You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository was archived by the owner on Nov 14, 2023. It is now read-only.
Adds auto-generated docs to make this library self-sufficient rather than relying on links in the ray docs.
---------
Signed-off-by: Justin Yu <[email protected]>
Copy file name to clipboardExpand all lines: README.md
+18-7Lines changed: 18 additions & 7 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -11,13 +11,13 @@ Here’s what tune-sklearn has to offer:
11
11
***Framework support**: tune-sklearn is used primarily for tuning Scikit-Learn models, but it also supports and provides examples for many other frameworks with Scikit-Learn wrappers such as Skorch (Pytorch) [[example](https://github.com/ray-project/tune-sklearn/blob/master/examples/torch_nn.py)], KerasClassifier (Keras) [[example](https://github.com/ray-project/tune-sklearn/blob/master/examples/keras_example.py)], and XGBoostClassifier (XGBoost) [[example](https://github.com/ray-project/tune-sklearn/blob/master/examples/xgbclassifier.py)].
12
12
***Scale up**: Tune-sklearn leverages [Ray Tune](http://tune.io/), a library for distributed hyperparameter tuning, to parallelize cross validation on multiple cores and even multiple machines without changing your code.
13
13
14
-
Check out our [API Documentation](https://docs.ray.io/en/master/tune/api_docs/sklearn.html) and [Walkthrough](https://docs.ray.io/en/master/tune/tutorials/tune-sklearn.html) (for `master` branch).
14
+
Check out our [API Documentation](docs) and [Walkthrough](https://docs.ray.io/en/master/tune/examples/tune-sklearn.html) (for `master` branch).
15
15
16
16
## Installation
17
17
18
18
### Dependencies
19
19
- numpy (>=1.16)
20
-
-[ray](http://docs.ray.io/)
20
+
-[ray](http://docs.ray.io/) (>=2.7.0)
21
21
- scikit-learn (>=0.23)
22
22
23
23
### User Installation
@@ -48,10 +48,10 @@ See [the Ray documentation for an overview of available stoppers](https://docs.r
48
48
49
49
## Examples
50
50
51
-
#### TuneGridSearchCV
51
+
#### [TuneGridSearchCV](docs/tune_gridsearch.md)
52
52
To start out, it’s as easy as changing our import statement to get Tune’s grid search cross validation interface, and the rest is almost identical!
53
53
54
-
`TuneGridSearchCV` accepts dictionaries in the format `{ param_name: str : distribution: list }` or a list of such dictionaries, just like scikit-learn's `GridSearchCV`. The distribution can also be the output of Ray Tune's [`tune.grid_search`](https://docs.ray.io/en/master/tune/api_docs/search_space.html#grid-search-api).
54
+
`TuneGridSearchCV` accepts dictionaries in the format `{ param_name: str : distribution: list }` or a list of such dictionaries, just like scikit-learn's `GridSearchCV`. The distribution can also be the output of Ray Tune's [`tune.grid_search`](https://docs.ray.io/en/master/tune/api/search_space.html).
55
55
56
56
```python
57
57
# from sklearn.model_selection import GridSearchCV
`TuneSearchCV` is an upgraded version of scikit-learn's `RandomizedSearchCV`.
116
116
117
-
It also provides a wrapper for several search optimization algorithms from Ray Tune's [`tune.suggest`](https://docs.ray.io/en/master/tune/api_docs/suggestion.html), which in turn are wrappers for other libraries. The selection of the search algorithm is controlled by the `search_optimization` parameter. In order to use other algorithms, you need to install the libraries they depend on (`pip install` column). The search algorithms are as follows:
117
+
It also provides a wrapper for several search optimization algorithms from Ray Tune's [searchers](https://docs.ray.io/en/master/tune/api/suggestion.html), which in turn are wrappers for other libraries. The selection of the search algorithm is controlled by the `search_optimization` parameter. In order to use other algorithms, you need to install the libraries they depend on (`pip install` column). The search algorithms are as follows:
118
118
119
119
| Algorithm |`search_optimization` value | Summary | Website |`pip install`|
@@ -126,7 +126,7 @@ It also provides a wrapper for several search optimization algorithms from Ray T
126
126
127
127
All algorithms other than RandomListSearcher accept parameter distributions in the form of dictionaries in the format `{ param_name: str : distribution: tuple or list }`.
128
128
129
-
Tuples represent real distributions and should be two-element or three-element, in the format `(lower_bound: float, upper_bound: float, Optional: "uniform" (default) or "log-uniform")`. Lists represent categorical distributions. [Ray Tune Search Spaces](https://docs.ray.io/en/master/tune/api_docs/search_space.html) are also supported and provide a rich set of potential distributions. Search spaces allow for users to specify complex, potentially nested search spaces and parameter distributions. Furthermore, each algorithm also accepts parameters in their own specific format. More information in [Tune documentation](https://docs.ray.io/en/master/tune/api_docs/suggestion.html).
129
+
Tuples represent real distributions and should be two-element or three-element, in the format `(lower_bound: float, upper_bound: float, Optional: "uniform" (default) or "log-uniform")`. Lists represent categorical distributions. [Ray Tune Search Spaces](https://docs.ray.io/en/master/tune/api/search_space.html) are also supported and provide a rich set of potential distributions. Search spaces allow for users to specify complex, potentially nested search spaces and parameter distributions. Furthermore, each algorithm also accepts parameters in their own specific format. More information in [Tune documentation](https://docs.ray.io/en/master/tune/api/suggestion.html).
130
130
131
131
Random Search (default) accepts dictionaries in the format `{ param_name: str : distribution: list }` or a list of such dictionaries, just like scikit-learn's `RandomizedSearchCV`.
132
132
@@ -191,5 +191,16 @@ Tune-sklearn also supports the use of other machine learning libraries such as P
-[`list_searcher`](./list_searcher.md#module-list_searcher): Helper class to support passing a
8
+
-[`tune_basesearch`](./tune_basesearch.md#module-tune_basesearch): Parent class for a cross-validation interface
9
+
-[`tune_gridsearch`](./tune_gridsearch.md#module-tune_gridsearch): Class for doing grid search over lists of hyperparameters
10
+
-[`tune_search`](./tune_search.md#module-tune_search): Class for cross-validation over distributions of hyperparameters
11
+
-[`utils`](./utils.md#module-utils)
12
+
13
+
## Classes
14
+
15
+
-[`list_searcher.ListSearcher`](./list_searcher.md#class-listsearcher): Custom search algorithm to support passing in a list of
16
+
-[`list_searcher.RandomListSearcher`](./list_searcher.md#class-randomlistsearcher): Custom search algorithm to support passing in a list of
17
+
-[`tune_basesearch.TuneBaseSearchCV`](./tune_basesearch.md#class-tunebasesearchcv): Abstract base class for TuneGridSearchCV and TuneSearchCV
18
+
-[`tune_gridsearch.TuneGridSearchCV`](./tune_gridsearch.md#class-tunegridsearchcv): Exhaustive search over specified parameter values for an estimator.
19
+
-[`tune_search.TuneSearchCV`](./tune_search.md#class-tunesearchcv): Generic, non-grid search on hyper parameters.
20
+
-[`utils.EarlyStopping`](./utils.md#class-earlystopping): An enumeration.
0 commit comments