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).
.. _NumPy: http://numpy.scipy.org/
=============
Release Notes
=============
Theano 1.0.5 (27th of July 2020)
================================
This is a maintenance release of Theano, version ``1.0.5``, with no
new features, but some deprecation fixes.
We recommend that everybody update to this version.
Highlights (since 1.0.4):
- Theano is now compatible with Python 3.9
- Fixed many deprecation warnings
A total of 13 people contributed to this release since ``1.0.4``:
- 1fish2
- Frederic Bastien
- Rebecca Palmer
- Miro Hrončok
- Dan Foreman-Mackey
- Adrian Seyboldt
- abergeron
- Tim Gates
- Tim Odonnell
- Robert P. Goldman
- Duc Nguyen
- Igor Varfolomeev
- Thomas Wiecki
Raw data
{
"_id": null,
"home_page": "http://deeplearning.net/software/theano/",
"name": "Theano",
"maintainer": "",
"docs_url": null,
"requires_python": "",
"maintainer_email": "",
"keywords": "theano math numerical symbolic blas numpy gpu autodiff differentiation",
"author": "LISA laboratory, University of Montreal",
"author_email": "theano-dev@googlegroups.com",
"download_url": "https://files.pythonhosted.org/packages/6b/97/bcd5654ba60f35f180931afabbd3b4c46c0379852f961c7a2819ff897f5d/Theano-1.0.5.tar.gz",
"platform": "Windows",
"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:\n\n * **tight integration with NumPy:** a similar interface to NumPy's. numpy.ndarrays are also used internally in Theano-compiled functions.\n * **transparent use of a GPU:** perform data-intensive computations up to 140x faster than on a CPU (support for float32 only).\n * **efficient symbolic differentiation:** Theano can compute derivatives for functions of one or many inputs.\n * **speed and stability optimizations:** avoid nasty bugs when computing expressions such as log(1 + exp(x)) for large values of x.\n * **dynamic C code generation:** evaluate expressions faster.\n * **extensive unit-testing and self-verification:** includes tools for detecting and diagnosing bugs and/or potential problems.\n\nTheano has been powering large-scale computationally intensive scientific\nresearch since 2007, but it is also approachable enough to be used in the\nclassroom (IFT6266 at the University of Montreal).\n\n.. _NumPy: http://numpy.scipy.org/\n\n\n=============\nRelease Notes\n=============\n\nTheano 1.0.5 (27th of July 2020)\n================================\n\nThis is a maintenance release of Theano, version ``1.0.5``, with no\nnew features, but some deprecation fixes.\n\nWe recommend that everybody update to this version.\n\nHighlights (since 1.0.4):\n\n - Theano is now compatible with Python 3.9\n - Fixed many deprecation warnings\n\nA total of 13 people contributed to this release since ``1.0.4``:\n\n - 1fish2\n - Frederic Bastien\n - Rebecca Palmer\n - Miro Hron\u010dok\n - Dan Foreman-Mackey\n - Adrian Seyboldt\n - abergeron\n - Tim Gates\n - Tim Odonnell\n - Robert P. Goldman\n - Duc Nguyen\n - Igor Varfolomeev\n - Thomas Wiecki\n",
"bugtrack_url": null,
"license": "BSD",
"summary": "Optimizing compiler for evaluating mathematical expressions on CPUs and GPUs.",
"version": "1.0.5",
"project_urls": {
"Homepage": "http://deeplearning.net/software/theano/"
},
"split_keywords": [
"theano",
"math",
"numerical",
"symbolic",
"blas",
"numpy",
"gpu",
"autodiff",
"differentiation"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "6b97bcd5654ba60f35f180931afabbd3b4c46c0379852f961c7a2819ff897f5d",
"md5": "d9275643c4b9c5aef77ece8ec144fac9",
"sha256": "6e9439dd53ba995fcae27bf20626074bfc2fff446899dc5c53cb28c1f9202e89"
},
"downloads": -1,
"filename": "Theano-1.0.5.tar.gz",
"has_sig": false,
"md5_digest": "d9275643c4b9c5aef77ece8ec144fac9",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 2842778,
"upload_time": "2020-07-27T16:13:54",
"upload_time_iso_8601": "2020-07-27T16:13:54.262781Z",
"url": "https://files.pythonhosted.org/packages/6b/97/bcd5654ba60f35f180931afabbd3b4c46c0379852f961c7a2819ff897f5d/Theano-1.0.5.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2020-07-27 16:13:54",
"github": false,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"lcname": "theano"
}
JSON