Spaces:
Sleeping
Sleeping
File size: 4,261 Bytes
aaf5028 5a4a6ea aaf5028 8390c1e b70e400 8390c1e 5a4a6ea 8390c1e 09f98e2 8390c1e d9c644f b70e400 d9c644f b70e400 5a4a6ea 571fd05 5a4a6ea d9c644f b70e400 8390c1e b70e400 d9c644f b70e400 d9c644f b70e400 d9c644f b70e400 8390c1e aaf5028 b70e400 5a4a6ea b70e400 09f98e2 | 1 2 3 4 5 6 7 8 9 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 | import os
import pandas as pd
from sklearn.metrics import mean_squared_error, mean_absolute_error, r2_score
from urllib.parse import urlparse
import mlflow
import mlflow.sklearn
import numpy as np
import joblib
from mlProject.entity.config_entity import ModelEvaluationConfig
from pathlib import Path
from mlProject.utils.common import save_json
from mlflow.models import ModelSignature
from mlflow.types import Schema, ColSpec
class ModelEvaluation:
def __init__(self, config: ModelEvaluationConfig):
self.config = config
def eval_metrics(self, actual, pred):
rmse = np.sqrt(mean_squared_error(actual, pred))
mae = mean_absolute_error(actual, pred)
r2 = r2_score(actual, pred)
return rmse, mae, r2
def log_into_mlflow(self):
experiment_name = "Elasticnet" # Set your experiment name
# Set the MLflow tracking URI (if needed)
mlflow.set_tracking_uri(self.config.mlflow_uri)
# mlflow.set_registry_uri(self.config.mlflow_uri)
# Set authentication credentials
mlflow_password= os.getenv("MLFLOW_TRACKING_PASSWORD")
username = "priteshptadvi29"
os.environ["MLFLOW_TRACKING_USERNAME"] = username
os.environ["MLFLOW_TRACKING_PASSWORD"] = mlflow_password
# Check if the experiment exists
experiment = mlflow.get_experiment_by_name(experiment_name)
if experiment is None:
# Create a new experiment
experiment_id = mlflow.create_experiment(experiment_name)
print(f"Created new experiment: {experiment_name} with ID {experiment_id}")
else:
print(f"Using existing experiment: {experiment_name} with ID {experiment.experiment_id}")
# Set the experiment for logging
mlflow.set_experiment(experiment_name)
test_data = pd.read_csv(self.config.test_data_path)
model = joblib.load(self.config.model_path)
test_x = test_data.drop([self.config.target_column], axis=1)
test_y = test_data[[self.config.target_column]]
# Create an input example (this should match your model's input shape)
input_example = test_data.drop([self.config.target_column], axis=1).iloc[0:1] # Take the first row as an example
# Define the model signature
input_schema = Schema([
ColSpec("float", "fixed acidity"), # Replace with your actual feature names and types
ColSpec("float", "volatile acidity"),
ColSpec("float", "citric acid"),
ColSpec("float", "residual sugar"),
ColSpec("float", "chlorides"),
ColSpec("float", "free sulfur dioxide"),
ColSpec("float", "total sulfur dioxide"),
ColSpec("float", "density"),
ColSpec("float", "pH"),
ColSpec("float", "sulphates"),
ColSpec("float", "alcohol"),
])
# Create the model signature
signature = ModelSignature(inputs=input_schema)
with mlflow.start_run():
# Rest of the logging code
predicted_qualities = model.predict(test_x)
(rmse, mae, r2) = self.eval_metrics(test_y, predicted_qualities)
# Saving metrics as local
scores = {"rmse": rmse, "mae": mae, "r2": r2}
save_json(path = Path(self.config.metric_file_name), data=scores)
mlflow.log_params(self.config.all_params)
mlflow.log_metric("rmse", rmse)
mlflow.log_metric("mae", mae)
mlflow.log_metric("r2", r2)
# Model registry does not work with file store
tracking_url_type_store = urlparse(mlflow.get_tracking_uri()).scheme
if tracking_url_type_store != "file":
# Register the model
mlflow.sklearn.log_model(model,
"model",
registered_model_name = "ElasticnetModel",
signature=signature,
input_example=input_example.to_dict(orient="records")[0]
)
else:
mlflow.sklearn.log_model(model, "model") |