{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/benchmarking-of-lstm-networks","title":"Benchmarking of LSTM Networks","arxiv_id":"1508.02774","date":"2015-08-11","proceeding":null,"authors":["Thomas M. Breuel"],"abstract":"LSTM (Long Short-Term Memory) recurrent neural networks have been highly\nsuccessful in a number of application areas. This technical report describes\nthe use of the MNIST and UW3 databases for benchmarking LSTM networks and\nexplores the effect of different architectural and hyperparameter choices on\nperformance. Significant findings include: (1) LSTM performance depends\nsmoothly on learning rates, (2) batching and momentum has no significant effect\non performance, (3) softmax training outperforms least square training, (4)\npeephole units are not useful, (5) the standard non-linearities (tanh and\nsigmoid) perform best, (6) bidirectional training combined with CTC performs\nbetter than other methods.","url_abs":"http://arxiv.org/abs/1508.02774v1","url_pdf":"http://arxiv.org/pdf/1508.02774v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"benchmarking-of-lstm-networks","repo_url":"https://github.com/tmbdev/clstm","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"benchmarking","task_name":"Benchmarking"}],"methods":[{"method_slug":"lstm","method_name":"LSTM"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}