You are building an ML model to detect anomalies in real-time sensor data. You will use Pub/Sub to handle incoming requests. You want to store the results for analytics and visualization. How should you configure the pipeline?
1 = Dataflow, 2 - Al Platform, 3 = BigQuery
1 = DataProc, 2 = AutoML, 3 = Cloud Bigtable
1 = BigQuery, 2 = AutoML, 3 = Cloud Functions
1 = BigQuery, 2 = Al Platform, 3 = Cloud Storage
Dataflow is a fully managed service for executing Apache Beam pipelines that can process streaming or batch data1.
Al Platform is a unified platform that enables you to build and run machine learning applications across Google Cloud2.
BigQuery is a serverless, highly scalable, and cost-effective cloud data warehouse designed for business agility3.
These services are suitable for building an ML model to detect anomalies in real-time sensor data, as they can handle large-scale data ingestion, preprocessing, training, serving, storage, and visualization. The other options are not as suitable because:
DataProc is a service for running Apache Spark and Apache Hadoop clusters, which are not optimized for streaming data processing4.
AutoML is a suite of machine learning products that enables developers with limited machine learning expertise to train high-quality models specific to their business needs5. However, it does not support custom models or real-time predictions.
Cloud Bigtable is a scalable, fully managed NoSQL database service for large analytical and operational workloads. However, it is not designed for ad hoc queries or interactive analysis.
Cloud Functions is a serverless execution environment for building and connecting cloud services. However, it is not suitable for storing or visualizing data.
Cloud Storage is a service for storing and accessing data on Google Cloud. However, it is not a data warehouse and does not support SQL queries or visualization tools.
You need to quickly build and train a model to predict the sentiment of customer reviews with custom categories without writing code. You do not have enough data to train a model from scratch. The resulting model should have high predictive performance. Which service should you use?
AutoML Natural Language
Cloud Natural Language API
AI Hub pre-made Jupyter Notebooks
AI Platform Training built-in algorithms
AutoML Natural Language is a service that allows you to build and train custom natural language models without writing code. You can use AutoML Natural Language to perform sentiment analysis with custom categories, such as positive, negative, or neutral. You can also use pre-trained models or transfer learning to leverage existing knowledge and reduce the amount of data required to train a model from scratch. AutoML Natural Language provides a user-friendly interface and a powerful AutoML engine that optimizes your model for high predictive performance.
Cloud Natural Language API is a service that provides pre-trained models for common natural language tasks, such as sentiment analysis, entity analysis, and syntax analysis. However, it does not allow you to customize the categories or use your own data for training.
AI Hub pre-made Jupyter Notebooks are interactive documents that contain code, text, and visualizations for various machine learning scenarios. However, they require some coding skills and data preparation to use them effectively.
AI Platform Training built-in algorithms are pre-configured machine learning algorithms that you can use to train models on AI Platform. However, they do not support sentiment analysis as a natural language task.
References:
AutoML Natural Language documentation
Cloud Natural Language API documentation
AI Hub documentation
AI Platform Training documentation
You work with a team of researchers to develop state-of-the-art algorithms for financial analysis. Your team develops and debugs complex models in TensorFlow. You want to maintain the ease of debugging while also reducing the model training time. How should you set up your training environment?
Configure a v3-8 TPU VM SSH into the VM to tram and debug the model.
Configure a v3-8 TPU node Use Cloud Shell to SSH into the Host VM to train and debug the model.
Configure a M-standard-4 VM with 4 NVIDIA P100 GPUs SSH into the VM and use
Parameter Server Strategy to train the model.
Configure a M-standard-4 VM with 4 NVIDIA P100 GPUs SSH into the VM and use
MultiWorkerMirroredStrategy to train the model.
