Since ``nni v2.0``, we provide a new way to launch experiments. Before that, you need to configure the experiment in the yaml configuration file and then use the experiment ``nnictl`` command to launch the experiment. Now, you can also configure and run experiments directly in python file. If you are familiar with python programming, this will undoubtedly bring you more convenience.
Since ``v2.0``, NNI provides a new way to launch the experiments. Before that, you need to configure the experiment in the YAML configuration file and then use the ``nnictl`` command to launch the experiment. Now, you can also configure and run experiments directly in the Python file. If you are familiar with Python programming, this will undoubtedly bring you more convenience.
Run a New Experiment
Run a New Experiment
--------------------
--------------------
After successfully installing ``nni``, you can start the experiment with a python script in the following 2 steps.
..
After successfully installing ``nni`` and prepare the `trial code <../TrialExample/Trials.rst>`__, you can start the experiment with a Python script in the following 2 steps.
Step 1 - Initialize an experiment instance and configure it
Step 1 - Initialize an experiment instance and configure it
@@ -44,11 +45,11 @@ Use the form like ``experiment.config.foo = 'bar'`` to configure your experiment
...
@@ -44,11 +45,11 @@ Use the form like ``experiment.config.foo = 'bar'`` to configure your experiment
See all real `builtin tuners <../builtin_tuner.rst>`__ supported in NNI.
See all real `builtin tuners <../builtin_tuner.rst>`__ supported in NNI.
See `parameter configuration <../reference/experiment_config.rst>`__ required by different training services.
See `configuration reference <../reference/experiment_config.rst>`__ for more detailed usage of these fields.
..
Step 2 - Just run
Step 2 - Just run
^^^^^^^^^^^^^^^^^
.. code-block:: python
.. code-block:: python
...
@@ -56,11 +57,14 @@ See `parameter configuration <../reference/experiment_config.rst>`__ required by
...
@@ -56,11 +57,14 @@ See `parameter configuration <../reference/experiment_config.rst>`__ required by
Now, you have successfully launched an NNI experiment. And you can type ``localhost:8080`` in your browser to observe your experiment in real time.
Now, you have successfully launched an NNI experiment. And you can type ``localhost:8080`` in your browser to observe your experiment in real time.
.. Note:: In this way, experiment will run in the foreground and will automatically exit when the experiment finished. If you want to run an experiment in an interactive way, use ``start()`` in Step 2.
In this way, experiment will run in the foreground and will automatically exit when the experiment finished.
.. Note:: If you want to run an experiment in an interactive way, use ``start()`` in Step 2. If you launch the experiment in Python script, please use ``run()``, as ``start()`` is designed for the interactive scenarios.
Example
Example
^^^^^^^
^^^^^^^
Below is an example for this new launching approach. You can also find this code in :githublink:`mnist-tfv2/launch.py <examples/trials/mnist-tfv2/launch.py>`.
Below is an example for this new launching approach. You can find this code in :githublink:`mnist-tfv2/launch.py <examples/trials/mnist-tfv2/launch.py>`.
.. code-block:: python
.. code-block:: python
...
@@ -89,18 +93,21 @@ Below is an example for this new launching approach. You can also find this code
...
@@ -89,18 +93,21 @@ Below is an example for this new launching approach. You can also find this code
experiment.run(8080)
experiment.run(8080)
Start and Manage a New Experiment
Start and Manage a New Experiment
---------------------------------
---------------------------------
We migrate the API in ``NNI Client`` to this new launching approach.
Launch the experiment by ``start()`` instead of ``run()``, then you can use these APIs in interactive mode.
NNI migrates the API in ``NNI Client`` to this new launching approach. Launch the experiment by ``start()`` instead of ``run()``, then you can use these APIs in interactive mode.
Please refer to `example usage <./python_api_start.rst>`__ and code file :githublink:`python_api_start.ipynb <examples/trials/sklearn/classification/python_api_start.ipynb>`.
Please refer to `example usage <./python_api_start.rst>`__ and code file :githublink:`python_api_start.ipynb <examples/trials/sklearn/classification/python_api_start.ipynb>`.
.. Note:: ``run()`` polls the experiment status and will automatically call ``stop()`` when the experiment finished. ``start()`` just launched a new experiment, so you need to manually stop the experiment by calling ``stop()``.
.. Note:: ``run()`` polls the experiment status and will automatically call ``stop()`` when the experiment finished. ``start()`` just launched a new experiment, so you need to manually stop the experiment by calling ``stop()``.
Connect and Manage an Exist Experiment
Connect and Manage an Exist Experiment
--------------------------------------
--------------------------------------
If you launch the experiment by ``nnictl`` and also want to use these APIs, you can use ``Experiment.connect()`` to connect to an existing experiment.
If you launch an experiment by ``nnictl`` and also want to use these APIs, you can use ``Experiment.connect()`` to connect to an existing experiment.
Please refer to `example usage <./python_api_connect.rst>`__ and code file :githublink:`python_api_connect.ipynb <examples/trials/sklearn/classification/python_api_connect.ipynb>`.
Please refer to `example usage <./python_api_connect.rst>`__ and code file :githublink:`python_api_connect.ipynb <examples/trials/sklearn/classification/python_api_connect.ipynb>`.
...
@@ -110,10 +117,203 @@ Resume/View and Manage a Stopped Experiment
...
@@ -110,10 +117,203 @@ Resume/View and Manage a Stopped Experiment
-------------------------------------------
-------------------------------------------
You can use ``Experiment.resume()`` and ``Experiment.view()`` to resume and view a stopped experiment, these functions behave like ``nnictl resume`` and ``nnictl view``.
You can use ``Experiment.resume()`` and ``Experiment.view()`` to resume and view a stopped experiment, these functions behave like ``nnictl resume`` and ``nnictl view``.
If you want to manage the experiment, set ``wait_completion`` as ``False`` and the functions will return an ``Experiment`` instance. For more parameters, please refer to API.
API
If you want to manage the experiment, set ``wait_completion`` as ``False`` and the functions will return an ``Experiment`` instance. For more parameters, please refer to API reference.
---
API Reference
-------------
Detailed usage could be found `here <../reference/experiment_config.rst>`__.
* `Experiment`_
* `Experiment Config <#Experiment-Config>`_
* `Algorithm Config <#Algorithm-Config>`_
* `Training Service Config <#Training-Service-Config>`_
