# hpt
> This repository is under construction :construction:
A minimal hyperparameter tuning framework to help you train hundreds of models.
It's essentially a set of helpful wrappers over optuna.
## Install
Install package from [PyPI](https://pypi.org/project/hyperparameter-tuning/):
`
pip install hyperparameter-tuning
`
## Getting started
```py
from hpt.tuner import ObjectiveFunction, OptunaTuner
obj_func = ObjectiveFunction(
X_train, y_train, X_test, y_test,
hyperparameter_space=HYPERPARAM_SPACE_PATH, # path to YAML file
eval_metric="accuracy",
s_train=s_train,
s_val=s_test,
threshold=0.50,
)
tuner = OptunaTuner(
objective_function=obj_func,
direction="maximize", # NOTE: can pass other useful study kwargs here (e.g. storage)
)
# Then just run optimize as you would for an optuna.Study object
tuner.optimize(n_trials=20, n_jobs=4)
# Results are stored in tuner.results
tuner.results
# You can reconstruct the best predictor with:
clf = obj_func.reconstruct_model(obj_func.best_trial)
```
## Defining a hyperparameter space
The hyperparameter space is provided either path to a YAML file, or as a `dict`
with the same structure.
Example hyperparameter spaces [here](examples/sklearn.small_hyperparam_space.yaml) and
[here](examples/sklearn.large_hyperparam_space.yaml).
The YAML file must follow this structure:
```yaml
# One or more top-level algorithms
DT:
# Full classpath of algorithm's constructor
classpath: sklearn.tree.DecisionTreeClassifier
# One or more key-word arguments to be passed to the constructor
kwargs:
# Kwargs may be sampled from a distribution
max_depth:
type: int # either 'int' or 'float'
range: [ 10, 100 ] # minimum and maximum values
log: True # (optionally) whether to use logarithmic scale
# Kwargs may be sampled from a fixed set of categories
criterion:
- 'gini'
- 'entropy'
# Kwargs may be a pre-defined value
min_samples_split: 4
# You may explore multiple algorithms at once
LR:
classpath: sklearn.linear_model.LogisticRegression
kwargs:
# An example of a float hyperparameter
C:
type: float
range: [ 0.01, 1.0 ]
log: True
```
Raw data
{
"_id": null,
"home_page": "https://github.com/AndreFCruz/hpt",
"name": "hyperparameter-tuning",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.8",
"maintainer_email": "",
"keywords": "ml,optimization,hyperparameter,tuning,fairness",
"author": "AndreFCruz",
"author_email": "",
"download_url": "https://files.pythonhosted.org/packages/c8/c5/40907d79c62e8528a9bfa21ee1776f0b66d63a5e7da7287fbfca0ff92c97/hyperparameter-tuning-0.3.1.tar.gz",
"platform": null,
"description": "# hpt\n\n> This repository is under construction :construction:\n\n\n\n\nA minimal hyperparameter tuning framework to help you train hundreds of models.\n\nIt's essentially a set of helpful wrappers over optuna.\n\n\n## Install\n\nInstall package from [PyPI](https://pypi.org/project/hyperparameter-tuning/):\n\n\n`\npip install hyperparameter-tuning\n`\n\n## Getting started\n\n```py\nfrom hpt.tuner import ObjectiveFunction, OptunaTuner\n\nobj_func = ObjectiveFunction(\n X_train, y_train, X_test, y_test,\n hyperparameter_space=HYPERPARAM_SPACE_PATH, # path to YAML file\n eval_metric=\"accuracy\",\n s_train=s_train,\n s_val=s_test,\n threshold=0.50,\n)\n\ntuner = OptunaTuner(\n objective_function=obj_func,\n direction=\"maximize\", # NOTE: can pass other useful study kwargs here (e.g. storage)\n)\n\n# Then just run optimize as you would for an optuna.Study object\ntuner.optimize(n_trials=20, n_jobs=4)\n\n# Results are stored in tuner.results\ntuner.results\n\n# You can reconstruct the best predictor with:\nclf = obj_func.reconstruct_model(obj_func.best_trial)\n```\n\n## Defining a hyperparameter space\n\nThe hyperparameter space is provided either path to a YAML file, or as a `dict` \nwith the same structure.\nExample hyperparameter spaces [here](examples/sklearn.small_hyperparam_space.yaml) and \n[here](examples/sklearn.large_hyperparam_space.yaml).\n\nThe YAML file must follow this structure:\n```yaml\n# One or more top-level algorithms\nDT: \n # Full classpath of algorithm's constructor\n classpath: sklearn.tree.DecisionTreeClassifier\n \n # One or more key-word arguments to be passed to the constructor\n kwargs:\n \n # Kwargs may be sampled from a distribution\n max_depth:\n type: int # either 'int' or 'float'\n range: [ 10, 100 ] # minimum and maximum values\n log: True # (optionally) whether to use logarithmic scale\n \n # Kwargs may be sampled from a fixed set of categories\n criterion:\n - 'gini'\n - 'entropy'\n \n # Kwargs may be a pre-defined value\n min_samples_split: 4\n\n\n# You may explore multiple algorithms at once\nLR:\n classpath: sklearn.linear_model.LogisticRegression\n kwargs:\n # An example of a float hyperparameter\n C:\n type: float\n range: [ 0.01, 1.0 ]\n log: True\n\n```\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "A minimal framework for running hyperparameter tuning",
"version": "0.3.1",
"project_urls": {
"Homepage": "https://github.com/AndreFCruz/hpt"
},
"split_keywords": [
"ml",
"optimization",
"hyperparameter",
"tuning",
"fairness"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "ef0636f0d4a9188ec9654a1ab96a9c3c4c1618d0a2cb566f76120e5cf89f451c",
"md5": "9e64f4a4b4c46964389cf79a1438c928",
"sha256": "df861403bbaad3263c826e369a215e76a2c8b038c85e856d8717343f639692da"
},
"downloads": -1,
"filename": "hyperparameter_tuning-0.3.1-py3-none-any.whl",
"has_sig": false,
"md5_digest": "9e64f4a4b4c46964389cf79a1438c928",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.8",
"size": 25935,
"upload_time": "2023-09-05T09:38:51",
"upload_time_iso_8601": "2023-09-05T09:38:51.987393Z",
"url": "https://files.pythonhosted.org/packages/ef/06/36f0d4a9188ec9654a1ab96a9c3c4c1618d0a2cb566f76120e5cf89f451c/hyperparameter_tuning-0.3.1-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "c8c540907d79c62e8528a9bfa21ee1776f0b66d63a5e7da7287fbfca0ff92c97",
"md5": "1f922c9a889393ade8031bf5d65b76b8",
"sha256": "451e0b8a51e989d63913eba9bdc162224fe010401f38fdc051c5fdd343f23613"
},
"downloads": -1,
"filename": "hyperparameter-tuning-0.3.1.tar.gz",
"has_sig": false,
"md5_digest": "1f922c9a889393ade8031bf5d65b76b8",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.8",
"size": 24211,
"upload_time": "2023-09-05T09:38:53",
"upload_time_iso_8601": "2023-09-05T09:38:53.530458Z",
"url": "https://files.pythonhosted.org/packages/c8/c5/40907d79c62e8528a9bfa21ee1776f0b66d63a5e7da7287fbfca0ff92c97/hyperparameter-tuning-0.3.1.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-09-05 09:38:53",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "AndreFCruz",
"github_project": "hpt",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "hyperparameter-tuning"
}
JSON