A TPU VM is a virtual machine that has direct access to a Cloud TPU device. TPU VMs provide a simpler and more flexible way to use Cloud TPUs, as they eliminate the need for a separate host VM and network setup. TPU VMs also support interactive debugging tools such as TensorFlow Debugger (tfdbg) and Python Debugger (pdb), which can help researchers develop and troubleshoot complex models. A v3-8 TPU VM has 8 TPU cores, which can provide high performance and scalability for training large models. SSHing into the TPU VM allows the user to run and debug the TensorFlow code directly on the TPU device, without any network overhead or data transfer issues. References:
1: TPU VMs Overview
2: TPU VMs Quickstart
3: Debugging TensorFlow Models on Cloud TPUs
You recently developed a deep learning model using Keras, and now you are experimenting with different training strategies. First, you trained the model using a single GPU, but the training process was too slow. Next, you distributed the training across 4 GPUs using tf.distribute.MirroredStrategy (with no other changes), but you did not observe a decrease in training time. What should you do?
Distribute the dataset with tf.distribute.Strategy.experimental_distribute_dataset
Create a custom training loop.
Use a TPU with tf.distribute.TPUStrategy.
Increase the batch size.
Option A is incorrect because distributing the dataset with tf.distribute.Strategy.experimental_distribute_dataset is not the most effective way to decrease the training time. This method allows you to distribute your dataset across multiple devices or machines, by creating a tf.data.Dataset instance that can be iterated over in parallel1. However, this option may not improve the training time significantly, as it does not change the amount of data or computation that each device or machine has to process. Moreover, this option may introduce additional overhead or complexity, as it requires you to handle the data sharding, replication, and synchronization across the devices or machines1.
Option B is incorrect because creating a custom training loop is not the easiest way to decrease the training time. A custom training loop is a way to implement your own logic for training your model, by using low-level TensorFlow APIs, such as tf.GradientTape, tf.Variable, or tf.function2. A custom training loop may give you more flexibility and control over the training process, but it also requires more effort and expertise, as you have to write and debug the code for each step of the training loop, such as computing the gradients, applying the optimizer, or updating the metrics2. Moreover, a custom training loop may not improve the training time significantly, as it does not change the amount of data or computation that each device or machine has to process.
Option C is incorrect because using a TPU with tf.distribute.TPUStrategy is not a valid way to decrease the training time. A TPU (Tensor Processing Unit) is a custom hardware accelerator designed for high-performance ML workloads3. A tf.distribute.TPUStrategy is a distribution strategy that allows you to distribute your training across multiple TPUs, by creating a tf.distribute.TPUStrategy instance that can be used with high-level TensorFlow APIs, such as Keras4. However, this option is not feasible, as Vertex AI Training does not support TPUs as accelerators for custom training jobs5. Moreover, this option may require significant code changes, as TPUs have different requirements and limitations than GPUs.
Option D is correct because increasing the batch size is the best way to decrease the training time. The batch size is a hyperparameter that determines how many samples of data are processed in each iteration of the training loop. Increasing the batch size may reduce the training time, as it reduces the number of iterations needed to train the model, and it allows each device or machine to process more data in parallel. Increasing the batch size is also easy to implement, as it only requires changing a single hyperparameter. However, increasing the batch size may also affect the convergence and the accuracy of the model, so it is important to find the optimal batch size that balances the trade-off between the training time and the model performance.
References:
tf.distribute.Strategy.experimental_distribute_dataset
Custom training loop
TPU overview
tf.distribute.TPUStrategy
Vertex AI Training accelerators
[TPU programming model]
[Batch size and learning rate]
[Keras overview]
[tf.distribute.MirroredStrategy]
[Vertex AI Training overview]
[TensorFlow overview]
You work on a growing team of more than 50 data scientists who all use AI Platform. You are designing a strategy to organize your jobs, models, and versions in a clean and scalable way. Which strategy should you choose?
Set up restrictive IAM permissions on the AI Platform notebooks so that only a single user or group can access a given instance.
Separate each data scientist’s work into a different project to ensure that the jobs, models, and versions created by each data scientist are accessible only to that user.
