Challenge: Initializing Model Weights and Biases
Task
Swipe to start coding
You are tasked with creating random tensors to initialize weights and biases for a simple neural network.
- Ensure reproducibility by setting a manual seed to an arbitrary number before generating the tensors.
- Create a
3x4
tensor filled with random values from a uniform distribution between0
and1
(weights for the first layer). - Create a
1x4
tensor filled with zeros (biases for the first layer). - Create a
4x2
tensor with random integers between-5
and5
(weights for the second layer).
Solution
Everything was clear?
Thanks for your feedback!
SectionΒ 1. ChapterΒ 6
single
Ask AI
Ask AI
Ask anything or try one of the suggested questions to begin our chat
Suggested prompts:
Why is initializing weights with random integers not recommended?
What are better ways to initialize weights in practice?
Can you explain the potential issues with using random integers for weights?
Awesome!
Completion rate improved to 5
Challenge: Initializing Model Weights and Biases
Swipe to show menu
Task
Swipe to start coding
You are tasked with creating random tensors to initialize weights and biases for a simple neural network.
- Ensure reproducibility by setting a manual seed to an arbitrary number before generating the tensors.
- Create a
3x4
tensor filled with random values from a uniform distribution between0
and1
(weights for the first layer). - Create a
1x4
tensor filled with zeros (biases for the first layer). - Create a
4x2
tensor with random integers between-5
and5
(weights for the second layer).
Solution
Everything was clear?
Thanks for your feedback!
Awesome!
Completion rate improved to 5SectionΒ 1. ChapterΒ 6
single