使用MLFlow和Sklearn记录模型时多个伪影路径

发布于 2025-02-09 07:39:52 字数 609 浏览 1 评论 0原文

我正在使用MLFlow来记录logistic回归的参数和工件，但是当我尝试记录模型时，我可以在mlflow UI中看到所有文件时，我会看到两个文件夹：一个文件夹：一个名为“模型”，另一个名为命名的文件夹“ logger”（我设置的那个）。

model = LogisticRegression()

mlflow.set_tracking_uri('file:///artifacts')
mlflow.set_experiment('test')
mlflow.autolog()

with mlflow.start_run(run_name=run_name) as run:
   model.train(X_train, y_train)
   mlflow.sklearn.log_model(model, 'logreg')

不确定我是否缺少某些东西或是否有配置。

希望有人可以帮助我！

原文

I'm using mlflow to log parameters and artifacts of a Logistic Regression, but when I try to log the model so I can see all the files in the Mlflow UI, I see two folders: one named 'model' and the other one named 'logger' (the one I set).

model = LogisticRegression()

mlflow.set_tracking_uri('file:///artifacts')
mlflow.set_experiment('test')
mlflow.autolog()

with mlflow.start_run(run_name=run_name) as run:
   model.train(X_train, y_train)
   mlflow.sklearn.log_model(model, 'logreg')