Machine Learning Model Serving Patterns and Best Practices: A definitive guide to deploying, monitoring, and providing accessibility to ML models in production

دانلود کتاب Machine Learning Model Serving Patterns and Best Practices: A definitive guide to deploying, monitoring, and providing accessibility to ML models in production

55000 تومان موجود

کتاب الگوهای ارائه مدل یادگیری ماشین و بهترین روش ها: راهنمای قطعی برای استقرار، نظارت و ارائه دسترسی به مدل های ML در تولید نسخه زبان اصلی

دانلود کتاب الگوهای ارائه مدل یادگیری ماشین و بهترین روش ها: راهنمای قطعی برای استقرار، نظارت و ارائه دسترسی به مدل های ML در تولید بعد از پرداخت مقدور خواهد بود
توضیحات کتاب در بخش جزئیات آمده است و می توانید موارد را مشاهده فرمایید


این کتاب نسخه اصلی می باشد و به زبان فارسی نیست.


امتیاز شما به این کتاب (حداقل 1 و حداکثر 5):

امتیاز کاربران به این کتاب:        تعداد رای دهنده ها: 2


توضیحاتی در مورد کتاب Machine Learning Model Serving Patterns and Best Practices: A definitive guide to deploying, monitoring, and providing accessibility to ML models in production

نام کتاب : Machine Learning Model Serving Patterns and Best Practices: A definitive guide to deploying, monitoring, and providing accessibility to ML models in production
عنوان ترجمه شده به فارسی : الگوهای ارائه مدل یادگیری ماشین و بهترین روش ها: راهنمای قطعی برای استقرار، نظارت و ارائه دسترسی به مدل های ML در تولید
سری :
نویسندگان :
ناشر : Packt Publishing
سال نشر : 2022
تعداد صفحات : 336
ISBN (شابک) : 9781803249902 , 1803249900
زبان کتاب : English
فرمت کتاب : pdf
حجم کتاب : 25 مگابایت



بعد از تکمیل فرایند پرداخت لینک دانلود کتاب ارائه خواهد شد. درصورت ثبت نام و ورود به حساب کاربری خود قادر خواهید بود لیست کتاب های خریداری شده را مشاهده فرمایید.


فهرست مطالب :


