Researchers present ChipNeMo, using domain adaptation to enhance LLMs for chip design, achieving up to 5x model size reduction with better performance.
Authors: Mingjie Liu, NVIDIA {Equal contribution}; Teodor-Dumitru Ene, NVIDIA {Equal contribution}; Robert Kirby, NVIDIA {Equal contribution}; Chris Cheng, NVIDIA {Equal contribution}; Nathaniel Pinckney, NVIDIA {Equal contribution}; Rongjian Liang, NVIDIA {Equal contribution}; Jonah Alben, NVIDIA; Himyanshu Anand, NVIDIA; Sanmitra Banerjee, NVIDIA; Ismet Bayraktaroglu, NVIDIA; Bonita Bhaskaran, NVIDIA; Bryan Catanzaro, NVIDIA; Arjun Chaudhuri, NVIDIA; Sharon Clay, NVIDIA; Bill Dally, NVIDIA;...
Mingjie Liu conducted DAPT and SFT model training. Teodor-Dumitru Ene, Robert Kirby developed inference and application evaluation infrastructure. Chris Cheng developed RAG framework. Nathaniel Pinckney collected and prepared data sets for training. Rongjian Liang developed custom tokenizers. Walker Turner, Charley Lind, George Kokai developed a general circuit design knowledge benchmark.
New York, NY, USA: Association for Computing Machinery, 2020, p. 27–32. . Available: https://doi.org/10.1145/3380446.3430634 “Beautiful Soup,” https://www.crummy.com/software/BeautifulSoup/, accessed: 10 Oct 2023. K. Sakaguchi et al., “Winogrande: An adversarial winograd schema challenge at scale,” arXiv preprint arXiv:1907.10641, 2019. R. Zellers et al.
Mingjie Liu conducted DAPT and SFT model training. Mingjie Liu Teodor-Dumitru Ene, Robert Kirby developed inference and application evaluation infrastructure. Teodor-Dumitru Ene, Robert Kirby Chris Cheng developed RAG framework. Chris Cheng Nathaniel Pinckney collected and prepared data sets for training. Nathaniel Pinckney Rongjian Liang developed custom tokenizers. Rongjian Liang Walker Turner, Charley Lind, George Kokai developed a general circuit design knowledge benchmark.
United Kingdom Latest News, United Kingdom Headlines
Similar News:You can also read news stories similar to this one that we have collected from other news sources.
ChipNeMo: Domain-Adapted LLMs for Chip Design: ChipNemo Domain Adaptation MethodsResearchers present ChipNeMo, using domain adaptation to enhance LLMs for chip design, achieving up to 5x model size reduction with better performance.
Read more »
ChipNeMo: Domain-Adapted LLMs for Chip Design: LLM ApplicationsResearchers present ChipNeMo, using domain adaptation to enhance LLMs for chip design, achieving up to 5x model size reduction with better performance.
Read more »
ChipNeMo: Domain-Adapted LLMs for Chip Design: Abstract and IntroResearchers present ChipNeMo, using domain adaptation to enhance LLMs for chip design, achieving up to 5x model size reduction with better performance.
Read more »
ChipNeMo: Domain-Adapted LLMs for Chip Design: AppendixResearchers present ChipNeMo, using domain adaptation to enhance LLMs for chip design, achieving up to 5x model size reduction with better performance.
Read more »
ChipNeMo: Domain-Adapted LLMs for Chip Design: Related WorksResearchers present ChipNeMo, using domain adaptation to enhance LLMs for chip design, achieving up to 5x model size reduction with better performance.
Read more »
ChipNeMo: Domain-Adapted LLMs for Chip Design: EvaluationsResearchers present ChipNeMo, using domain adaptation to enhance LLMs for chip design, achieving up to 5x model size reduction with better performance.
Read more »