Ajouter le serveur de suivi MLflow et l'application Streamlit

Dockerfile

@@ -13,7 +13,7 @@ ENV HOME=/home/user \
 WORKDIR $HOME/app
 
 # Install basic dependencies
-RUN pip install boto3 pandas gunicorn mlflow streamlit scikit-learn matplotlib seaborn plotly
+RUN pip install boto3 pandas gunicorn mlflow streamlit scikit-learn matplotlib seaborn plotly openpyxl
 
 COPY --chown=user . $HOME/app
 
@@ -22,4 +22,8 @@ RUN pip install -r /dependencies/requirements.txt
 
 COPY . $HOME/app
 
-CMD ["streamlit", "run", "app.py", "--server.port", "7860", "--server.address", "0.0.0.0"]
+# Exposer les ports pour Streamlit et MLflow
+EXPOSE 8000 5000
+
+# Lancer Streamlit et le serveur de suivi MLflow
+CMD ["sh", "-c", "streamlit run app.py --server.port 8000 --server.address 0.0.0.0 & python mlflow_app.py"]

mlflow_app.py

@@ -0,0 +1,5 @@
+import mlflow
+import os
+
+if __name__ == "__main__":
+    os.system("mlflow server --host 0.0.0.0 --port 5000")

train.py

@@ -12,6 +12,8 @@ from sklearn.compose import ColumnTransformer
 from sklearn.linear_model import LinearRegression
 from sklearn.metrics import mean_squared_error
 
+# Définir l'URI de suivi MLflow
+mlflow.set_tracking_uri("http://0.0.0.0:5000")
 
 if __name__ == "__main__":
 
@@ -79,7 +81,7 @@ if __name__ == "__main__":
         run_id = run.info.run_id
         with open("run_id.txt", "w") as f:
             f.write(run_id)
-            
+
     print("...Done!")
     print("Saving model...")
     mlflow.sklearn.log_model(model, "model", signature=infer_signature(X_train, predictions))