Use labels to organize resources into descriptive categories. Apply a label to each created resource so that users can filter the results by label when viewing or monitoring the resources.
Set up a BigQuery sink for Cloud Logging logs that is appropriately filtered to capture information about AI Platform resource usage. In BigQuery, create a SQL view that maps users to the resources they are using
Labels are key-value pairs that you can attach to AI Platform resources such as jobs, models, and versions. Labels can help you organize your resources into descriptive categories that reflect your business needs. For example, you can use labels to indicate the owner, purpose, environment, or status of a resource. You can also use labels to filter the results when you list or monitor your resources on the Google Cloud Console or the Cloud SDK. Using labels can help you manage your resources in a clean and scalable way, without requiring separate projects or restrictive permissions.
References:
Using labels to organize AI Platform resources
Creating and managing labels
QUESTION 52
You are training a deep learning model for semantic image segmentation with reduced training time. While using a Deep Learning VM Image, you receive the following error: The resource 'projects/deeplearning-platforn/zones/europe-west4-c/acceleratorTypes/nvidia-tesla-k80' was not found. What should you do?
A. Ensure that you have GPU quota in the selected region.
B. Ensure that the required GPU is available in the selected region.
C. Ensure that you have preemptible GPU quota in the selected region.
D. Ensure that the selected GPU has enough GPU memory for the workload.
Answer: B
The error message indicates that the selected GPU type (nvidia-tesla-k80) is not available in the selected region (europe-west4-c). This can happen when the GPU type is not supported in the region, or when the GPU quota is exhausted in the region. To avoid this error, you should ensure that the required GPU is available in the selected region before creating a Deep Learning VM Image. You can use the following steps to check the GPU availability and quota:
To check the GPU availability, you can use the gcloud compute accelerator-types list command with the --filter flag to specify the GPU type and the region. For example, to check the availability of nvidia-tesla-k80 in europe-west4-c, you can run:
gcloud compute accelerator-types list --filter="name=nvidia-tesla-k80 AND zone:europe-west4-c"
If the command returns an empty result, it means that the GPU type is not supported in the region. You can either choose a different GPU type or a different region that supports the GPU type. You can use the same command without the --filter flag to list all the available GPU types and regions. For example, to list all the available GPU types in europe-west4-c, you can run:
gcloud compute accelerator-types list --filter="zone:europe-west4-c"
To check the GPU quota, you can use the gcloud compute regions describe command with the --format flag to specify the region and the quota metric. For example, to check the quota for nvidia-tesla-k80 in europe-west4-c, you can run:
gcloud compute regions describe europe-west4-c --format="value(quotas.NVIDIA_K80_GPUS)"
If the command returns a value of 0, it means that the GPU quota is exhausted in the region. You can either request more quota from Google Cloud or choose a different region that has enough quota for the GPU type.
References:
Troubleshooting | Deep Learning VM Images | Google Cloud
Checking GPU availability
Checking GPU quota
You need to train a natural language model to perform text classification on product descriptions that contain millions of examples and 100,000 unique words. You want to preprocess the words individually so that they can be fed into a recurrent neural network. What should you do?
Create a hot-encoding of words, and feed the encodings into your model.
Identify word embeddings from a pre-trained model, and use the embeddings in your model.
Sort the words by frequency of occurrence, and use the frequencies as the encodings in your model.
Assign a numerical value to each word from 1 to 100,000 and feed the values as inputs in your model.
Option A is incorrect because creating a one-hot encoding of words, and feeding the encodings into your model is not an efficient way to preprocess the words individually for a natural language model. One-hot encoding is a method of representing categorical variables as binary vectors, where each element corresponds to a category and only one element is 1 and the rest are 01. However, this method is not suitable for high-dimensional and sparse data, such as words in a large vocabulary, because it requires a lot of memory and computation, and does not capture the semantic similarity or relationship between words2.
