/hyperopt

Distributed Asynchronous Hyperparameter Optimization in Python

Primary LanguagePythonOtherNOASSERTION

Hyperopt: Distributed Hyperparameter Optimization

Build Status PyPI version Anaconda-Server Badge

Hyperopt is a Python library for serial and parallel optimization over awkward search spaces, which may include real-valued, discrete, and conditional dimensions.

Getting started

Install hyperopt from PyPI

pip install hyperopt

to run your first example

# define an objective function
def objective(args):
    case, val = args
    if case == 'case 1':
        return val
    else:
        return val ** 2

# define a search space
from hyperopt import hp
space = hp.choice('a',
    [
        ('case 1', 1 + hp.lognormal('c1', 0, 1)),
        ('case 2', hp.uniform('c2', -10, 10))
    ])

# minimize the objective over the space
from hyperopt import fmin, tpe
best = fmin(objective, space, algo=tpe.suggest, max_evals=100)

print best
# -> {'a': 1, 'c2': 0.01420615366247227}
print hyperopt.space_eval(space, best)
# -> ('case 2', 0.01420615366247227}

If you're a developer, clone this repository and install from source:

git clone https://github.com/jaberg/hyperopt.git
cd hyperopt && python setup.py develop &&  pip install -e '.[MongoTrials, SparkTrials, ATPE]'

Algorithms

Currently three algorithms are implemented in hyperopt:

Hyperopt has been designed to accommodate Bayesian optimization algorithms based on Gaussian processes and regression trees, but these are not currently implemented.

All algorithms can be parallelized in two ways, using:

Documentation

Hyperopt documentation can be found here, but is partly still hosted on the wiki. Here are some quick links to the most relevant pages:

Examples

See projects using hyperopt on the wiki.

Announcements mailing list

Announcments

Discussion mailing list

Discussion

Cite

If you use this software for research, plase cite the following paper:

Bergstra, J., Yamins, D., Cox, D. D. (2013) Making a Science of Model Search: Hyperparameter Optimization in Hundreds of Dimensions for Vision Architectures. To appear in Proc. of the 30th International Conference on Machine Learning (ICML 2013).

Thanks

This project has received support from

  • National Science Foundation (IIS-0963668),
  • Banting Postdoctoral Fellowship program,
  • National Science and Engineering Research Council of Canada (NSERC),
  • D-Wave Systems, Inc.