{"id":"W7106839461","doi":"10.5281/zenodo.17719731","title":"Efficient LLM Self-Hosting using Adapters and VLLM Deployment","year":2025,"lang":"en","type":"article","venue":"Zenodo (CERN European Organization for Nuclear Research)","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Carleton University","funders":"","keywords":"Software deployment; Modular design; Orchestration; Abstraction; Dependency (UML); Personalization; Architecture; Inference","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":[0.002705206,0.0005370373,0.0004633823,0.0009233046,0.0004726534,0.00208738,0.002047878,0.0008158509,0.002881639],"category_scores_gemma":[0.008330837,0.0004938942,0.0008472069,0.0007166403,0.0007117199,0.003762331,0.00343831,0.00137878,0.001515935],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008883051,"about_ca_system_score_gemma":0.0009286799,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001754897,"about_ca_topic_score_gemma":0.001769014,"domain_scores_codex":[0.9981216,0.0006416359,0.0001547564,0.0003584552,0.0005183099,0.000205231],"domain_scores_gemma":[0.9968373,0.0007751301,0.0001835913,0.001777287,0.0003283856,0.00009836445],"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":[0.0004337263,0.0003366621,0.01077351,0.0007653895,0.0002017447,0.001056366,0.002331773,0.05884364,0.07774653,0.09398905,0.01705075,0.7364708],"study_design_scores_gemma":[0.00008880967,0.000260405,0.005811982,0.0002245438,0.0001663873,0.001000497,0.0008630239,0.6654749,0.1155127,0.08153784,0.1288985,0.0001603835],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04849971,0.0005025724,0.9054534,0.0004965751,0.0001111674,0.0002138115,0.0002190137,0.03582201,0.008681697],"genre_scores_gemma":[0.5004892,0.0005211167,0.4872372,0.0004092762,0.0000623346,0.0002912277,0.001226678,0.003523471,0.00623947],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002881639,"threshold_uncertainty_score":0.01430666,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02276446153855251,"score_gpt":0.2615022000031927,"score_spread":0.2387377384646402,"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."}}