-
I was trying to use my second-order optimizer [ESGD-M](https://github.com/crowsonkb/esgd) with BatchL2Grad in order to collect information on within-batch gradient variance to estimate stochastic nois…
-
add a standalone, reusable function with options
multivariate version is used in discrete_margins, we had brief discussion, my version is in miscmodels
one issues is how to handle different multi-di…
-
How much work would it be to add something like `valhessian!(H, schain, x, p)`? Where `H` is the Hessian of a simple chain with loss.
jbrea updated
7 months ago
-
## 🐛 Bug
Hi everyone, I found that autograd.backward() doesn’t trigger reduction under certain circumstance when using DDP.
@mrshenli @rohan-varma @pritamdamania
## To Reproduce
Here is a min…
-
## Description
The ODE functors aren't compatible with `fvar` inputs and fail to compile
#### Current Version:
v4.4.0
-
Not sure if this is the right spot, sorry if not!
I would like to state (in my case for an optimization problem I would like to solve with ipopt) the equality constraints as a vector.
No problem doi…
-
@S-Dafarra asked about this. This derivative have the nice properties that everything boils down to a lot of 6D cross product between all the joint motion subspace vectors.
References on this:
*…
-
Hi, when I installed pyke, I can not import it. I used both the two methods in pyke.keplerscience.org/install.html. What's the problem? Thanks for your help!
from pyke import kepextract
---------…
-
Hello,
thank you for this fantastic product. The more I am using it, the more I'm enjoying it. But I'm new and not aware of all the intricacies. So I am bumped again. I have two questions:
a) is t…
-
I'd like to add a `Matrix` class to complement `Vector`.
A key design question is what this needs to support. In particular: do we need to support multiple axes that correspond to flattened pytrees…