Cover\nTitle Page\nCopyright and Credits\nContributors\nAbout the reviewers\nTable of Contents\nPreface\nPart 1:Introduction to Model Serving\nChapter 1: Introducing Model Serving\n Technical requirements\n What is serving?\n What are models?\n What is model serving?\n Understanding the importance of model serving\n Using existing tools to serve models\n Summary\nChapter 2: Introducing Model Serving Patterns\n Design patterns in software engineering\n Understanding the value of model serving patterns\n ML serving patterns\n Serving philosophy patterns\n Patterns of serving approaches\n Summary\n Further reading\nPart 2:Patterns and Best Practices of Model Serving\nChapter 3: Stateless Model Serving\n Technical requirements\n Understanding stateful and stateless functions\n Stateless functions\n Stateful functions\n Extracting states from stateful functions\n Using stateful functions\n States in machine learning models\n Using input data as states\n Mitigating the impact of states from the ML model\n Summary\nChapter 4: Continuous Model Evaluation\n Technical requirements\n Introducing continuous model evaluation\n What to monitor in model evaluation\n Challenges of continuous model evaluation\n The necessity of continuous model evaluation\n Monitoring errors\n Deciding on retraining\n Enhancing serving resources\n Understanding business impact\n Common metrics for training and monitoring\n Continuous model evaluation use cases\n Evaluating a model continuously\n Collecting the ground truth\n Plotting metrics on a dashboard\n Selecting the threshold\n Setting a notification for performance drops\n Monitoring model performance when predicting rare classes\n Summary\n Further reading\nChapter 5: Keyed Prediction\n Technical requirements\n Introducing keyed prediction\n Exploring keyed prediction use cases\n Multi-threaded programming\n Multiple instances of the model running asynchronously\n Why the keyed prediction model is needed\n Exploring techniques for keyed prediction\n Passing keys with features from the clients\n Removing keys before the prediction\n Tagging predictions with keys\n Creating keys\n Summary\n Further reading\nChapter 6: Batch Model Serving\n Technical requirements\n Introducing batch model serving\n What is batch model serving?\n Different types of batch model serving\n Manual triggers\n Automatic periodic triggers\n Using continuous model evaluation to retrain\n Serving for offline inference\n Serving for on-demand inference\n Example scenarios of batch model serving\n Case 1 – recommendation\n Case 2 – sentiment analysis\n Techniques in batch model serving\n Setting up a periodic batch update\n Storing the predictions in a persistent store\n Pulling predictions by the server application\n Limitations of batch serving\n Summary\n Further reading\nChapter 7: Online Learning Model Serving\n Technical requirements\n Introducing online model serving\n Serving requests\n Use cases for online model serving\n Case 1 – recommending the nearest emergency center during a pandemic\n Case 2 – predicting the favorite soccer team in a tournament\n Case 3 – predicting the path of a hurricane or storm\n Case 4 – predicting the estimated delivery time of delivery trucks\n Challenges in online model serving\n Challenges in using newly arrived data for training\n Underperforming of the model after online training\n Overfitting and class imbalance\n Increasing of latency\n Handling concurrent requests\n Implementing online model serving\n Summary\n Further reading\nChapter 8: Two-Phase Model Serving\n Technical requirements\n Introducing two-phase model serving\n Exploring two-phase model serving techniques\n Training and saving an MNIST model\n Full integer quantization of the model and saving the converted model\n Comparing the size and accuracy of the models\n Separately trained phase one model with reduced features\n Separately trained different models\n Use cases of two-phase model serving\n Case 4 – route planners\n Summary\n Further reading\nChapter 9: Pipeline Pattern Model Serving\n Technical requirements\n Introducing the pipeline pattern\n A DAG\n Stages of the machine learning pipeline\n Introducing Apache Airflow\n Getting started with Apache Airflow\n Creating and starting a pipeline using Apache Airflow\n Demonstrating a machine learning pipeline using Airflow\n Advantages and disadvantages of the pipeline pattern\n Summary\n Further reading\nChapter 10: Ensemble Model Serving Pattern\n Technical requirements\n Introducing the ensemble pattern\n Using ensemble pattern techniques\n Model update\n Aggregation\n Model selection\n Combining responses\n End-to-end dummy example of serving the model\n Summary\nChapter 11: Business Logic Pattern\n Technical requirements\n Introducing the business logic pattern\n Type of business logic\n Technical approaches to business logic in model serving\n Data validation\n Feature transformation\n Prediction post-processing\n Summary\nPart 3: Introduction to Tools for Model Serving\nChapter 12: Exploring TensorFlow Serving\n Technical requirements\n Introducing TensorFlow Serving\n Servable\n Loader\n Source\n Aspired versions\n Manager\n Using TensorFlow Serving to serve models\n TensorFlow Serving with Docker\n Using advanced model configurations\n Summary\n Further reading\nChapter 13: Using Ray Serve\n Technical requirements\n Introducing Ray Serve\n Deployment\n ServeHandle\n Ingress deployment\n Deployment graph\n Using Ray Serve to serve a model\n Using the ensemble pattern in Ray Serve\n Using Ray Serve with the pipeline pattern\n Summary\n Further reading\nChapter 14: Using BentoML\n Technical requirements\n Introducing BentoML\n Preparing models\n Services and APIs\n Bento\n Using BentoML to serve a model\n Summary\n Further reading\nPart 4:Exploring Cloud Solutions\nChapter 15: Serving ML Models Using a Fully Managed Cloud Solution\n Technical requirements\n Introducing Amazon SageMaker\n Amazon SageMaker features\n Using Amazon SageMaker to serve a model\n Creating a notebook in Amazon SageMaker\n Serving the model using Amazon SageMaker\n Summary\nIndex\nOther Books You May Enjoy




پست ها تصادفی