반응형 arxiv572 [2022-11-20] 오늘의 자연어처리 Zero-Shot Dynamic Quantization for Transformer Inference We introduce a novel run-time method for significantly reducing the accuracy loss associated with quantizing BERT-like models to 8-bit integers. Existing methods for quantizing models either modify the training procedure,or they require an additional calibration step to adjust parameters that also requires a selected held-out dataset. Our .. 2022. 11. 20. [2022-11-19] 오늘의 자연어처리 Style Classification of Rabbinic Literature for Detection of Lost Midrash Tanhuma Material Midrash collections are complex rabbinic works that consist of text in multiple languages, which evolved through long processes of unstable oral and written transmission. Determining the origin of a given passage in such a compilation is not always straightforward and is often a matter of dispute among sch.. 2022. 11. 19. [2022-11-18] 오늘의 자연어처리 Noisy Pairing and Partial Supervision for Opinion Summarization Current opinion summarization systems simply generate summaries reflecting important opinions from customer reviews, but the generated summaries may not attract the reader's attention. Although it is helpful to automatically generate professional reviewer-like summaries from customer reviews, collecting many training pairs of custom.. 2022. 11. 18. [2022-11-17] 오늘의 자연어처리 FolkScope: Intention Knowledge Graph Construction for Discovering E-commerce Commonsense As stated by Oren Etzioni, ``commonsense is the dark matter of artificial intelligence''. In e-commerce, understanding users' needs or intentions requires substantial commonsense knowledge, e.g., ``A user bought an iPhone and a compatible case because the user wanted the phone to be protected''. In this pape.. 2022. 11. 17. 이전 1 ··· 107 108 109 110 111 112 113 ··· 143 다음 반응형