-
At the moment we use lbfgs_b from scipy as the default optimizer in GPflow.
Particularly with #185 and using float32 and float64 we have some anecdotal evidence that lbfgs_b is aggressively finding …
-
### 🚀 The feature, motivation and pitch
I have a python code that solve linear equations with torch.optim.LBFGS. And I want to make it work in C++. One posible way is to use libtorch. But I wander …
-
running `test_swa.py` on a device with multiple GPUs results in the following:
```
Test output:
> test_swa (test.test_swa.TestSWA) ... ERROR
>
> ================================================…
-
I played around a bit to test the `vmlmb` method for `DeconvOptim.jl`. It worked nicely, but I did not really find any advantage compared to the standard LBFGS implementation provided by `Opim.jl`. In…
-
Dheevatsa, have you seen any errors of the following sort?
```
Traceback (most recent call last):
File "lbfgs_tests.py", line 114, in
obj, grad, lr, backtracks, clos_evals, grad_evals, fail…
-
It would be useful for testing numerical inference algorithms to have differentiable likelihoods in lsbi. In theory I think the whole package can swap to jax, however things like rng are quite differe…
-
For this minimal example of fitting a 1-parameter poissonian likelihood `Pois(10|par)` where you would expect the best-fit value to be 10.0 tfp seems to fail:
```
tf.reset_default_graph()
sess = …
-
## ❓ Questions and Help
I'm currently using Ceres with the following options:
```cpp
linear_solver_type = ceres::DENSE_NORMAL_CHOLESKY;
minimizer_type = TRUST_REGION;
trust_region_strategy_type =…
Muon2 updated
4 months ago
-
We created a synthetic 1-D dataset that can be described as the following:
> We consider n =1, 024, xi i.i.d. ∼ N (0, 25) and yn ∼ N (0, σf**2 * Kf,n + σe**2 * In). Kf,n is an RBF kernel matrix wit…
-
Hi, I am new to GPytorch. I was trying a very simple example with training data
```
import torch
import gpytorch
from matplotlib import pyplot as plt
train_x = torch.arange(0, 100)
train_y =…