Optimizer precision - #135
Conversation
|
@marcelluethi As discussed today, I will add tests before we merge. |
|
Thanks for implementing this. This looks like a very good solution to me. |
…pendent on params
eac7a09 to
b9f60f1
Compare
| def --(using TensorTree[P], FloatTree[P, V]): P = Tensor0(VType[V])(p2) --! p1 | ||
| def **(using TensorTree[P], FloatTree[P, V]): P = Tensor0(VType[V])(p2) **! p1 | ||
| def `//!`[P](p1: P)(using TensorTree[P], FloatTree[P, V]): P = Tensor0(VType[V])(p2) `//!` p1 | ||
|
|
There was a problem hiding this comment.
Will this not be confusing on single tensors, as they are both tensor trees and tensors and consequently will have +, +! and ++!. Another problem might be that ! usually indicates broadcasting and has here a different meaning.
marcelluethi
left a comment
marcelluethi
left a comment
There was a problem hiding this comment.
Thanks for the fix. I think it is now the right way to handle the precision of the optimizers. I also like the optimizer parameters are just Double and hence easy to fix. The only thing I am not sure about are the !++ extension methods. However, I am fine with merging if this is not an accident but a conscious decision.
2 participants
No description provided.