10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170 | class BaseTracker:
"""
Basic Tracker object that don't have any requirements nor parameters to be used
"""
def __init__(self, tracking_uri:str, username: str = None, password: str = None)-> None:
"""
Initialize the tracker object
Parameters:
tracking_uri: URI to the MLflow server
username: Username to access the MLflow server
password: Password to access the MLflow server
Returns:
None
"""
mlflow.set_tracking_uri(tracking_uri)
os.environ['MLFLOW_TRACKING_USERNAME'] = username
os.environ['MLFLOW_TRACKING_PASSWORD'] = password
# Create an MLflow client
self.client = mlflow.tracking.MlflowClient()
def log_model(self, sk_model: Any, artifact_path: Any, extra_pip_requirements: Any | None = None,**kwargs):
"""
Log a model to MLflow.
Parameters:
sk_model: scikit-learn model to be saved
artifact_path: name of the model
extra_pip_requirements: additional pip requirements to be installed
Returns:
None
"""
mlflow.sklearn.log_model(sk_model, artifact_path, extra_pip_requirements,**kwargs)
def log_params(self, params: Dict[str, Any], **kwargs) -> None:
"""
Log parameters to MLflow.
Parameters:
params: Dict with the parameters to log.
Returns:
None
"""
mlflow.log_params(params, **kwargs)
def log_metrics(self, params: Dict[str, float], **kwargs) -> None:
"""
Log metrics to MLflow.
Parameters:
params: Dict with the metrics to log.
Returns:
None
"""
mlflow.log_metrics(params, **kwargs)
def log_artifact(self, local_path: str, artifact_path: str | None = None, **kwargs) -> None:
"""
Log an artifact to MLflow.
Parameters:
local_path: Path to the file to write.
artifact_path: If provided, the directory in artifact_uri to write to.
Returns:
None
"""
mlflow.log_artifact(local_path, artifact_path, **kwargs)
def set_project(self, project_name: str)-> None:
"""
Check if the experiment already exists; if not, create it
Parameters:
project_name (str): The name of the project.
Returns:
None
"""
project = mlflow.get_experiment_by_name(
project_name)
if project is None:
mlflow.create_experiment(name=project_name)
self.project_name = project_name
print(f"Experiment {project_name} created")
else:
# Activate the experiment for tracking
mlflow.set_experiment(project_name)
self.project_name = project_name
print(f"Experiment {project_name} already exists")
def save_experiment(self):
"""
Start a run in MLflow.
Parameters:
**kwargs: The parameters to log.
Returns:
None
"""
print("Saving experiment", self.project_name)
runs = mlflow.search_runs(experiment_names=[
self.project_name], order_by=["start_time desc"])
next_run_number = len(runs) + 1
active_run = mlflow.start_run(run_name=f"{self.project_name}-{next_run_number}")
return active_run
def get_model_run_id(self, model_name: str, stage: str = "Production"):
"""
Get the run_id of a model based on its name and stage.
Parameters:
model_name (str): The name of the model.
stage (str): The stage of the model.
Returns:
str: The run_id of the model.
"""
latest_versions = self.client.get_latest_versions(
name=model_name, stages=[stage])
run_id = latest_versions[0].source.split("/")
return run_id[2]
def get_model_experiment_id(self, model_name: str):
"""
Get the project name associated with a model.
Parameters:
model_name (str): The name of the model.
Returns:
str: The project name.
"""
latest_production_id = self.get_model_run_id(model_name=model_name, stage="Production")
run_info = mlflow.get_run(latest_production_id)
return run_info.info.experiment_id
def get_run_name(self, run_id: str):
"""
Get the run name associated with a run ID.
Parameters:
run_id (str): Run ID.
Returns:
str: Run name.
"""
if run_id:
run_info = mlflow.get_run(run_id)
run_name = run_info.info.run_name
else:
run_name = None
return run_name
|