Vits batch size #1824
Answered
by
harmlessman
KyuubiYoru
asked this question in
General Q&A
Vits batch size
#1824
-
|
What effect has the batch size on the training result? |
Beta Was this translation helpful? Give feedback.
Answered by
harmlessman
Aug 4, 2022
Replies: 1 comment
-
|
Both the quality and speed of the model affect. If the batch size is too large, there is a risk of overfitting, and if the batch size is too small, there is a risk of noise. If the batch size is large, the speed is slow, and if the batch size is small, the speed is fast. The batch size is recommended between 32 and 64. |
Beta Was this translation helpful? Give feedback.
0 replies
Answer selected by
KyuubiYoru
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Both the quality and speed of the model affect.
If the batch size is too large, there is a risk of overfitting, and if the batch size is too small, there is a risk of noise.
If the batch size is large, the speed is slow, and if the batch size is small, the speed is fast.
The batch size is recommended between 32 and 64.