プãƒãƒ³ãƒ—トを高性能ã«æœ€é©åŒ–ã™ã‚‹Prompt Tuningã¨ã¯ï¼Ÿ3ã¤ã®è¦ç‚¹ âœ”ï¸ å‡çµã—ãŸæ±Žç”¨LLMã®Prompt部分ã®ã¿ã‚’å¦ç¿’ã™ã‚‹Prompt Tuningã‚’æ案 âœ”ï¸ Fine-tuningã«è¿‘ã„精度をå©ã出ã—ãŸ âœ”ï¸ å¤§å¹…ãªãƒ‘ラメータ削減をå¯èƒ½ã«ã™ã‚‹ The Power of Scale for Parameter-Efficient Prompt Tuning written by Brian Lester, Rami Al-Rfou, Noah Constant (Submitted on 18 Apr 2021 (v1), last revised 2 Sep 2021 (this version, v2)) Comments: Accepted to EMNLP 2021 Subjects: Computation and Language (cs.CL) code： 本記事ã§ä½¿ç”¨ã—ã¦ã„ã‚‹ç”»åƒã¯è«–æ–‡ä¸ã®ã‚‚ã®ã€ç´¹ä»‹ã‚¹ãƒ©ã‚¤ãƒ‰ã®ã‚‚ã®ã€ã¾ãŸã¯ãã‚Œ
{{#tags}}- {{label}}
{{/tags}}