{"id":23464113,"url":"https://github.com/gwtaylor/theano-rnn","last_synced_at":"2025-04-06T12:09:03.809Z","repository":{"id":3910061,"uuid":"4998872","full_name":"gwtaylor/theano-rnn","owner":"gwtaylor","description":"Demonstration of recurrent neural network implemented with Theano","archived":false,"fork":false,"pushed_at":"2017-04-18T08:53:56.000Z","size":43,"stargazers_count":377,"open_issues_count":7,"forks_count":164,"subscribers_count":42,"default_branch":"master","last_synced_at":"2025-03-30T10:09:42.471Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"bsd-3-clause","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/gwtaylor.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"License.md","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null}},"created_at":"2012-07-12T04:17:01.000Z","updated_at":"2024-09-06T08:34:42.000Z","dependencies_parsed_at":"2022-09-09T22:40:47.451Z","dependency_job_id":null,"html_url":"https://github.com/gwtaylor/theano-rnn","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/gwtaylor%2Ftheano-rnn","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/gwtaylor%2Ftheano-rnn/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/gwtaylor%2Ftheano-rnn/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/gwtaylor%2Ftheano-rnn/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/gwtaylor","download_url":"https://codeload.github.com/gwtaylor/theano-rnn/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":247478323,"owners_count":20945266,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2024-12-24T09:25:35.938Z","updated_at":"2025-04-06T12:09:03.791Z","avatar_url":"https://github.com/gwtaylor.png","language":"Python","funding_links":[],"categories":["Codes"],"sub_categories":[],"readme":"#theano-rnn\n\nDemonstration of recurrent neural network implemented with Theano\n\n## Dependencies\n\n* [Theano](http://deeplearning.net/software/theano/)\n* [Scikit-learn](http://scikit-learn.org/stable/)\nThis relies on scikits-learn simply because I subclass their BaseEstimator\nclass, but this dependency could easily be removed.\n* A reasonably good Python distribution with numpy and scipy. I\nrecommend [Enthought](http://enthought.com/) because it is heavily optimized and it has a free\nacademic license.\n* If you want to use the Hessian-Free optimizer then you will also need:\n[theano-hf](https://github.com/boulanni/theano-hf)\n\n## Description\n\n* rnn.py : this is the most basic implementation of a \"vanilla\" RNN. It\nis designed for readability. It processes each sequence at a time.\nThere are three test functions which show how to train an RNN with\nreal-valued, binary or softmax outputs using stochastic gradient\ndescent.\n* rnn_minibatch.py : this is similar to rnn.py but it is slightly more\ncomplicated code and processes multiple sequences at once for speed\n(i.e. in \"mini-batches\"). It also hooks into scipy.optimize to use\nmore sophisticated optimization methods. It again includes three test\nfunctions based on output type.\n* hf_example.py: this uses the class defined by rnn.py but instead of\ntraining it with stochastic gradient descent, it trains it with\n[Martens and Sutskever's variant of Hessian-Free optimization](http://www.cs.toronto.edu/~jmartens/docs/RNN_HF.pdf).\n\n## Other implementations\n\nThere are other Theano rnn implementations publicly available, for example:\n* [Razvan Pascanu's implementation](https://github.com/pascanur/trainingRNNs).\n* Razvan also provides a [simple sketch](http://groups.google.com/group/theano-users/browse_thread/thread/39c755b93675f437). This was the starting point for this code on the Theano-users list. \n* [Matthias Zoehrer's implementation](https://github.com/mzoehr/Theano/tree/rnn_benchmark/benchmark/rnn).\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fgwtaylor%2Ftheano-rnn","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fgwtaylor%2Ftheano-rnn","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fgwtaylor%2Ftheano-rnn/lists"}