Option B is correct because identifying word embeddings from a pre-trained model, and using the embeddings in your model is a good way to preprocess the words individually for a natural language model. Word embeddings are low-dimensional and dense vectors that represent the meaning and usage of words in a continuous space3. Word embeddings can be learned from a large corpus of text using neural networks, such as word2vec, GloVe, or BERT4. Using pre-trained word embeddings can save time and resources, and improve the performance of the natural language model, especially when the training data is limited or noisy5.
Option C is incorrect because sorting the words by frequency of occurrence, and using the frequencies as the encodings in your model is not a meaningful way to preprocess the words individually for a natural language model. This method implies that the frequency of a word is a good indicator of its importance or relevance, which may not be true. For example, the word “the” is very frequent but not very informative, while the word “unicorn” is rare but more distinctive. Moreover, this method does not capture the semantic similarity or relationship between words, and may introduce noise or bias into the model.
Option D is incorrect because assigning a numerical value to each word from 1 to 100,000 and feeding the values as inputs in your model is not a valid way to preprocess the words individually for a natural language model. This method implies an ordinal relationship between the words, which may not be true. For example, assigning the values 1, 2, and 3 to the words “apple”, “banana”, and “orange” does not make sense, as there is no inherent order among these fruits. Moreover, this method does not capture the semantic similarity or relationship between words, and may confuse the model with irrelevant or misleading information.
References:
One-hot encoding
Word embeddings
Word embedding
Pre-trained word embeddings
Using pre-trained word embeddings in a Keras model
[Term frequency]
[Term frequency-inverse document frequency]
[Ordinal variable]
[Encoding categorical features]
You want to rebuild your ML pipeline for structured data on Google Cloud. You are using PySpark to conduct data transformations at scale, but your pipelines are taking over 12 hours to run. To speed up development and pipeline run time, you want to use a serverless tool and SQL syntax. You have already moved your raw data into Cloud Storage. How should you build the pipeline on Google Cloud while meeting the speed and processing requirements?
Use Data Fusion's GUI to build the transformation pipelines, and then write the data into BigQuery
Convert your PySpark into SparkSQL queries to transform the data and then run your pipeline on Dataproc to write the data into BigQuery.
Ingest your data into Cloud SQL convert your PySpark commands into SQL queries to transform the data, and then use federated queries from BigQuery for machine learning
Ingest your data into BigQuery using BigQuery Load, convert your PySpark commands into BigQuery SQL queries to transform the data, and then write the transformations to a new table
BigQuery is a serverless, scalable, and cost-effective data warehouse that allows users to run SQL queries on large volumes of data. BigQuery Load is a tool that can ingest data from Cloud Storage into BigQuery tables. BigQuery SQL is a dialect of SQL that supports many of the same functions and operations as PySpark, such as window functions, aggregate functions, joins, and subqueries. By using BigQuery Load and BigQuery SQL, you can rebuild your ML pipeline for structured data on Google Cloud without having to manage any servers or clusters, and with faster performance and lower cost than using PySpark on Dataproc. You can also use BigQuery ML to create and evaluate ML models using SQL commands. References:
BigQuery documentation
BigQuery Load documentation
BigQuery SQL reference
BigQuery ML documentation
You have recently trained a scikit-learn model that you plan to deploy on Vertex Al. This model will support both online and batch prediction. You need to preprocess input data for model inference. You want to package the model for deployment while minimizing additional code What should you do?
1 Upload your model to the Vertex Al Model Registry by using a prebuilt scikit-learn prediction container
2 Deploy your model to Vertex Al Endpoints, and create a Vertex Al batch prediction job that uses the instanceConfig.inscanceType setting to transform your input data
