Machine learning

Current implementation of Go binding can not specify options.

GPUOptions struct is in internal package. And go generate doesn't work for protobuf directory. So we can't specify GPUOptions for NewSession.

The metrics are slowly leaving Datasets (they are being deprecated as we speak) to move to the Evaluate library. We are looking for contributors to help us with the move.

Normally, the migration should be as easy as replacing the import of load_metric from Datasets to the load function in Evaluate. See a u

🐛 Describe the bug

The function einsum_helper in torch/onnx/symbolic_opset12.py is a helper. It is not meant for public consumption. We should rename it to _einsum_helper and remove it from __all__

Versions

N/A

Describe the bug

The acceptable values for max_samples are (0, 1]. One possible option for class_weight is balanced_subsample. However, for values of max_samples near zero and class_weight='balanced_subsample', the model fails with an unhelpful error related to having an empty array as an indexer.

This is only likely to come up in a grid search, as it seems unlikely that some

This subject was initially discussed in https://discourse.julialang.org/t/efficient-way-to-split-string-at-specific-index/83115/17 regarding a way to split a string into equal-length substrings.

The solutions discussed involved (for a string composed of 10 blocks of 8 strings)

• using a comprehension split8(str::String) = [str[i+1:8] for i in 0:8:length(str)]
• using a comprehension and `S

Currently many more Python projects like dask and optuna are using Python type hints. With the Python package of xgboost gaining more and more features, we should also adopt mypy as a safe guard against some type errors and for better code documentation.

This is a nice first issue:

Add types/docments to the Learner.get_preds function. This function is essential to any fastai user and has almost no documentation. Add types and text to the variables as we have in many places now.