Updated all syntax to 0.55.0

customer-satisfaction/config.yaml

@@ -6,7 +6,7 @@ settings:
       - mlflow
 
 # configuration of the Model Control Plane
-model_version:
+model:
   name: Customer_Satisfaction_Predictor
   license: Apache 2.0
   description: Predictor of Customer Satisfaction.

customer-satisfaction/pipelines/deployment_pipeline.py

@@ -2,7 +2,7 @@
 
 from zenml.integrations.mlflow.steps import mlflow_model_deployer_step
 
-from zenml import pipeline, ModelVersion
+from zenml import pipeline
 
 from pipelines.training_pipeline import customer_satisfaction_training_pipeline
 from steps import predictor

customer-satisfaction/steps/evaluation.py

@@ -45,7 +45,7 @@ def evaluation(
         mlflow.log_metric("rmse", rmse)
 
         # Also add the metrics to the Model within the ZenML Model Control Plane
-        artifact = get_step_context().model_version.get_artifact("sklearn_regressor")
+        artifact = get_step_context().model.get_artifact("sklearn_regressor")
 
         log_artifact_metadata(
             metadata={

customer-satisfaction/steps/model_loader.py

@@ -1,5 +1,5 @@
 from sklearn.base import RegressorMixin
-from zenml import step, ModelVersion
+from zenml import step, Model
 from zenml.client import Client
 
 
@@ -12,9 +12,9 @@ def model_loader(
     Args:
         model_name: Name of the Model to load
     """
-    model_version = ModelVersion(
+    model = Model(
         name=model_name,
         version="production"
     )
-    model_artifact: RegressorMixin = model_version.load_artifact("sklearn_regressor")
+    model_artifact: RegressorMixin = model.load_artifact("sklearn_regressor")
     return model_artifact

customer-satisfaction/steps/model_promoter.py

@@ -1,4 +1,4 @@
-from zenml import get_step_context, step, ModelVersion
+from zenml import get_step_context, step, Model
 from zenml.logger import get_logger
 
 logger = get_logger(__name__)
@@ -21,37 +21,37 @@ def model_promoter(
     Returns:
         Whether the model was promoted or not.
     """
-    # Get the current model_version produced by the current pipeline
-    model_version = get_step_context().model_version
+    # Get the current model produced by the current pipeline
+    zenml_model = get_step_context().model
 
     # Get the previous model version at the production stage
-    previous_production_model = ModelVersion(
-        name=model_version.name,
+    previous_production_model = Model(
+        name=zenml_model.name,
         version="production"
     )
 
     try:
         # In case there already is a model version at the correct stage
-        previous_production_model_version_mse = float(
+        previous_production_model_mse = float(
             previous_production_model.get_artifact("sklearn_regressor").run_metadata["metrics"].value["mse"]
         )
     except RuntimeError:
         # In case no model version has been promoted before,
         #   default to a threshold value well above the new mse
-        previous_production_model_version_mse = mse + 1000
+        previous_production_model_mse = mse + 1000
 
-    if mse > previous_production_model_version_mse:
+    if mse > previous_production_model_mse:
         logger.info(
             f"Model mean-squared error {mse:.2f} is higher than"
             f" the mse of the previous production model "
-            f"{previous_production_model_version_mse:.2f} ! "
+            f"{previous_production_model_mse:.2f} ! "
             f"Not promoting model."
         )
         is_promoted = False
     else:
         logger.info(f"Model promoted to {stage}!")
         is_promoted = True
-        model_version = get_step_context().model_version
-        model_version.set_stage(stage, force=True)
+        zenml_model = get_step_context().model
+        zenml_model.set_stage(stage, force=True)
 
