-
It would be beneficial to go away from using native pytorch `Dataset/Dataloader` and use pytorch lightning `DataModule` instead([see here](https://lightning.ai/docs/pytorch/stable/data/datamodule.html…
-
# Title
Uncertainty quantification: methods, use cases, challenges
# Description
Uncertainty quantification (UQ) methods enable us to equip ML model predictions with "error bars". These are u…
-
I have found your project is quite interesting for learning NER and RE. I have succeeded in training NER and RE. However I got the following error in running PipelineModel.py
```
python PipelineMo…
-
## Describe the bug ##
By default memory_limit is set to 3GB for machine learning models. I do get following error even when I use data of size ~3.5MB.
`ValueError: Dummy prediction failed with run…
-
I want to use the size of the vocabularies as a parameter for my transformations. For example,
```python
def preprocessing_fn(inputs, custom_config):
...
outputs["labels"] = tf.sparse.t…
-
**Describe the bug**
I have been trying to use Error Analysis on my XGBoost Regression model for quite some time now but it seems like I am unlucky when it comes to the dependency issues.
**To Rep…
-
Currently, our estimators are named to follow method names introduced in R's `yaImpute` package. We likely want to conform to scikit-learn's model of being more explicit about whether estimators are …
-
### Describe your issue.
In MNE-Python we run a pip-pre job with thhe latest `scipy-wheels-nightly` builds. The latest 1.11.0.dev0+1956.7c74503 pip-pre wheel appears to be buggy:
- https://githu…
-
Hello!
Using function `hcp.preprocessing.apply_ref_correction(raw)` on HCP dataset returns the following error:
```python
File ~/miniconda3/envs/XXXXX/lib/python3.8/site-packages/hcp/preprocessi…
-
I am interested in probabilistic GBM and was doing some research and found your PGBM repository.
So I read the README and according to the Feature overview, pgbm.torch.PGBM is compatible with NaN.
H…