Half precision spectograms - Mixed precision training
Tomsen1410 opened this issue · comments
Hey,
Is it sufficient to have the spectograms in half precision format (float16)? And did you train your model with mixed precision?
Hi there,
Is it sufficient to have the spectograms in half precision format (float16)?
I haven't used half-precision before, but I think it is worth having a try.
And did you train your model with mixed precision?
Not for SSAST as it seems torch.cuda.amp
doesn't directly support multiple forward
functions (e.g., mpg
, mpc
, etc in https://github.com/YuanGongND/ssast/blob/main/src/models/ast_models.py). But if you can find a workaround, I don't think that hurts the performance, as we use torch.cuda.amp
in our original AST project (single forward function), see
and