Tensor operations (add, mul, matmul) in PyTorch - ML Experiment: Train & Evaluate

Experiment - Tensor operations (add, mul, matmul)

Problem:You want to perform basic tensor operations like addition, element-wise multiplication, and matrix multiplication using PyTorch tensors.

Current Metrics:No current metrics as this is a basic tensor operation task.

Issue:You are unsure how to correctly use PyTorch functions for add, mul, and matmul operations and want to see their outputs.

Your Task

Write PyTorch code to create two tensors and perform addition, element-wise multiplication, and matrix multiplication. Verify the outputs are correct.

Use PyTorch tensors only.

Use built-in PyTorch functions or operators for add, mul, and matmul.

Print the results to verify correctness.

Hint 1

Hint 2

Hint 3

Hint 4

Hint 5

Solution

PyTorch

import torch

# Create two tensors
A = torch.tensor([[1, 2], [3, 4]])
B = torch.tensor([[5, 6], [7, 8]])

# Addition
add_result = A + B

# Element-wise multiplication
mul_result = A * B

# Matrix multiplication
matmul_result = torch.matmul(A, B)

# Print results
print('Addition result:\n', add_result)
print('Element-wise multiplication result:\n', mul_result)
print('Matrix multiplication result:\n', matmul_result)

Created two 2x2 tensors A and B.

Performed addition using + operator.

Performed element-wise multiplication using * operator.

Performed matrix multiplication using torch.matmul().

Printed all results to verify correctness.

Results Interpretation

Before: No tensor operations performed.

After: Addition, element-wise multiplication, and matrix multiplication results are correctly computed and printed.

This experiment shows how to use PyTorch tensor operations for basic arithmetic and matrix multiplication, which are foundational for building neural networks and other ML models.

Bonus Experiment

Try performing the same operations on tensors with different shapes and observe what errors or broadcasting behaviors occur.

💡 Hint

Use tensors of shape (2,3) and (3,2) and try addition and matrix multiplication to see shape compatibility requirements.