In one to two sentences, state the recommended timeline for creating an initial dev/test set and metric on a new project.
Question: In one to two sentences, describe the recommended timeline for creating an initial dev/test set and metric when starting a new project.
Sample answer: Teams should try to create an initial dev/test set and metric in less than one week, rarely longer, when starting out on a new project.
Key points:
- Less than one week for new projects
- Rarely longer than a week
- Applies to brand new applications, not mature ones
Rubric: Full credit states the under-one-week (or 'less than a week') timeline specifically for new projects. Partial credit gives a vague or incorrect timeframe.
0
1
References
Machine Learning Yearning (Deeplearning.ai)
Machine Learning Yearning (Deeplearning.ai)
Machine Learning Yearning (Deeplearning.ai)
Machine Learning Yearning (Deeplearning.ai)
Machine Learning Yearning (Deeplearning.ai)
Machine Learning Yearning (Deeplearning.ai)
Machine Learning Yearning (Deeplearning.ai)
Tags
Machine Learning
Deep Learning
Supervised Learning
Dive into Deep Learning @ D2L
Data Science
Machine Learning Strategy
Machine Learning Yearning @ DeepLearning.AI
Related
For a brand new application, roughly how quickly should a team establish initial dev/test sets and a metric?
True or False: The one-week timeline for establishing dev/test sets applies equally to mature applications.
When starting a new project, quickly choosing dev/test sets gives the team a well-defined _____ to aim for.
Match each project situation to its recommended dev/test set timeline.
Order the steps a team should follow when quickly establishing dev/test sets for a new application.
Why does establishing dev/test sets quickly benefit a team working on a new application?
A team debates spending three weeks perfecting dev/test sets for a brand new application before starting development.
In one to two sentences, state the recommended timeline for creating an initial dev/test set and metric on a new project.
What is the main reason to quickly choose dev/test sets when starting a new project?
True or False: It is acceptable for the initial dev/test set and metric on a new project to be imperfect.