{"id":"abr-ai-niagara-38m-batch-en","nom":"niagara-38m-batch.en","editeur":"Applied Brain Research","pays":"Canada","poids_ouverts":true,"poids_selon":null,"licence":"Applied Brain Research Open License","licence_selon":null,"liens":{"page":"https://huggingface.co/abr-ai/niagara-38m-batch.en","essayer":"https://huggingface.co/spaces/abr-ai/niagara-asr-demo"},"fiche_auto":true,"page":"https://quelleia.com/voix/modeles/abr-ai-niagara-38m-batch-en/","releve":"2026-10-19","en_bref":"49e sur 53 en transcription en anglais (Open ASR).","presentation":"niagara-38m-batch.en est un modèle compact de reconnaissance vocale anglaise d'Applied Brain Research, de type modèle à espace d'états avec attention, entraîné sur environ 50 000 heures de parole. Cette variante non causale sert à la comparaison sur le classement Open ASR, une version en streaming étant vendue sous licence commerciale.","sortie":"2026-04-15","taille":"38 millions de paramètres","langues":{"annoncees":{"annoncees":"anglais seulement","nombre":1,"francais":false,"liste":["en"]},"prises_en_charge_selon_open_tts":null,"evaluees":["en"]},"preuves":[{"appuie":["taille"],"lu":"https://huggingface.co/abr-ai/niagara-38m-batch.en/raw/main/README.md (« 38M parameters. »), 17/10/2026"},{"appuie":["langues"],"lu":"https://huggingface.co/abr-ai/niagara-38m-batch.en/raw/main/README.md (« Automatic speech recognition model transcribing speech audio to text in English. »), 17/10/2026"},{"appuie":["sortie"],"lu":"https://huggingface.co/abr-ai/niagara-38m-batch.en/raw/main/README.md (« Release Date: April 15, 2026 »), 17/10/2026"},{"appuie":["presentation"],"lu":"https://huggingface.co/abr-ai/niagara-38m-batch.en/raw/main/README.md (« A live streaming variant of this english ASR model with sub-120ms latency and features such as punctuation, capitalization and custom vocabulary »), 17/10/2026"}],"classements":[{"id":"transcription/open-asr-anglais","nom":"Transcription en anglais, Open ASR","rang":49,"sur":53,"valeur":11.14,"unite":"%","chiffres":{"rtfx":4546.42,"jeux":{"AMI-Cleaned":13.87,"Earnings22-Cleaned-AA-chunked":12.94,"Gigaspeech-Cleaned":11.38,"LS Clean":3.81,"LS Other":9.28,"Private (conversational)":19.65,"Private (scripted)":10.05,"SPGISpeech":3.08,"URGENT24":22.88,"Voice Arena Monsoon":10.41,"Voxpopuli-AA-Cleaned":5.16}},"voisins":["nvidia-nemotron-3-5-asr-streaming-0-6b","meta-omniasr-ctc-7b-v2"],"source":{"nom":"Open ASR Leaderboard","mention":"Source : Open ASR Leaderboard, Hugging Face","url":"https://huggingface.co/spaces/hf-audio/open_asr_leaderboard","releve":"2026-10-19"}}],"noms_dans_les_sources":{"hf_audio":"abr-ai/niagara-38m-batch.en"}}