
experiment_summary = Summary(value=[
Summary.Value(tag=EXPERIMENT_TAG, metadata=experiment_summary_metadata)
])
return experiment_summary
def add_image(self, tag, img, global_step=None, walltime=None):
r"""Add an image."""
self._get_file_writer().add_summary(
Summary(value=[Summary.Value(tag=tag, image=img)])
)