1 Wrap your model in a custom prediction routine (CPR). and build a container image from the CPR local model
2 Upload your sci-kit learn model container to Vertex Al Model Registry
3 Deploy your model to Vertex Al Endpoints, and create a Vertex Al batch prediction job
1. Create a custom container for your sci-kit learn model,
2 Define a custom serving function for your model
3 Upload your model and custom container to Vertex Al Model Registry
4 Deploy your model to Vertex Al Endpoints, and create a Vertex Al batch prediction job
1 Create a custom container for your sci-kit learn model.
2 Upload your model and custom container to Vertex Al Model Registry
3 Deploy your model to Vertex Al Endpoints, and create a Vertex Al batch prediction job that uses the instanceConfig. instanceType setting to transform your input data
The best option for deploying a scikit-learn model on Vertex AI with minimal additional code is to wrap the model in a custom prediction routine (CPR) and build a container image from the CPR local model. Upload your scikit-learn model container to Vertex AI Model Registry. Deploy your model to Vertex AI Endpoints, and create a Vertex AI batch prediction job. This option allows you to leverage the power and simplicity of Google Cloud to deploy and serve a scikit-learn model that supports both online and batch prediction. Vertex AI is a unified platform for building and deploying machine learning solutions on Google Cloud. Vertex AI can deploy a trained scikit-learn model to an online prediction endpoint, which can provide low-latency predictions for individual instances. Vertex AI can also create a batch prediction job, which can provide high-throughput predictions for a large batch of instances. A custom prediction routine (CPR) is a Python script that defines the logic for preprocessing the input data, running the prediction, and postprocessing the output data. A CPR can help you customize the prediction behavior of your model, and handle complex or non-standard data formats. A CPR can also help you minimize the additional code, as you only need to write a few functions to implement the prediction logic. A container image is a package that contains the model, the CPR, and the dependencies. A container image can help you standardize and simplify the deployment process, as you only need to upload the container image to Vertex AI Model Registry, and deploy it to Vertex AI Endpoints. By wrapping the model in a CPR and building a container image from the CPR local model, uploading the scikit-learn model container to Vertex AI Model Registry, deploying the model to Vertex AI Endpoints, and creating a Vertex AI batch prediction job, you can deploy a scikit-learn model on Vertex AI with minimal additional code1.
The other options are not as good as option B, for the following reasons:
Option A: Uploading your model to the Vertex AI Model Registry by using a prebuilt scikit-learn prediction container, deploying your model to Vertex AI Endpoints, and creating a Vertex AI batch prediction job that uses the instanceConfig.instanceType setting to transform your input data would not allow you to preprocess the input data for model inference, and could cause errors or poor performance. A prebuilt scikit-learn prediction container is a container image that is provided by Google Cloud, and contains the scikit-learn framework and the dependencies. A prebuilt scikit-learn prediction container can help you deploy a scikit-learn model without writing any code, but it also limits your customization options. A prebuilt scikit-learn prediction container can only handle standard data formats, such as JSON or CSV, and cannot perform any preprocessing or postprocessing on the input or output data. If your input data requires any transformation or normalization before running the prediction, you cannot use a prebuilt scikit-learn prediction container. The instanceConfig.instanceType setting is a parameter that determines the machine type and the accelerator type for the batch prediction job. The instanceConfig.instanceType setting can help you optimize the performance and the cost of the batch prediction job, but it cannot help you transform your input data2.
Option C: Creating a custom container for your scikit-learn model, defining a custom serving function for your model, uploading your model and custom container to Vertex AI Model Registry, and deploying your model to Vertex AI Endpoints, and creating a Vertex AI batch prediction job would require more skills and steps than using a CPR and a container image. A custom container is a container image that contains the model, the dependencies, and a web server. A custom container can help you customize the prediction behavior of your model, and handle complex or non-standard data formats. A custom serving function is a Python function that defines the logic for running the prediction on the model. A custom serving function can help you implement the prediction logic of your model, and handle complex or non-standard data formats. However, creating a custom container and defining a custom serving function would require more skills and steps than using a CPR and a container image. You would need to write code, build and test the container image, configure the web server, and implement the prediction logic. Moreover, creating a custom container and defining a custom serving function would not allow you to preprocess the input data for model inference, as the custom serving function only runs the prediction on the model3.
