{"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/fast-conformer-with-linearly-scalable","title":"Fast Conformer with Linearly Scalable Attention for Efficient Speech Recognition","arxiv_id":"2305.05084","date":"2023-05-08","proceeding":null,"authors":["Dima Rekesh","Nithin Rao Koluguri","Samuel Kriman","Somshubra Majumdar","Vahid Noroozi","He Huang","Oleksii Hrinchuk","Krishna Puvvada","Ankur Kumar","Jagadeesh Balam","Boris Ginsburg"],"abstract":"Conformer-based models have become the dominant end-to-end architecture for speech processing tasks. With the objective of enhancing the conformer architecture for efficient training and inference, we carefully redesigned Conformer with a novel downsampling schema. The proposed model, named Fast Conformer(FC), is 2.8x faster than the original Conformer, supports scaling to Billion parameters without any changes to the core architecture and also achieves state-of-the-art accuracy on Automatic Speech Recognition benchmarks. To enable transcription of long-form speech up to 11 hours, we replaced global attention with limited context attention post-training, while also improving accuracy through fine-tuning with the addition of a global token. Fast Conformer, when combined with a Transformer decoder also outperforms the original Conformer in accuracy and in speed for Speech Translation and Spoken Language Understanding.","url_abs":"https://arxiv.org/abs/2305.05084v6","url_pdf":"https://arxiv.org/pdf/2305.05084v6.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":[],"tasks":[{"task_slug":"automatic-speech-recognition-2","task_name":"Automatic Speech Recognition"},{"task_slug":"decoder","task_name":"Decoder"},{"task_slug":"speech-recognition","task_name":"Speech Recognition"},{"task_slug":"spoken-language-understanding","task_name":"Spoken Language Understanding"},{"task_slug":"translation","task_name":"Translation"},{"task_slug":"speech-recognition-1","task_name":"speech-recognition"}],"methods":[{"method_slug":"absolute-position-encodings","method_name":"Absolute Position Encodings"},{"method_slug":"adam","method_name":"Adam"},{"method_slug":"attention","method_name":"Attention"},{"method_slug":"bpe","method_name":"BPE"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"label-smoothing","method_name":"Label Smoothing"},{"method_slug":"layer-normalization","method_name":"Layer Normalization"},{"method_slug":"linear-layer","method_name":"Linear Layer"},{"method_slug":"multi-head-attention","method_name":"Multi-Head Attention"},{"method_slug":"position-wise-feed-forward-layer","method_name":"Position-Wise Feed-Forward Layer"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"speed","method_name":"SPEED"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"transformer","method_name":"Transformer"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/speech-recognition-on-common-voice-english","task":"Speech Recognition","dataset":"Common Voice English","model":"parakeet-rnnt-1.1b","rank_in_archive_order":1,"of":2,"metrics":{"Word Error Rate (WER)":"5.8%"},"uses_additional_data":true},{"leaderboard":"/sota/speech-recognition-on-librispeech-test-clean","task":"Speech Recognition","dataset":"LibriSpeech test-clean","model":"parakeet-rnnt-1.1b","rank_in_archive_order":6,"of":64,"metrics":{"Word Error Rate (WER)":"1.46"},"uses_additional_data":true},{"leaderboard":"/sota/speech-recognition-on-spgispeech","task":"Speech Recognition","dataset":"SPGISpeech","model":"parakeet-rnnt-1.1b","rank_in_archive_order":2,"of":3,"metrics":{"Word Error Rate (WER)":"3.11"},"uses_additional_data":true},{"leaderboard":"/sota/speech-recognition-on-tedlium","task":"Speech Recognition","dataset":"Tedlium","model":"parakeet-rnnt-1.1b","rank_in_archive_order":2,"of":4,"metrics":{"Word Error Rate (WER)":"3.92"},"uses_additional_data":true}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2305.05084","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}