Review Cloud ML's capabilities to monitor performance of models and identify bottlenecks in this video.
- [Instructor] So, you have built and deployed a model.…That's great.…Now, you need to see how it's doing.…Cloud ML provides a user interface…to monitor performance of models in real time.…Go to the models page, and select the model.…Then select the version, and scroll down the page.…You will see performance data for this specific version.…You can choose a chart interval based on…the interval for which you want to analyze performance.…
Predictions per second shows…the incoming prediction load requests.…This tells you how much load is being put on the model.…Prediction errors keep track of errors that happen…during the prediction process.…If there is any code-related errors, or data-related errors,…it is going to show up here.…Requests per second contains the number…of requests coming in.…Each request might contain multiple predictions.…Response codes indicate the HTTP response code,…returned through the prediction process.…
A 200 OK response code indicates success.…Other codes need to be analyzed for why they are happening,…
- Evaluating the machine learning tools in GCP
- Understanding the predictive analytics process
- Building models
- Training models with jobs
- Building and running predictions
- Best practices for cost control, testing, and performance monitoring
Skill Level Intermediate
Predictive Customer Analyticswith Kumaran Ponnambalam1h 37m Intermediate
1. ML Options in GCP
2. Cloud ML Basics
3. Model Building with Cloud ML
4. Predictions in Cloud ML
5. Cloud ML Best Practices
- Mark as unwatched
- Mark all as unwatched
Are you sure you want to mark all the videos in this course as unwatched?
This will not affect your course history, your reports, or your certificates of completion for this course.Cancel
Take notes with your new membership!
Type in the entry box, then click Enter to save your note.
1:30Press on any video thumbnail to jump immediately to the timecode shown.
Notes are saved with you account but can also be exported as plain text, MS Word, PDF, Google Doc, or Evernote.