Option D: Creating a custom container for your scikit-learn model, uploading your model and custom container to Vertex AI Model Registry, deploying your model to Vertex AI Endpoints, and creating a Vertex AI batch prediction job that uses the instanceConfig.instanceType setting to transform your input data would not allow you to preprocess the input data for model inference, and could cause errors or poor performance. A custom container is a container image that contains the model, the dependencies, and a web server. A custom container can help you customize the prediction behavior of your model, and handle complex or non-standard data formats. However, creating a custom container would require more skills and steps than using a CPR and a container image. You would need to write code, build and test the container image, and configure the web server. The instanceConfig.instanceType setting is a parameter that determines the machine type and the accelerator type for the batch prediction job. The instanceConfig.instanceType setting can help you optimize the performance and the cost of the batch prediction job, but it cannot help you transform your input data23.
References:
Preparing for Google Cloud Certification: Machine Learning Engineer, Course 3: Production ML Systems, Week 2: Serving ML Predictions
Google Cloud Professional Machine Learning Engineer Exam Guide, Section 3: Scaling ML models in production, 3.1 Deploying ML models to production
Official Google Cloud Certified Professional Machine Learning Engineer Study Guide, Chapter 6: Production ML Systems, Section 6.2: Serving ML Predictions
Custom prediction routines
Using pre-built containers for prediction
Using custom containers for prediction
You recently used XGBoost to train a model in Python that will be used for online serving Your model prediction service will be called by a backend service implemented in Golang running on a Google Kubemetes Engine (GKE) cluster Your model requires pre and postprocessing steps You need to implement the processing steps so that they run at serving time You want to minimize code changes and infrastructure maintenance and deploy your model into production as quickly as possible. What should you do?
Use FastAPI to implement an HTTP server Create a Docker image that runs your HTTP server and deploy it on your organization's GKE cluster.
Use FastAPI to implement an HTTP server Create a Docker image that runs your HTTP server Upload the image to Vertex Al Model Registry and deploy it to a Vertex Al endpoint.
Use the Predictor interface to implement a custom prediction routine Build the custom contain upload the container to Vertex Al Model Registry, and deploy it to a Vertex Al endpoint.
Use the XGBoost prebuilt serving container when importing the trained model into Vertex Al Deploy the model to a Vertex Al endpoint Work with the backend engineers to implement the pre- and postprocessing steps in the Golang backend service.
The best option for implementing the processing steps so that they run at serving time, minimizing code changes and infrastructure maintenance, and deploying the model into production as quickly as possible, is to use the Predictor interface to implement a custom prediction routine. Build the custom container, upload the container to Vertex AI Model Registry, and deploy it to a Vertex AI endpoint. This option allows you to leverage the power and simplicity of Vertex AI to serve your XGBoost model with minimal effort and customization. Vertex AI is a unified platform for building and deploying machine learning solutions on Google Cloud. Vertex AI can deploy a trained XGBoost model to an online prediction endpoint, which can provide low-latency predictions for individual instances. A custom prediction routine (CPR) is a Python script that defines the logic for preprocessing the input data, running the prediction, and postprocessing the output data. A CPR can help you customize the prediction behavior of your model, and handle complex or non-standard data formats. A CPR can also help you minimize the code changes, as you only need to write a few functions to implement the prediction logic. A Predictor interface is a class that inherits from the base class aiplatform.Predictor, and implements the abstract methods predict() and preprocess(). A Predictor interface can help you create a CPR by defining the preprocessing and prediction logic for your model. A container image is a package that contains the model, the CPR, and the dependencies. A container image can help you standardize and simplify the deployment process, as you only need to upload the container image to Vertex AI Model Registry, and deploy it to Vertex AI Endpoints. By using the Predictor interface to implement a CPR, building the custom container, uploading the container to Vertex AI Model Registry, and deploying it to a Vertex AI endpoint, you can implement the processing steps so that they run at serving time, minimize code changes and infrastructure maintenance, and deploy the model into production as quickly as possible1.
