Trial API¶
Model Definition Interfaces¶
To create a Trial API model definition, we should implement the Trial interface provided by Determined. This interface returns information about the machine learning task the user wants to perform, like the model architecture to use or the validation metrics that should be computed.
Determined provides versions of the Trial interface for each of the application frameworks it supports:
Create an Experiment via det.experimental.create()
¶
A user can submit an experiment from Python by executing the
determined.experimental.create()
function:
class MyTrial(...):
...
det.experimental.create(trial_def=MyTrial, context_dir=".")
In addition to a trial class definition, the create()
API requires a
context directory (context_dir
). The context directory specifies the root
directory of the code containing the trial implementation – for a majority of
users this is the current working directory (.). The create()
API also
accepts two boolean keyword arguments:
local
(bool
):local=False
will sumbit the experiment to a Determined cluster.local=True
will execute the the training loop in your local Python environment (although currently, local training is not implemented, so you must also settest=True
). Defaults to False.test
(bool
):test=True
will execute a minimal trianing loop rather than a full experiment. This can be useful for porting or debugging a model because many common errors will surface quickly. Defaults to False.
Create an Experiment via the CLI¶
A user can submit an experiment via the det experiment create
CLI command:
$ det experiment create <YAML config file> <context directory>
The context directory of Python files that contain the Trial API implementation
should include an accompanying entrypoint
that specifies from where to load
a trial class. The entrypoint specification is expected to take the form:
<module>:<object reference>
<module>
specifies the module containing the trial class within the model
definition, relative to the root. It may be an empty string if the model
definition is a Python package and the trial
class is exposed in the top-level __init__.py
file.
<object reference>
specifies the naming of the trial class within the
module. It may be a nested object delimited by dots.
Examples:
:MNistTrial
expects anMNistTrial
class that is exposed in a__init__.py
file at the top level of the model definition.model_def:CIFAR10Trial
expects aCIFAR10Trial
class that is defined in a filemodel_def.py
at the top level of the model definition.determined_lib.trial:trial_classes.NestedTrial
expects aNestedTrial
class that is an attribute oftrial_classes
, wheretrial_classes
is defined in a filedetermined_lib/trial.py
.
Note that this follows the Entry points specification defined in the Python
Packaging User Guide with a single
difference: the directory name of the model definition is prefixed to
<module>
, or used as the module if <module>
is empty.
Since project directories might include large artifacts that should not
be packaged as part of the model definition (e.g., data sets or compiled
binaries), users can optionally include a .detignore
file at the
top level that specifies file paths to be omitted from the model
definition. The .detignore
file uses the same syntax as
.gitignore. Note that
byte-compiled Python files (e.g., .pyc
files or __pycache__
directories) are always ignored.