def _save_model(
model: BaseEstimator,
machine: Union[Machine, dict],
output_dir: Union[os.PathLike, str],
):
"""
Save the model according to the expected Argo workflow procedure.
Parameters
----------
model: BaseEstimator
The model to save to the directory with gordo serializer.
machine: Union[Machine, dict]
Machine instance used to build this model.
output_dir: Union[os.PathLike, str]
The directory where to save the model, will create directories if needed.
Returns
-------
Union[os.PathLike, str]
Path to the saved model
"""
os.makedirs(output_dir, exist_ok=True) # Ok if some dirs exist
serializer.dump(
model,
output_dir,
metadata=machine.to_dict()
if isinstance(machine, Machine) else machine,
)
return output_dir
def test_pipeline_serialization(self):
pipe = Pipeline(
[
("pca1", PCA(n_components=10)),
(
"fu",
FeatureUnion(
[
("pca2", PCA(n_components=3)),
(
"pipe",
Pipeline(
[
("minmax", MinMaxScaler()),
("truncsvd", TruncatedSVD(n_components=7)),
]
),
),
]
),
),
("ae", KerasAutoEncoder(kind="feedforward_hourglass")),
]
)
X = np.random.random(size=100).reshape(10, 10)
pipe.fit(X.copy(), X.copy())
with TemporaryDirectory() as tmp:
# Test dump
metadata = {"key": "value"}
serializer.dump(pipe, tmp, metadata=metadata)
# Test load from the serialized pipeline above
pipe_clone = serializer.load(tmp)
metadata_clone = serializer.load_metadata(tmp)
# Ensure the metadata was saved and loaded back
self.assertEqual(metadata, metadata_clone)
# Verify same state for both pipelines
y_hat_pipe1 = pipe.predict(X.copy()).flatten()
y_hat_pipe2 = pipe_clone.predict(X.copy()).flatten()
self.assertTrue(np.allclose(y_hat_pipe1, y_hat_pipe2))
# Now use dumps/loads
serialized = serializer.dumps(pipe)
pipe_clone = serializer.loads(serialized)
# Verify same state for both pipelines
y_hat_pipe1 = pipe.predict(X.copy()).flatten()
y_hat_pipe2 = pipe_clone.predict(X.copy()).flatten()
self.assertTrue(np.allclose(y_hat_pipe1, y_hat_pipe2))
def test_dump_load_models(model):
X = np.random.random(size=100).reshape(10, 10)
model.fit(X.copy(), X.copy())
model_out = model.predict(X.copy())
with TemporaryDirectory() as tmp:
serializer.dump(model, tmp)
model_clone = serializer.load(tmp)
model_clone_out = model_clone.predict(X.copy())
assert np.allclose(model_out.flatten(), model_clone_out.flatten())
def trained_model_directory(
model_collection_directory: str, config_str: str, gordo_name: str
):
"""
Fixture: Train a basic AutoEncoder and save it to a given directory
will also save some metadata with the model
"""
# Model specific to the model being trained here
model_dir = os.path.join(model_collection_directory, gordo_name)
os.makedirs(model_dir, exist_ok=True)
builder = local_build(config_str=config_str)
model, metadata = next(builder) # type: ignore
serializer.dump(model, model_dir, metadata=metadata.to_dict())
yield model_dir
def trained_model_directories(model_collection_directory: str,
config_str: str):
"""
Fixture: Train a basic AutoEncoder and save it to a given directory
will also save some metadata with the model
"""
# Model specific to the model being trained here
builder = local_build(config_str=config_str)
model_directories = {}
for model, metadata in builder:
metadata_dict = metadata.to_dict()
model_name = metadata_dict.get("name")
model_dir = os.path.join(model_collection_directory, model_name)
os.makedirs(model_dir, exist_ok=True)
serializer.dump(model, model_dir, metadata=metadata.to_dict())
model_directories[model_name] = model_dir
yield model_directories
def download_model(ctx: click.Context, output_dir: str, target: typing.List[str]):
"""
Download the actual model from the target and write to an output directory
"""
client = Client(*ctx.obj["args"], **ctx.obj["kwargs"])
models = client.download_model(targets=target)
# Iterate over mapping of models and save into their own sub dirs of the output_dir
for model_name, model in models.items():
model_out_dir = os.path.join(output_dir, model_name)
os.mkdir(model_out_dir)
click.secho(
f"Writing model '{model_name}' to directory: '{model_out_dir}'...", nl=False
)
serializer.dump(model, model_out_dir)
click.secho(f"done")
click.secho(f"Wrote all models to directory: {output_dir}", fg="green")