Nick Becker
Nick Becker
Is this ready for another round of reviews?
Changing the title before merging, as this PR only applies this change to random forest models.
This will also close https://github.com/rapidsai/cuml/issues/691
@carterrees yes, you can do this by starting a Dask CUDA cluster and setting `use_dask=True`. This brief recording shows an example. https://www.youtube.com/watch?v=7z4OJQdY_mw ```python from dask.distributed import Client from dask_cuda import...
Hi @GitAronas . It looks like you're using Windows, based on your screenshot. The cuML configuration requires the library cuML, which is not available natively on Windows. cuML requires Ubuntu...
You should use `n_jobs=1` (the default). cuML is currently designed for the "one process per GPU" paradigm". Additionally, how are you setting up your Dask cluster? It might be valuable...
Hi @chris1610 ! I just came across this issue due to the cuDF mention. This looks like a really cool project. Happy to help answer any questions about cuDF if...
In case it's useful, KMeans and KNN are also implemented in [cuML](https://github.com/rapidsai/cuml). Both algorithms are supported multi-GPU and multi-node. If you'd like to try it, I recommend using the [RAPIDS...
> Problem description: When there are a large number of string columns, data in feather format is much smaller than that in arrow format. When data is transmitted to the...
Writing the file out to Feather will compress the data using LZ4 by default. Writing out to Parquet will compress the data using snappy by default, but you can select...