     return is_promoted

huggingface-sagemaker/run.ipynb

@@ -5,7 +5,7 @@
    "id": "51690802-31a7-4e6d-9f88-e6457c6c4a96",
    "metadata": {},
    "source": [
-    "# Huggingface Model to Sagemaker Endpoint: Automating MLOps with ZenML\r\n",
+    "# Huggingface Model to Sagemaker Endpoint: Automating MLOps with ZenML\n",
     "Deploying Huggingface models to AWS Sagemaker endpoints typically only requires a few lines of code. However, there's a growing demand to not just deploy, but to seamlessly automate the entire flow from training to production with comprehensive lineage tracking. ZenML adeptly fills this niche, providing an end-to-end MLOps solution for Huggingface users wishing to deploy to Sagemaker. Below, we’ll walk through the architecture that ZenML employs to bring a Huggingface model into production with AWS Sagemaker. Of course all of this can be adapted to not just Sagemaker, but any other model deployment service like GCP Vertex or Azure ML Platform.\n",
     "\n",
     "This blog post showcases one way of using ZenML pipelines to achieve this:\n",
@@ -30,16 +30,13 @@
     "import numpy as np\n",
     "from datasets import DatasetDict, load_dataset\n",
     "from typing_extensions import Annotated\n",
-    "from zenml import step\n",
+    "from zenml import step, pipeline, Model\n",
     "from zenml.logger import get_logger\n",
     "\n",
     "import os\n",
     "from typing import Optional\n",
     "from datetime import datetime as dt\n",
     "\n",
-    "from zenml import pipeline\n",
-    "from zenml.model import ModelConfig\n",
-    "\n",
     "from steps import (\n",
     "    data_loader,\n",
     "    notify_on_failure,\n",
@@ -68,9 +65,9 @@
    "metadata": {},
    "source": [
     "# 🍳Breaking it down\n",
-    "\r\n",
-    "\r\n",
-    "\r\n"
+    "\n",
+    "\n",
+    "\n"
    ]
   },
   {
@@ -250,12 +247,10 @@
     "\n",
     "# This executes all steps in the pipeline in the correct order using the orchestrator\n",
     "# stack component that is configured in your active ZenML stack.\n",
-    "model_config = ModelConfig(\n",
+    "zenml_model = Model(\n",
     "    name=zenml_model_name,\n",
     "    license=\"Apache 2.0\",\n",
     "    description=\"Show case Model Control Plane.\",\n",
-    "    create_new_model_version=True,\n",
-    "    delete_new_version_on_failure=True,\n",
     "    tags=[\"sentiment_analysis\", \"huggingface\"],\n",
     ")\n",
     "\n",
@@ -265,7 +260,7 @@
     "    pipeline_args[\"enable_cache\"] = False\n",
     "\n",
     "# Execute Feature Engineering Pipeline\n",
-    "pipeline_args[\"model_config\"] = model_config\n",
+    "pipeline_args[\"model\"] = zenml_model\n",
     "pipeline_args[\"config_path\"] = os.path.join(\"configs\", \"feature_engineering_config.yaml\")\n",
     "run_args_feature = {\n",
     "    \"max_seq_length\": max_seq_length,\n",
@@ -299,7 +294,7 @@
    "id": "78ab8771-4421-4975-a3d5-12892a56b805",
    "metadata": {},
    "source": [
-    "## 💪 Step 2: Train the model with Huggingface Hub as the model registry\r\n",
+    "## 💪 Step 2: Train the model with Huggingface Hub as the model registry\n",
     " "
    ]
   },
@@ -342,7 +337,7 @@
     "# run_args_train[\"tokenizer_artifact_id\"] = latest_run.steps['tokenizer_loader'].output.id\n",
     "\n",
     "# Configure the model\n",
-    "pipeline_args[\"model_config\"] = model_config\n",
+    "pipeline_args[\"model\"] = zenml_model\n",
     "\n",
     "pipeline_args[\n",
     "    \"run_name\"\n",
@@ -400,7 +395,7 @@
    "id": "be79f454-a45d-4f5f-aa93-330d52069124",
    "metadata": {},
    "source": [
-    "## 🫅 Step 3: Promote the model to production\r\n"
+    "## 🫅 Step 3: Promote the model to production\n"
    ]
   },
   {
@@ -431,10 +426,10 @@
    "outputs": [],
    "source": [
     "run_args_promoting = {}\n",
-    "model_config = ModelConfig(name=zenml_model_name)\n",
+    "zenml_model = Model(name=zenml_model_name)\n",
     "pipeline_args[\"config_path\"] = os.path.join(\"configs\", \"promoting_config.yaml\")\n",
     "\n",
-    "pipeline_args[\"model_config\"] = model_config\n",
+    "pipeline_args[\"model\"] = zenml_model\n",
     "\n",
     "pipeline_args[\n",
     "    \"run_name\"\n",
@@ -458,7 +453,7 @@
    "id": "6efc4968-35fd-42e3-ba62-d8e1557aa0d6",
    "metadata": {},
    "source": [
-    "## 💯 Step 4: Deploy the model to AWS Sagemaker Endpoints\r\n"
+    "## 💯 Step 4: Deploy the model to AWS Sagemaker Endpoints\n"
    ]
   },
   {
@@ -491,11 +486,11 @@
     "pipeline_args[\"config_path\"] = os.path.join(\"configs\", \"deploying_config.yaml\")\n",
     "\n",
     "# Deploying pipeline has new ZenML model config\n",
-    "model_config = ModelConfig(\n",
+    "zenml_model = Model(\n",
     "    name=zenml_model_name,\n",
     "    version=ModelStages.PRODUCTION,\n",
     ")\n",
-    "pipeline_args[\"model_config\"] = model_config\n",
+    "pipeline_args[\"model\"] = zenml_model\n",
     "pipeline_args[\"enable_cache\"] = False\n",
     "run_args_deploying = {}\n",
     "pipeline_args[\n",
@@ -520,10 +515,10 @@
    "id": "594ee4fc-f102-4b99-bdc3-2f1670c87679",
    "metadata": {},
    "source": [
-    "ZenML builds upon the straightforward deployment capability of Huggingface models to AWS Sagemaker, and transforms it into a sophisticated, repeatable, and transparent MLOps workflow. It takes charge of the intricate steps necessary for modern ML systems, ensuring that software engineering leads can focus on iteration and innovation rather than operational intricacies.\r\n",
-    "\r\n",
-    "To delve deeper into each stage, refer to the comprehensive guide on GitHub[: zenml-io/zenml-huggingface-sagemak](https://github.com/zenml-io/zenml-huggingface-sagemaker)er. Additionally[, this YouTube playli](https://www.youtube.com/watch?v=Q1EH2H8Akgo&list=PLhNrLW_IWplw6dBbmGcL828-atJMu3CwF)st provides a detailed visual walkthrough of the entire pipeline: Huggingface to Sagemaker ZenML tutorial.\r\n",
-    "\r\n",
+    "ZenML builds upon the straightforward deployment capability of Huggingface models to AWS Sagemaker, and transforms it into a sophisticated, repeatable, and transparent MLOps workflow. It takes charge of the intricate steps necessary for modern ML systems, ensuring that software engineering leads can focus on iteration and innovation rather than operational intricacies.\n",
+    "\n",
+    "To delve deeper into each stage, refer to the comprehensive guide on GitHub[: zenml-io/zenml-huggingface-sagemak](https://github.com/zenml-io/zenml-huggingface-sagemaker)er. Additionally[, this YouTube playli](https://www.youtube.com/watch?v=Q1EH2H8Akgo&list=PLhNrLW_IWplw6dBbmGcL828-atJMu3CwF)st provides a detailed visual walkthrough of the entire pipeline: Huggingface to Sagemaker ZenML tutorial.\n",
+    "\n",
     "Interested in standardizing your MLOps workflows? ZenML Cloud is now available to all - get a managed ZenML server with important features such as RBAC and pipeline trigge[rs. Book a ](https://zenml.io/book-a-demo)demo with us now to learn how you can create your own MLOps pipelines today."
    ]
   }

