Tensor.quick_gelu()
Returns the value of the Quick Gaussian Error Linear Unit function applied to the specified tensor.
Usage
from tinygrad.tensor import Tensor
a1 = Tensor([[1, 2, 3], [3, 4, 5]])
a1 = a1.quick_gelu()
print(a1.numpy())
Return value
[[0.8457958 1.9356586 2.9819288]
[2.9819288 3.995585 4.998993 ]]