{"id":"W3022810465","doi":"10.1109/micro50266.2020.00071","title":"GOBO: Quantizing Attention-Based NLP Models for Low Latency and Energy Efficient Inference","year":2020,"lang":"en","type":"preprint","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":11,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto","funders":"","keywords":"Computer science; Quantization (signal processing); Computation; Inference; Latency (audio); Language model; Computer engineering; Parallel computing; Computer hardware; Algorithm; Artificial intelligence","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":[0.0007428518,0.0007817984,0.0006990916,0.0006346194,0.0003695713,0.000919938,0.001777631,0.0008997652,0.005010807],"category_scores_gemma":[0.003708655,0.0004420132,0.0005984882,0.000655612,0.0006695917,0.002384693,0.001603406,0.001557621,0.00111468],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001033129,"about_ca_system_score_gemma":0.00102652,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01042909,"about_ca_topic_score_gemma":0.01985732,"domain_scores_codex":[0.9996945,0.00006678696,0.00001605041,0.00007875374,0.0001061877,0.00003773596],"domain_scores_gemma":[0.9993625,0.0003581657,0.00005074385,0.0001058076,0.00009177445,0.00003105695],"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":[0.0004986929,0.0001663097,0.001205754,0.0003851979,0.0001399688,0.0001546491,0.0002859771,0.5403324,0.02986532,0.03277383,0.01250503,0.3816868],"study_design_scores_gemma":[0.0000109008,0.00001317794,0.00006962147,0.000007028066,0.000006413154,0.000009924203,0.000009573185,0.9875259,0.002258815,0.009297215,0.0007865493,0.000004875122],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01782248,0.0004641166,0.9738513,0.0003568479,0.00005980517,0.0000728352,0.0003433799,0.005080035,0.001949204],"genre_scores_gemma":[0.5428099,0.0005069075,0.4493875,0.00055167,0.000090602,0.0002816205,0.001245882,0.0008161198,0.004309665],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01042909,"threshold_uncertainty_score":0.02073675,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06401963398422118,"score_gpt":0.2784773016417433,"score_spread":0.2144576676575221,"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."}}