wassname/ray
1
0
Fork 0
mirror of https://github.com/wassname/ray.git synced 2026-09-12 12:51:15 +08:00
Code Issues Packages Projects Releases Wiki Activity

[tune] HyperOpt Support (v2) (#1763)

Browse Source
This commit is contained in:
Richard Liaw
2018-04-04 11:08:26 -07:00
committed by GitHub
parent 5a9e83761d
commit 888e70f1be
13 changed files with 348 additions and 44 deletions
Download Patch File Download Diff File Expand all files Collapse all files
doc/source/tune.rst
+18 -2
View File
@@ -62,7 +62,7 @@ Features
Ray Tune has the following features:
- Scalable implementations of search algorithms such as `Population Based Training (PBT) <pbt.html>`__, `Median Stopping Rule <hyperband.html#median-stopping-rule>`__, and `HyperBand <hyperband.html>`__.
- Scalable implementations of search algorithms such as `Population Based Training (PBT) <pbt.html>`__, `Median Stopping Rule <hyperband.html#median-stopping-rule>`__, Model-Based Optimization (HyperOpt), and `HyperBand <hyperband.html>`__.
- Integration with visualization tools such as `TensorBoard <https://www.tensorflow.org/get_started/summaries_and_tensorboard>`__, `rllab's VisKit <https://media.readthedocs.org/pdf/rllab/latest/rllab.pdf>`__, and a `parallel coordinates visualization <https://en.wikipedia.org/wiki/Parallel_coordinates>`__.
@@ -94,12 +94,28 @@ You can find the code for Ray Tune `here on GitHub <https://github.com/ray-proje
Trial Schedulers
----------------
By default, Ray Tune schedules trials in serial order with the ``FIFOScheduler`` class. However, you can also specify a custom scheduling algorithm that can early stop trials, perturb parameters, or incorporate suggestions from an external service. Currently implemented trial schedulers include `Population Based Training (PBT) <pbt.html>`__, `Median Stopping Rule <hyperband.html#median-stopping-rule>`__, and `HyperBand <hyperband.html>`__.
By default, Ray Tune schedules trials in serial order with the ``FIFOScheduler`` class. However, you can also specify a custom scheduling algorithm that can early stop trials, perturb parameters, or incorporate suggestions from an external service. Currently implemented trial schedulers include `Population Based Training (PBT) <pbt.html>`__, `Median Stopping Rule <hyperband.html#median-stopping-rule>`__, Model-Based Optimization (HyperOpt), and `HyperBand <hyperband.html>`__.
.. code-block:: python
run_experiments({...}, scheduler=AsyncHyperBandScheduler())
HyperOpt Integration
--------------------
The``HyperOptScheduler`` is a Trial Scheduler that is backed by HyperOpt to perform sequential model-based hyperparameter optimization.
In order to use this scheduler, you will need to install HyperOpt via the following command:
.. code-block:: bash
$ pip install --upgrade git+git://github.com/hyperopt/hyperopt.git
An example of this can be found in `hyperopt_example.py <https://github.com/ray-project/ray/blob/master/python/ray/tune/examples/hyperopt_example.py>`__.
.. autoclass:: ray.tune.hpo_scheduler.HyperOptScheduler
Visualizing Results
-------------------