-
On Mac OS X with a Nvidia GPU and CUDA, ``python setup.py bdist_wheel`` runs with no errors. I can see a series of copying commands are executed:
```
creating build/bdist.macosx-10.12-intel/wheel
c…
-
Hi!
I used the Fortran-code of LBFGSB by Nocedal et al frequently in the past and tried out your code.
My question is if it is possible to implement the version of Nocedal with strong Wolfe-Powell l…
-
Does anyone here have a workaround for repeated integer/categorical entries? For example, if I am searching the space
`dimensions = [(1,10), (1,10),["red","green","blue"]]`
With a local optimum of:
`…
-
Hi,
This is very cool tech! I was curious if any thought had been put into how to make this approach to scale to higher resolutions; either by a CPU implementation or better yet have it work across mu…
-
From Marcus on stan-users in response to a model and question from David C. Cohen:
- The optimal solution (from nnls) is very sparse, i.e., all but maybe a dozen coefficients are zero. This solution …
-
I understand that "kernel" will become deprecated in the new version. How can one specify the Gpy Matern ARD 5/2 kernel with the newest API?
`GPyOpt.methods.bayesian_optimization.BayesianOptimizat…
ghost updated
7 years ago
-
I'm trying to use Fminbox with LBFGS, but it gives me the following error:
``` julia
ERROR: unrecognized keyword argument "eta"
in optimize at /home/riseth/.julia/v0.4/Optim/src/fminbox.jl:188
```
…
-
All default parameters should be double checked, as I believe some of our choices are currently suboptimal.
- [x] gp_minimize: default kernel?
- [x] gp_minimize: default noise level alpha for the GP? …
-
The build_loss_objective is not focused enough from both the math and the user point of view.
First "build" should refer to the general class of problems being solved here: the NL objective function…
-
I figured I should update details in separate issues instead of the one big issue #200 .
I am planning to start off by testing on CUTEst unconstrained functions, through the CUTEst.jl package. I inst…