{"id":"W4287752972","doi":"10.48550/arxiv.2006.13343","title":"One Model to Pronounce Them All: Multilingual Grapheme-to-Phoneme\\n Conversion With a Transformer Ensemble","year":2020,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"","keywords":"Grapheme; Transformer; Computer science; Word error rate; Task (project management); Speech recognition; Language model; Artificial intelligence; Natural language processing; Speech synthesis; Engineering; Voltage","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0007969144,0.002037168,0.0008669111,0.0007056326,0.0006019868,0.001285584,0.001707995,0.001176851,0.005891907],"category_scores_gemma":[0.001603827,0.0005087623,0.001263512,0.0006141122,0.000536507,0.002823228,0.001949884,0.003540358,0.005807059],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004840474,"about_ca_system_score_gemma":0.001023998,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005885868,"about_ca_topic_score_gemma":0.01311915,"domain_scores_codex":[0.9995377,0.0001034919,0.00001903064,0.0002191196,0.00006630303,0.00005438508],"domain_scores_gemma":[0.9994892,0.0001649402,0.00002066127,0.0001532637,0.0001156876,0.00005633241],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0008411803,0.0004944766,0.003140986,0.0002556045,0.0005447372,0.000639536,0.0004294596,0.1700978,0.04191466,0.01633817,0.04101035,0.7242931],"study_design_scores_gemma":[0.00004040146,0.0001090796,0.000397953,0.00002110747,0.0001323764,0.0002277041,0.0000916124,0.9610005,0.01480733,0.01413346,0.008997791,0.00004073029],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05162027,0.001182574,0.9087976,0.001471052,0.001025618,0.0001128474,0.001591366,0.0202332,0.01396553],"genre_scores_gemma":[0.7166188,0.0008535509,0.2453032,0.001119778,0.0003448097,0.0001804888,0.00564886,0.001959396,0.02797112],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005891907,"threshold_uncertainty_score":0.01971042,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09770618884152414,"score_gpt":0.2212921747534571,"score_spread":0.1235859859119329,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}