medium10 pts · first answer countspytorchtensor-shapes
Shape after a batched matmul
What does this code print?
import torch a = torch.randn(8, 10, 64) w = torch.randn(64, 32) print((a @ w).shape)
Discussion 0
Join free to post in the discussion.
Related questions
- Dropout after eval()Machine Learning Engineer
- Different results every inference runMachine Learning Engineer
- Gradient of a sum of squaresMachine Learning Engineer
- Training loop without zero_gradMachine Learning Engineer
- Softmax before CrossEntropyLossMachine Learning Engineer
- view() after a transposeMachine Learning Engineer