Uhandisi wa Maagizo — Ubunifu wa Maelekezo kwa Miundo Mikuu ya Lugha
Uhandisi wa maagizo ni mazoezi ya kuunda maelekezo rasmi ya lugha asilia — maagizo — ili kupata matokeo yaliyolengwa kutoka kwa miundo mikuu ya lugha (LLMs). Uliwekwa rasmi na Brown et al. (2020) katika muktadha wa GPT-3 na kupanuliwa na Wei et al. (2022) kwa kutumia uhandisi wa mfuatano wa mawazo, unajumuisha mikakati minne mikuu: sifuri-risasi, wachache-risasi, mfuatano-wa-mawazo, na mti-wa-mawazo. Badala ya kutoa mafunzo upya kwa modeli, mchambuzi huunda tabia ya modeli kupitia muundo wa maandishi ya pembejeo.
Soma mbinu kamili
Ingia kwa akaunti ya bure ili kusoma sehemu hii.
Method map
The neighbourhood of related methods — select a node to explore.
Vyanzo
- Brown, T. et al. (2020). Language Models are Few-Shot Learners. Advances in Neural Information Processing Systems (NeurIPS), 33, 1877-1901. link ↗
- Wei, J. et al. (2022). Chain-of-Thought Prompting Elicits Reasoning in Large Language Models. Advances in Neural Information Processing Systems (NeurIPS), 35. link ↗
Jinsi ya kunukuu ukurasa huu
ScholarGate. (2026, June 1). Prompt Engineering (Instruction Design for Large Language Models). ScholarGate. https://scholargate.app/sw/text-mining/prompt-engineering
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Uainishaji wa maandishi kwa michache ya mifano (Few-Shot Text Classification)Uchimbaji wa Matini↔ compare
- Utekelezaji wa GPT (GPT Fine-Tuning)Ujifunzaji wa Kina↔ compare
- LoRA na PEFTUjifunzaji wa Kina↔ compare
- Uzalishaji wa Lugha AsiliaUchimbaji wa Matini↔ compare
- Uzalishaji Ulioimarishwa kwa Urejeshaji (RAG)Uchimbaji wa Matini↔ compare
- Uainishaji wa MaandishiUchimbaji wa Matini↔ compare
- Uainishaji wa Hali-ya-SifuriUchimbaji wa Matini↔ compare
Umeona tatizo kwenye ukurasa huu? Ripoti au pendekeza marekebisho →