pub fn dequantize_device(
w: impl AsRef<Array>,
scales: impl AsRef<Array>,
biases: impl AsRef<Array>,
group_size: impl Into<Option<i32>>,
bits: impl Into<Option<i32>>,
stream: impl AsRef<Stream>,
) -> Result<Array>
Expand description
Dequantize the matrix w
using the provided scales
and biases
and the group_size
and
bits
configuration.
For details, please see this documentation