-
Using the scikit-learn main branch:
```python-traceback
Traceback (most recent call last):
File "/usr/share/miniconda/envs/sklbench/lib/python3.8/runpy.py", line 194, in _run_module_as_main
…
-
Efficient approximate kNN-graph construction, and efficient approximate kNN search using such a graph would be of significant potential benefit for scikit-learn. Such methods are independent of data d…
-
**Describe the bug**
Circular import error with PyTorch nightly. If I uninstall deepspeed it works fine.
```
Traceback (most recent call last):
File "/test/oss.py", line 322, in
mp.spawn…
-
### Summary
With #24678, we make it easier for the `Criterion` class to be inherited. However, the Splitter class can also leverage this improvement. We should separate the current `Splitter` class i…
-
This issue suggests some ideas, which you might use to improve your testing. Normally writing tests can be a very time-consuming task and it is crucial to have a large number of tests for good coverag…
-
### Summary
There exists other criterions, where additional arguments would need to be passed in. The current `Criterion` class is not actually a generic Criterion class since it explicitly assumes…
-
Which models for time series forecasting fedot uses? Can you provide examples? Is there any comparison with competitors? @RepoPilotAssistant please help
-
It would be interesting to test/benchmark the ball tree + haversine approach versus the kd-tree + conversion lat/lon spherical -> xyz cartersian.
The second approach is also used by libpysal, see […
-
### Describe the workflow you want to enable
I would like the team to add an argument(s) for using only part of the entire dataset to train the KNNImputer.
### Describe your proposed solution
This …
-
**>**
We need to have a performance test suite in KFServing. Performance tests will help us identify potential bottlenecks and create reproducible resources for KFServing production users.
**Goa…