Skip to main content
Help us improve Python packaging – donate today!

Optimizing compiler for evaluating mathematical expressions on CPUs and GPUs.

Project Description

Theano is a Python library that allows you to define, optimize, and efficiently evaluate mathematical expressions involving multi-dimensional arrays. It is built on top of NumPy. Theano features:

  • tight integration with NumPy: a similar interface to NumPy’s. numpy.ndarrays are also used internally in Theano-compiled functions.
  • transparent use of a GPU: perform data-intensive computations up to 140x faster than on a CPU (support for float32 only).
  • efficient symbolic differentiation: Theano can compute derivatives for functions of one or many inputs.
  • speed and stability optimizations: avoid nasty bugs when computing expressions such as log(1 + exp(x)) for large values of x.
  • dynamic C code generation: evaluate expressions faster.
  • extensive unit-testing and self-verification: includes tools for detecting and diagnosing bugs and/or potential problems.

Theano has been powering large-scale computationally intensive scientific research since 2007, but it is also approachable enough to be used in the classroom (IFT6266 at the University of Montreal).

Release Notes

Theano 1.0.1 (6th of December, 2017)

This is a maintenance release of Theano, version 1.0.1, with no new features, but some important bug fixes.

We recommend that everybody update to this version.

Highlights (since 1.0.0):

  • Fixed compilation and improved float16 support for topK on GPU
    • NB: topK support on GPU is experimental and may not work for large input sizes on certain GPUs
  • Fixed cuDNN reductions when axes to reduce have size 1
  • Attempted to prevent re-initialization of the GPU in a child process
  • Fixed support for temporary paths with spaces in Theano initialization
  • Spell check pass on the documentation

A total of 6 people contributed to this release since 1.0.0:

  • Frederic Bastien
  • Steven Bocco
  • Arnaud Bergeron
  • Sam Johnson
  • Edward Betts
  • Simon Lefrancois

Release history Release notifications

This version
History Node

1.0.1

History Node

1.0.0

History Node

1.0.0rc1

History Node

0.10.0beta4

History Node

0.10.0beta3

History Node

0.10.0b2

History Node

0.10.0b1

History Node

0.9.0

History Node

0.9.0rc4

History Node

0.9.0rc3

History Node

0.9.0rc2

History Node

0.9.0rc1

History Node

0.9.0b1

History Node

0.8.2

History Node

0.8.1

History Node

0.8.0

History Node

0.8.0rc1

History Node

0.7.0

History Node

0.7.0rc2

History Node

0.7.0rc1

History Node

0.6.0

History Node

0.6.0rc5

History Node

0.6.0rc3

History Node

0.6.0rc2

History Node

0.6.0rc1

History Node

0.5.0

History Node

0.5.0rc2

History Node

0.5.0rc1

History Node

0.4.1

History Node

0.4.1rc2

History Node

0.4.1rc1

History Node

0.4.0

History Node

0.4.0rc4

History Node

0.4.0rc3

History Node

0.4.0rc2

History Node

0.4.0rc1

History Node

0.3.1

History Node

0.3.1rc2

History Node

0.3.1rc1

History Node

0.3.0

History Node

0.3.0rc4

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Filename, size & hash SHA256 hash help File type Python version Upload date
Theano-1.0.1.tar.gz (2.8 MB) Copy SHA256 hash SHA256 Source None Dec 7, 2017

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging CloudAMQP CloudAMQP RabbitMQ AWS AWS Cloud computing Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page