{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,6,28]],"date-time":"2025-06-28T06:24:06Z","timestamp":1751091846319,"version":"3.28.0"},"reference-count":17,"publisher":"IEEE","license":[{"start":{"date-parts":[[2022,4,6]],"date-time":"2022-04-06T00:00:00Z","timestamp":1649203200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2022,4,6]],"date-time":"2022-04-06T00:00:00Z","timestamp":1649203200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2022,4,6]]},"DOI":"10.1109\/isqed54688.2022.9806238","type":"proceedings-article","created":{"date-parts":[[2022,6,29]],"date-time":"2022-06-29T19:46:20Z","timestamp":1656531980000},"page":"1-6","source":"Crossref","is-referenced-by-count":3,"title":["Low-Precision Quantization Techniques for Hardware-Implementation-Friendly BERT Models"],"prefix":"10.1109","author":[{"given":"Xinpei","family":"Zhang","sequence":"first","affiliation":[{"name":"The University of Tokyo"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yi","family":"Ding","sequence":"additional","affiliation":[{"name":"The University of Tokyo"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Mingfei","family":"Yu","sequence":"additional","affiliation":[{"name":"The University of Tokyo"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Shin-Ichi","family":"O'Uchi","sequence":"additional","affiliation":[{"name":"AIST-UTokyo AI Chip Design Open Innovation"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Masahiro","family":"Fujita","sequence":"additional","affiliation":[{"name":"The University of Tokyo"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"article-title":"Learned step size quantization","year":"2019","author":"esser","key":"ref10"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.3390\/computers5040020"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA45697.2020.00086"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2016.7471821"},{"key":"ref14","first-page":"7873","article-title":"CLIP-Q : Deep Network Compression Learning by In-Parallel","author":"tung","year":"2018","journal-title":"Proceedings - 31th IEEE Conference on Computer Vision and Pattern Recognition CVPR 2018"},{"key":"ref15","first-page":"7543","article-title":"Improving neural network quantization without retraining using outlier channel splitting","author":"zhao","year":"2019","journal-title":"International Conference on Machine Learning"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/W18-5446"},{"article-title":"Pushing the limits of narrow precision inferencing at cloud scale with microsoft floating point","year":"2020","author":"rouhani","key":"ref17"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.emnlp-main.37"},{"article-title":"Kdlsq-bert: A quantized bert combining knowledge distillation with learned step size quantization","year":"2021","author":"jin","key":"ref3"},{"key":"ref6","first-page":"10271","article-title":"Pushing the Limits of Narrow Precision Inferencing at Cloud Scale with Microsoft Floating Point","volume":"33","author":"rouhani","year":"2020","journal-title":"Advances in neural information processing systems"},{"article-title":"Aciq: Analytical clipping for integer quantization of neural networks","year":"2018","author":"banner","key":"ref5"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v34i05.6409"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00286"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.1109\/EMC2-NIPS53020.2019.00016"},{"article-title":"Bert: Pre-training of deep bidirectional transformers for language understanding","year":"2018","author":"devlin","key":"ref1"},{"article-title":"I-bert: Integer-only bert quantization","year":"2021","author":"kim","key":"ref9"}],"event":{"name":"2022 23rd International Symposium on Quality Electronic Design (ISQED)","start":{"date-parts":[[2022,4,6]]},"location":"Santa Clara, CA, USA","end":{"date-parts":[[2022,4,7]]}},"container-title":["2022 23rd International Symposium on Quality Electronic Design (ISQED)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/9806045\/9806137\/09806238.pdf?arnumber=9806238","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,7,25]],"date-time":"2022-07-25T20:14:02Z","timestamp":1658780042000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/9806238\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,4,6]]},"references-count":17,"URL":"https:\/\/doi.org\/10.1109\/isqed54688.2022.9806238","relation":{},"subject":[],"published":{"date-parts":[[2022,4,6]]}}}