huggingface-sagemaker/run.py

@@ -23,7 +23,7 @@
 from zenml.client import Client
 from zenml.enums import ModelStages
 from zenml.logger import get_logger
-from zenml.model.model_version import ModelVersion
+from zenml import Model
 
 from pipelines import (
     sentinment_analysis_deploy_pipeline,
@@ -200,12 +200,10 @@ def main(
             os.path.dirname(os.path.realpath(__file__)),
             "configs",
     )
-    model_version = ModelVersion(
+    zenml_model = Model(
         name=zenml_model_name,
         license="Apache 2.0",
         description="Show case Model Control Plane.",
-        create_new_model_version=True,
-        delete_new_version_on_failure=True,
         tags=["sentiment_analysis", "huggingface"],
     )
 
@@ -216,7 +214,7 @@ def main(
 
     # Execute Feature Engineering Pipeline
     if feature_pipeline:
-        pipeline_args["model_version"] = model_version
+        pipeline_args["model"] = zenml_model
         pipeline_args["config_path"] = os.path.join(config_folder, "feature_engineering_config.yaml")
         run_args_feature = {
             "max_seq_length": max_seq_length,
@@ -259,7 +257,7 @@ def main(
             run_args_train["dataset_artifact_id"] = tokenized_dataset_artifact.id
             run_args_train["tokenizer_artifact_id"] = tokenized_tokenizer_artifact.id
 
