mirror of
https://github.com/vale981/ray
synced 2025-03-09 04:46:38 -04:00
236951ee4c
This PR introduces a TrialCheckpoint class which is returned e.g. by ExperimentAnalysis.best_checkpoint. The class enables easy access to cloud storage locations (rather than just local directories before). It also comes with utilities to download, upload, and save trial checkpoints to local and cloud targets.
51 lines
No EOL
1.4 KiB
ReStructuredText
51 lines
No EOL
1.4 KiB
ReStructuredText
.. _tune-analysis-docs:
|
|
|
|
Analysis (tune.analysis)
|
|
========================
|
|
|
|
You can use the ``ExperimentAnalysis`` object for analyzing results. It is returned automatically when calling ``tune.run``.
|
|
|
|
.. code-block:: python
|
|
|
|
analysis = tune.run(
|
|
trainable,
|
|
name="example-experiment",
|
|
num_samples=10,
|
|
)
|
|
|
|
Here are some example operations for obtaining a summary of your experiment:
|
|
|
|
.. code-block:: python
|
|
|
|
# Get a dataframe for the last reported results of all of the trials
|
|
df = analysis.results_df
|
|
|
|
# Get a dataframe for the max accuracy seen for each trial
|
|
df = analysis.dataframe(metric="mean_accuracy", mode="max")
|
|
|
|
# Get a dict mapping {trial logdir -> dataframes} for all trials in the experiment.
|
|
all_dataframes = analysis.trial_dataframes
|
|
|
|
# Get a list of trials
|
|
trials = analysis.trials
|
|
|
|
You may want to get a summary of multiple experiments that point to the same ``local_dir``. This is also supported by the ``ExperimentAnalysis`` class.
|
|
|
|
.. code-block:: python
|
|
|
|
from ray.tune import ExperimentAnalysis
|
|
analysis = ExperimentAnalysis("~/ray_results/example-experiment")
|
|
|
|
.. _exp-analysis-docstring:
|
|
|
|
ExperimentAnalysis (tune.ExperimentAnalysis)
|
|
--------------------------------------------
|
|
|
|
.. autoclass:: ray.tune.ExperimentAnalysis
|
|
:members:
|
|
|
|
TrialCheckpoint (tune.cloud.TrialCheckpoint)
|
|
--------------------------------------------
|
|
|
|
.. autoclass:: ray.tune.cloud.TrialCheckpoint
|
|
:members: |