{"id":"W4378801057","doi":"10.1145/3583781.3590302","title":"High-Throughput Edge Inference for BERT Models via Neural Architecture Search and Pipeline","year":2023,"lang":"en","type":"article","venue":"","topic":"Advanced Neural Network Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Computer science; Pipeline (software); Throughput; Inference; Enhanced Data Rates for GSM Evolution; Architecture; Artificial neural network; Computer architecture; Parallel computing; Distributed computing; Artificial intelligence; Operating system; Wireless","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":[0.0007919937,0.001038317,0.0008010654,0.0005558668,0.0005184124,0.0009000688,0.001620528,0.0008864411,0.005209838],"category_scores_gemma":[0.003389455,0.0006672842,0.0006390443,0.0005315993,0.0005976807,0.001907567,0.0009543215,0.001661993,0.0006064594],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001238476,"about_ca_system_score_gemma":0.001963645,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01646046,"about_ca_topic_score_gemma":0.04073166,"domain_scores_codex":[0.9997017,0.00008051434,0.00001363799,0.00006213038,0.00008374386,0.00005829661],"domain_scores_gemma":[0.9988878,0.0007449569,0.0000701264,0.0001213121,0.0001295503,0.00004619539],"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":[0.0001338514,0.00004730373,0.001183682,0.00005944045,0.00003506288,0.00007340994,0.00003129392,0.9423382,0.002310864,0.009265728,0.001457962,0.04306322],"study_design_scores_gemma":[0.000003481323,0.000006210119,0.00002963454,0.000001113621,0.000002275968,0.000003465529,0.000002823324,0.9967475,0.0002799086,0.002813488,0.0001085773,0.000001484624],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05102098,0.0003444004,0.9403782,0.0003325859,0.00003896924,0.00004815607,0.0001703212,0.003099522,0.004566852],"genre_scores_gemma":[0.7606335,0.0001936061,0.2337471,0.00019713,0.00003409447,0.00008976411,0.0003961681,0.0002713302,0.004437222],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01646046,"threshold_uncertainty_score":0.03272927,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04471021459728105,"score_gpt":0.3134345893846869,"score_spread":0.2687243747874059,"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."}}