-
Hi Paul,
I'm trying to allow the full variance-covariance matrix of the random effects to vary across groups using the gr(re, by = "cat_var") syntax (from issue [365](https://github.com/paul-buerkn…
-
I think in the mobilenet paper, it is mentioned that they trained in a similar way as the inception v3. In the inception V3 paper, they used a learning rate of learning rate decay of 0.94 every 2 epoc…
-
### Motivation: Why do you think this is important?
When using conditionals in a workflow, I found that the graph doesn't represent correctly the inputs and outputs dependencies between tasks. To the…
-
Downloading the hparams table as a csv omits the Trial ID column. It would be helpful if this was included along with the logged hyperparameters and metrics for completeness.
-
GPNN trains a model using a randomly subsampled dataset to extract the optimal hyperparameters. Then for each test data, they use the nearest neighbor to get prediction. This work aims to study the sc…
-
when `growth_rate_deriv` effect is selected in https://github.com/idem-lab/epiwave/blob/main/R/create_infection_timeseries.R
model loses traction over GP hyperparameters, leading to model failing t…
-
/kind feature
**Describe the solution you'd like**
[A clear and concise description of what you want to happen.]
When describing a katib experiment page, there is one graph showing all trials w…
-
Hello,
unfortunately, my results are nowhere close to the demo clip shown https://www.youtube.com/watch?v=c5VG7HkDs8I.
- lips not moving, worse than reference papers, blurry.
- The movements of th…
-
Hello author. I was unable to reproduce the results in the paper.
I used the hyperparameters of provided in the github repository as well as the hyperparameters provided in the original article for …
-
When the eval() function in the acquisition function class involves gradient backward, because line 102 of hebo/acq_optimizers/evolution_optimizer.py is ''with torch.no_grad():'', the gradient cannot …