|
What is the vectorized gradient formula with label smoothing? |
Answered by
Ultron09
Aug 28, 2026
Replies: 1 comment
|
The gradient simplifies to Softmax(logits) - one_hot(targets), with zero auxiliary tensor allocations. |
0 replies
Answer selected by
Ultron09
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
The gradient simplifies to Softmax(logits) - one_hot(targets), with zero auxiliary tensor allocations.