{"id":"W4389520158","doi":"10.18653/v1/2023.findings-emnlp.404","title":"NASH: A Simple Unified Framework of Structured Pruning for Accelerating Encoder-Decoder Language Models","year":2023,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Kootenay Association for Science & Technology","funders":"Institute for Information and Communications Technology Promotion; Ministry of Science and ICT, South Korea; Korea Advanced Institute of Science and Technology","keywords":"Computer science; Encoder; Speedup; Inference; Pruning; Language model; Algorithm; Artificial intelligence; Parallel computing","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":[0.002375242,0.001198989,0.00101643,0.001190193,0.0007442328,0.001503901,0.002897811,0.001382153,0.003999636],"category_scores_gemma":[0.01046352,0.0007206454,0.001062843,0.0009177654,0.001029797,0.003573987,0.002530857,0.002376378,0.001503562],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001139144,"about_ca_system_score_gemma":0.002389201,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005222887,"about_ca_topic_score_gemma":0.01402378,"domain_scores_codex":[0.9987416,0.0004752915,0.00008797705,0.0002378831,0.0003546026,0.0001026898],"domain_scores_gemma":[0.997304,0.001586898,0.0001356006,0.0004961757,0.0003728666,0.0001044962],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0003229695,0.0001640548,0.001560944,0.0003316643,0.0001634333,0.0004156661,0.0004857667,0.445529,0.01571351,0.160505,0.008290324,0.3665176],"study_design_scores_gemma":[0.00002172581,0.00003331938,0.00007410749,0.00001486701,0.00002999607,0.00005850517,0.00001964989,0.9558577,0.003264573,0.03840536,0.002209005,0.00001127451],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.005376516,0.0001996622,0.9908054,0.0001394076,0.00004138089,0.00006479172,0.00009320509,0.001914058,0.001365588],"genre_scores_gemma":[0.2054235,0.0003902711,0.7886105,0.0002688452,0.0000974811,0.0003034015,0.0005421262,0.000759779,0.003604212],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005222887,"threshold_uncertainty_score":0.01338017,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06707218725630042,"score_gpt":0.3174647955264347,"score_spread":0.2503926082701343,"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."}}