The other options are not as good as option C, for the following reasons:
Option A: Using FastAPI to implement an HTTP server, creating a Docker image that runs your HTTP server, and deploying it on your organization’s GKE cluster would require more skills and steps than using the Predictor interface to implement a CPR, building the custom container, uploading the container to Vertex AI Model Registry, and deploying it to a Vertex AI endpoint. FastAPI is a framework for building web applications and APIs in Python. FastAPI can help you implement an HTTP server that can handle prediction requests and responses, and perform data preprocessing and postprocessing. A Docker image is a package that contains the model, the HTTP server, and the dependencies. A Docker image can help you standardize and simplify the deployment process, as you only need to build and run the Docker image. GKE is a service that can create and manage Kubernetes clusters on Google Cloud. GKE can help you deploy and scale your Docker image on Google Cloud, and provide high availability and performance. However, using FastAPI to implement an HTTP server, creating a Docker image that runs your HTTP server, and deploying it on your organization’s GKE cluster would require more skills and steps than using the Predictor interface to implement a CPR, building the custom container, uploading the container to Vertex AI Model Registry, and deploying it to a Vertex AI endpoint. You would need to write code, create and configure the HTTP server, build and test the Docker image, create and manage the GKE cluster, and deploy and monitor the Docker image. Moreover, this option would not leverage the power and simplicity of Vertex AI, which can provide online prediction natively integrated with Google Cloud services2.
Option B: Using FastAPI to implement an HTTP server, creating a Docker image that runs your HTTP server, uploading the image to Vertex AI Model Registry, and deploying it to a Vertex AI endpoint would require more skills and steps than using the Predictor interface to implement a CPR, building the custom container, uploading the container to Vertex AI Model Registry, and deploying it to a Vertex AI endpoint. FastAPI is a framework for building web applications and APIs in Python. FastAPI can help you implement an HTTP server that can handle prediction requests and responses, and perform data preprocessing and postprocessing. A Docker image is a package that contains the model, the HTTP server, and the dependencies. A Docker image can help you standardize and simplify the deployment process, as you only need to build and run the Docker image. Vertex AI Model Registry is a service that can store and manage your machine learning models on Google Cloud. Vertex AI Model Registry can help you upload and organize your Docker image, and track the model versions and metadata. Vertex AI Endpoints is a service that can provide online prediction for your machine learning models on Google Cloud. Vertex AI Endpoints can help you deploy your Docker image to an online prediction endpoint, which can provide low-latency predictions for individual instances. However, using FastAPI to implement an HTTP server, creating a Docker image that runs your HTTP server, uploading the image to Vertex AI Model Registry, and deploying it to a Vertex AI endpoint would require more skills and steps than using the Predictor interface to implement a CPR, building the custom container, uploading the container to Vertex AI Model Registry, and deploying it to a Vertex AI endpoint. You would need to write code, create and configure the HTTP server, build and test the Docker image, upload the Docker image to Vertex AI Model Registry, and deploy the Docker image to Vertex AI Endpoints. Moreover, this option would not leverage the power and simplicity of Vertex AI, which can provide online prediction natively integrated with Google Cloud services2.
