[enc][dask] Support training continuation. #11609
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
trivialfis
force-pushed
the
enc-dask-doc
branch
from
ed18017 to
f011826
Compare
Evaluation. Fixes. Documents.
Co-authored-by: Copilot <[email protected]>
trivialfis
force-pushed
the
enc-dask-doc
branch
from
4de4b85 to
b7910a6
Compare
Contributor
There was a problem hiding this comment.
Pull Request Overview
This PR adds support for training continuation in the Dask interface with categorical encoding. The main goal is to enable XGBoost to preserve categorical encodings when continuing training from a previous model, ensuring consistent categorical handling across training sessions.
- Adds a
run_recodefunction and related testing infrastructure for categorical re-coding functionality - Refactors the dispatch system to use
DispatchAnyconsistently across CPU and CUDA implementations - Updates DMatrix creation functions to accept model parameters for extracting reference categories
Reviewed Changes
Copilot reviewed 37 out of 37 changed files in this pull request and generated 5 comments.
Show a summary per file
| File | Description |
|---|---|
| python-package/xgboost/testing/dask.py | Adds run_recode function to test categorical re-coding with training continuation |
| python-package/xgboost/dask/data.py | Updates DMatrix creation functions to handle model categories and reference categories |
| src/data/proxy_dmatrix.h | Refactors dispatch system and adds support for reference categories |
| src/data/proxy_dmatrix.cuh | Updates CUDA dispatch functions to use consistent DispatchAny interface |
| tests/test_distributed/test_with_dask/test_with_dask.py | Adds test for re-coding functionality |
Comments suppressed due to low confidence (1)
src/data/proxy_dmatrix.h:88
- The parameter type changed from pass-by-value to pointer without updating the function documentation. This is a breaking change that should be clearly documented or the parameter should maintain consistent passing semantics.
void SetColumnar(StringView data);
20 tasks
trivialfis
changed the title
Member
Author
|
cc @rongou . |
rongou
approved these changes
Aug 6, 2025
doc/tutorials/categorical.rst
Outdated
| Internally, XGBoost attempts to extract the categories from the dataframe inputs. For | ||
| inference (predict), the re-coding happens on the fly and there's no data copy (baring | ||
| from some internal transformations performed by the dataframe itself). For training |
doc/tutorials/categorical.rst
Outdated
| Starting with 3.1, the *Python* interface can remember the encoding and perform recoding | ||
| during inference and training continuation when the input is a dataframe (`pandas`, | ||
| `cuDF`, `polars`, `pyarrow`, `modin`). The feature support focuses on basic usage. It has | ||
| some restrictions on the types of inputs that can be accepted. Firstly, category names |
doc/tutorials/categorical.rst
Outdated
| - integer, from 8-bit to 64-bit, both signed and unsigned are supported. | ||
| - 32-bit or 64-bit floating point | ||
|
|
||
| Other category types are not supported. Secondly, the input types must be strictly |
doc/tutorials/categorical.rst
Outdated
| from some internal transformations performed by the dataframe itself). For training | ||
| continuation however, re-coding requires some extra steps if you are using the native | ||
| interface. The sklearn interface and the Dask interface can handle training continuation | ||
| automatically. Lastly, please note that using the re-coder with the native interface is |
Member
Author
There was a problem hiding this comment.
Thank you for going through the doc. I fixed the grammar based on your comments.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Ref #11088