-        pipeline_args["model_version"] = model_version
+        pipeline_args["model"] = zenml_model
 
         pipeline_args[
             "run_name"
@@ -274,13 +272,13 @@ def main(
     if promoting_pipeline:
         run_args_promoting = {}
         # Promoting pipeline always check latest version
-        model_version = ModelVersion(
+        zenml_model = Model(
             name=zenml_model_name,
             version=ModelStages.LATEST,
         )
         pipeline_args["config_path"] = os.path.join(config_folder, "promoting_config.yaml")
 
-        pipeline_args["model_version"] = model_version
+        pipeline_args["model"] = zenml_model
 
         pipeline_args[
             "run_name"
@@ -294,11 +292,11 @@ def main(
         pipeline_args["config_path"] = os.path.join(config_folder, "deploying_config.yaml")
 
         # Deploying pipeline has new ZenML model config
-        model_version = ModelVersion(
+        zenml_model = Model(
             name=zenml_model_name,
             version=ModelStages.PRODUCTION,
         )
-        pipeline_args["model_version"] = model_version
+        pipeline_args["model"] = zenml_model
         pipeline_args["enable_cache"] = False
         run_args_deploying = {}
         pipeline_args[

huggingface-sagemaker/steps/deploying/sagemaker_deployment.py

@@ -49,7 +49,7 @@ def deploy_hf_to_sagemaker(
     #  Otherwise, use the provided values.
     if repo_id is None or revision is None:
         context = get_step_context()
-        mv = context.model_version
+        mv = context.model
         deployment_metadata = mv.get_data_artifact(name="huggingface_url").run_metadata
         repo_id = deployment_metadata["repo_id"].value
         revision = deployment_metadata["revision"].value

huggingface-sagemaker/steps/deploying/save_model.py

@@ -44,11 +44,11 @@ def save_model_to_deploy():
         f" Loading latest version of the model for stage {pipeline_extra['target_env']}..."
     )
     # Get the current model version
-    latest_version = get_step_context().model_version
+    current_zenml_model = get_step_context().model
 
     # Load model and tokenizer from Model Control Plane
-    model = latest_version.load_artifact(name="model")
-    tokenizer = latest_version.load_artifact(name="tokenizer")
+    model = current_zenml_model.load_artifact(name="model")
+    tokenizer = current_zenml_model.load_artifact(name="tokenizer")
     # Save the model and tokenizer locally
     model_path = "./gradio/"  # replace with the actual path
     tokenizer_path = "./gradio/"  # replace with the actual path

huggingface-sagemaker/steps/promotion/promote_get_metrics.py

@@ -25,9 +25,6 @@
 
 logger = get_logger(__name__)
 
-model_registry = Client().active_stack.model_registry
-
-
 @step
 def promote_get_metrics() -> (
     Tuple[
@@ -54,24 +51,23 @@ def promote_get_metrics() -> (
     zenml_client = Client()
 
     # Get current model version metric in current run
-    model_version = get_step_context().model_version
-    current_version = model_version._get_model_version()
-    current_metrics = current_version.get_model_artifact("model").run_metadata["metrics"].value
+    current_zenml_model = get_step_context().model
+    current_metrics = current_zenml_model.get_model_artifact("model").run_metadata["metrics"].value
     logger.info(f"Current model version metrics are {current_metrics}")
 
     # Get latest saved model version metric in target environment
     try:
-        latest_version = zenml_client.get_model_version(
-            model_name_or_id=model_version.name,
+        latest_zenml_model = zenml_client.get_model_version(
+            model_name_or_id=current_zenml_model.name,
             model_version_name_or_number_or_id=ModelStages(
                 pipeline_extra["target_env"]
             ),
         )
     except KeyError:
-        latest_version = None
-    if latest_version:
+        latest_zenml_model = None
+    if latest_zenml_model:
         latest_metrics = (
-            latest_version.get_model_artifact("model").run_metadata["metrics"].value
+            latest_zenml_model.get_model_artifact("model").run_metadata["metrics"].value
         )
         logger.info(f"Current model version metrics are {latest_metrics}")
     else:

huggingface-sagemaker/steps/promotion/promote_metric_compare_promoter.py

@@ -78,9 +78,8 @@ def promote_metric_compare_promoter(
             should_promote = False
 
     if should_promote:
-        model_version = get_step_context().model_version
-        model_version = model_version._get_model_version()
-        model_version.set_stage(pipeline_extra["target_env"], force=True)
+        zenml_model = get_step_context().model
+        zenml_model.set_stage(pipeline_extra["target_env"], force=True)
 
     logger.info(
         f"Promoted current model version to {pipeline_extra['target_env']} environment"