Option D: Using the XGBoost prebuilt serving container when importing the trained model into Vertex AI, deploying the model to a Vertex AI endpoint, working with the backend engineers to implement the pre- and postprocessing steps in the Golang backend service would not allow you to implement the processing steps so that they run at serving time, and could increase the code changes and infrastructure maintenance. A XGBoost prebuilt serving container is a container image that is provided by Google Cloud, and contains the XGBoost framework and the dependencies. A XGBoost prebuilt serving container can help you deploy a XGBoost model without writing any code, but it also limits your customization options. A XGBoost prebuilt serving container can only handle standard data formats, such as JSON or CSV, and cannot perform any preprocessing or postprocessing on the input or output data. If your input data requires any transformation or normalization before running the prediction, you cannot use a XGBoost prebuilt serving container. A Golang backend service is a service that is implemented in Golang, a programming language that can be used for web development and system programming. A Golang backend service can help you handle the prediction requests and responses from the frontend, and communicate with the Vertex AI endpoint. However, using the XGBoost prebuilt serving container when importing the trained model into Vertex AI, deploying the model to a Vertex AI endpoint, working with the backend engineers to implement the pre- and postprocessing steps in the Golang backend service would not allow you to implement the processing steps so that they run at serving time, and could increase the code changes and infrastructure maintenance. You would need to write code, import the trained model into Vertex AI, deploy the model to a Vertex AI endpoint, implement the pre- and postprocessing steps in the Golang backend service, and test and monitor the Golang backend service. Moreover, this option would not leverage the power and simplicity of Vertex AI, which can provide online prediction natively integrated with Google Cloud services2.
References:
Preparing for Google Cloud Certification: Machine Learning Engineer, Course 3: Production ML Systems, Week 2: Serving ML Predictions
Google Cloud Professional Machine Learning Engineer Exam Guide, Section 3: Scaling ML models in production, 3.1 Deploying ML models to production
Official Google Cloud Certified Professional Machine Learning Engineer Study Guide, Chapter 6: Production ML Systems, Section 6.2: Serving ML Predictions
Custom prediction routines
Using pre-built containers for prediction
Using custom containers for prediction
You developed an ML model with Al Platform, and you want to move it to production. You serve a few thousand queries per second and are experiencing latency issues. Incoming requests are served by a load balancer that distributes them across multiple Kubeflow CPU-only pods running on Google Kubernetes Engine (GKE). Your goal is to improve the serving latency without changing the underlying infrastructure. What should you do?
Significantly increase the max_batch_size TensorFlow Serving parameter
Switch to the tensorflow-model-server-universal version of TensorFlow Serving
Significantly increase the max_enqueued_batches TensorFlow Serving parameter
Recompile TensorFlow Serving using the source to support CPU-specific optimizations Instruct GKE to choose an appropriate baseline minimum CPU platform for serving nodes
TensorFlow Serving is a service that allows you to deploy and serve TensorFlow models in a scalable and efficient way. TensorFlow Serving supports various platforms and hardware, such as CPU, GPU, and TPU. However, the default TensorFlow Serving binaries are built with generic CPU instructions, which may not leverage the full potential of the CPU architecture. To improve the serving latency and performance, you can recompile TensorFlow Serving using the source code and enable CPU-specific optimizations, such as AVX, AVX2, and FMA1. These optimizations can speed up the computation and inference of the TensorFlow models, especially for deep neural networks.
Google Kubernetes Engine (GKE) is a service that allows you to run and manage containerized applications on Google Cloud using Kubernetes. GKE supports various types and sizes of nodes, which are the virtual machines that run the containers. GKE also supports different CPU platforms, which are the generations and models of the CPUs that power the nodes. GKE allows you to choose a baseline minimum CPU platform for your node pool, which is a group of nodes with the same configuration. By choosing a baseline minimum CPU platform, you can ensure that your nodes have the CPU features and capabilities that match your workload requirements2.
For the use case of serving a few thousand queries per second and experiencing latency issues, the best option is to recompile TensorFlow Serving using the source to support CPU-specific optimizations, and instruct GKE to choose an appropriate baseline minimum CPU platform for serving nodes. This option can improve the serving latency and performance without changing the underlying infrastructure, as it only involves rebuilding the TensorFlow Serving binary and selecting the CPU platform for the GKE nodes. This option can also take advantage of the CPU-only pods that are running on GKE, as it can optimize the CPU utilization and efficiency. Therefore, recompiling TensorFlow Serving using the source to support CPU-specific optimizations and instructing GKE to choose an appropriate baseline minimum CPU platform for serving nodes is the best option for this use case.
References:
Building TensorFlow Serving from source
Specifying a minimum CPU platform for a node pool
