{"id":"W4312108337","doi":"10.48550/arxiv.2212.11146","title":"The Challenges of HTR Model Training: Feedback from the Project Donner le gout de l'archive a l'ere numerique","year":2022,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Handwritten Text Recognition Techniques","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université du Québec à Rimouski; Université de Montréal","funders":"","keywords":"Computer science; Handwriting; Process (computing); Handwriting recognition; Base (topology); Scale (ratio); Language model; Character (mathematics); Natural language processing; Word error rate; Data science; Artificial intelligence; Speech recognition; Feature extraction; Programming language; Cartography","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":[0.04575191,0.001770953,0.001250412,0.001385372,0.002094703,0.003867838,0.002787129,0.003457849,0.004515812],"category_scores_gemma":[0.1880143,0.0005534479,0.0007218867,0.001419104,0.002422317,0.00532645,0.003700557,0.004471529,0.003551896],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002776372,"about_ca_system_score_gemma":0.0035222,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0172554,"about_ca_topic_score_gemma":0.02186824,"domain_scores_codex":[0.9670364,0.01751745,0.001275576,0.002954592,0.0104262,0.0007896827],"domain_scores_gemma":[0.7925724,0.1459845,0.002409607,0.009190097,0.04561345,0.004230032],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"qualitative","study_design_scores_codex":[0.001304412,0.001955799,0.02182011,0.001941303,0.0001673709,0.001612131,0.03319894,0.04784796,0.01818323,0.007702331,0.1810946,0.6831718],"study_design_scores_gemma":[0.0003030933,0.00188783,0.02788606,0.001513611,0.0001433547,0.002467952,0.03319815,0.21747,0.05220442,0.01478622,0.6476831,0.0004561428],"study_design_candidate":"qualitative","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5334485,0.01092238,0.3118071,0.06338266,0.003363396,0.001098573,0.006341071,0.01106072,0.05857563],"genre_scores_gemma":[0.7286136,0.004766958,0.2096039,0.004730597,0.0007829408,0.000575111,0.012044,0.004307963,0.03457481],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.04575191,"threshold_uncertainty_score":0.2419621,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1201216217698076,"score_gpt":0.2170851636345285,"score_spread":0.09696354186472089,"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."}}