lab8

Jenkinsfile_artifact

@@ -0,0 +1,23 @@
+pipeline {
+    agent {
+        dockerfile {
+            args '-v /mlruns:/mlruns'
+        }
+    }
+    parameters {
+        buildSelector(
+			defaultSelector: lastSuccessful(),
+			description: 'select build from which to copy artifacts',
+			name: 'BUILD_SELECTOR'
+		)
+    }
+    stages {
+        stage('Copy artifacts') {
+            steps {
+                copyArtifacts filter: 'mlruns.tar.gz', projectName: 's449288-training/master', selector: buildParameter('BUILD_SELECTOR')
+                sh 'mkdir -p mlrunsArtifact && tar xzf mlruns.tar.gz -C mlrunsArtifact --strip-components 1'
+                sh "python ./predictArtifact.py"
+            }
+        }
+    }
+}

Jenkinsfile_mlflow

@@ -0,0 +1,15 @@
+pipeline {
+    agent {
+        dockerfile {
+            args '-v /mlruns:/mlruns'
+        }
+    }    
+    stages {
+        stage('Prediction') {
+            steps {
+                sh 'ls -la'
+                sh "python ./predictMlflow.py"
+            }
+        }
+    }
+}

MLproject

@@ -0,0 +1,13 @@
+name: tutorial
+# conda_env: conda.yaml 
+    
+docker_env:
+  image: ksero/ium:mlflow
+
+entry_points:
+  main:
+    parameters:
+      epochs: {type: float, default: 1500}
+    command: "python ./pytorch/pytorch.py {epochs}"
+  test:
+    command: "python ./evaluation.py"

predictArtifact.py

@@ -0,0 +1,14 @@
+import mlflow
+import numpy as np
+import json
+
+logged_model = 'mlrunsArtifact/1/6b2323cf51794581bf1e2f6d060d50f6/artifacts/model'
+loaded_model = mlflow.pyfunc.load_model(logged_model)
+
+with open(f'{logged_model}/input_example.json') as f:
+    input_example_data = json.load(f)
+
+input_example = np.array(input_example_data['inputs']).reshape(-1,)
+
+print(f'Input: {input_example}')
+print(f'Prediction: {loaded_model.predict(input_example)}')

predictMlflow.py

@@ -0,0 +1,14 @@
+import mlflow
+import numpy as np
+import json
+
+logged_model = '/mlruns/20/80fe21a0804844088147d15a3cebb3e5/artifacts/lego'
+loaded_model = mlflow.pyfunc.load_model(logged_model)
+
+with open(f'{logged_model}/input_example.json') as f:
+    input_example_data = json.load(f)
+
+input_example = np.array(input_example_data['inputs']).reshape(-1,)
+
+print(f'Input: {input_example}')
+print(f'Prediction: {loaded_model.predict(input_example)}')

pytorch/pytorch.py

@@ -21,6 +21,8 @@ import sys
 from sacred import Experiment
 from sacred.observers import FileStorageObserver
 from sacred.observers import MongoObserver
+import mlflow
+import mlflow.keras
 
 # In[2]:
 ex = Experiment(save_git_info=False)
@@ -30,7 +32,8 @@ ex.observers.append(FileStorageObserver('my_runs'))
 # ex.observers.append(MongoObserver(url='mongodb://mongo_user:mongo_password_IUM_2021@localhost:27017', db_name='sacred'))
 ex.observers.append(MongoObserver(url='mongodb://admin:IUM_2021@172.17.0.1:27017', db_name='sacred'))
 
-
+mlflow.set_tracking_uri("http://172.17.0.1:5000")
+mlflow.set_experiment('s444354')
 
 try: 
     numberOfEpochParam = int(sys.argv[1])
@@ -162,21 +165,39 @@ def fit(epochs, lr, model, train_loader, val_loader, _log, _run, opt_func=torch.
     epochs=epochs
     history = []
     optimizer = opt_func(model.parameters(), lr)
-    for epoch in range(epochs):
-        for batch in train_loader:
-            loss = model.training_step(batch)
-            loss.backward()
-            optimizer.step()
-            optimizer.zero_grad()
-        result = evaluate(model, val_loader)
-        model.epoch_end(epoch, result, epochs)
-        history.append(result)
+    with mlflow.start_run():
+    
+        for epoch in range(epochs):
+            for batch in train_loader:
+                loss = model.training_step(batch)
+                loss.backward()
+                optimizer.step()
+                optimizer.zero_grad()
+            result = evaluate(model, val_loader)
+            model.epoch_end(epoch, result, epochs)
+            history.append(result)
+            
+        mlflow.log_param('epochs', epochs)
+        mlflow.log_param('lr', lr)
     
     torch.save(model, 'saved_model.pth')
     ex.add_artifact("saved_model.pth")
     
     _run.info["epochs"] = epochs
     
+    signature = mlflow.models.signature.infer_signature(house_price_features, linear_model.predict(house_price_features))
+  
+    tracking_url_type_store = urlparse(mlflow.get_tracking_uri()).scheme
+  
+    sampleInp = [0.1, 0.1, 546.0, 555.223, 1., 1., 33.16376, 84.12426]
+  
+    if tracking_url_type_store != "file":
+        mlflow.keras.log_model(model, "model", registered_model_name="red-wine-quality", signature=signature)
+    else:
+        mlflow.keras.log_model(model, "model", signature=signature, input_example=np.array(sampleInp))
+    
+    
+  
     return history