Analisis Komparatif Kinerja Llama Murni, Rag Native, dan Rag Fine-Tuning

Authors

  • Nindya Desyana Universitas Dian Nuswantoro
  • Ardytha Luthfiarta Universitas Dian Nuswantoro

DOI:

https://doi.org/10.25077/TEKNOSI.v11i3.2025.263-272

Keywords:

Retrieval-Augmented Generation (RAG), Llama 3, LangChain, AI Hallucination, System-level Tuning, Domain-Specific Chatbot, ROUGE-L, BLEU

Abstract

Penelitian ini mengatasi masalah halusinasi pada Large Language Models (LLM) seperti LLaMA ketika menjawab kueri domain spesifik. Tujuan penelitian adalah membandingkan kinerja tiga arsitektur chatbot: LLaMA murni, Retrieval-Augmented Generation (RAG) berbasis LangChain (RAG Native), dan RAG Fine-Tuning. Implementasi RAG pada penelitian ini menggunakan framework LangChain sebagai sistem penghubung antara model LLaMA dan sumber pengetahuan eksternal (vector database). Framework ini menyediakan pipeline retriever-reader yang memungkinkan integrasi antara model bahasa dan data kontekstual melalui embedding serta pencarian vektor. Metode evaluasi dilakukan secara kuantitatif menggunakan metrik ROUGE-L dan BLEU pada dataset studi kasus. Hasil penelitian menunjukkan peningkatan kinerja yang progresif: arsitektur LLaMA murni (baseline) memperoleh skor ROUGE-L sebesar 46.48, implementasi RAG Native (LangChain) meningkat menjadi 61.42, dan model RAG Fine-Tuning (LangChain Optimized) mencapai kinerja tertinggi dengan skor 83.06. Penelitian ini menyimpulkan bahwa integrasi arsitektur RAG melalui framework LangChain secara signifikan meningkatkan akurasi respons chatbot, dan proses fine-tuning pada konfigurasi RAG merupakan langkah optimasi krusial untuk mencapai performa terbaik pada domain spesifik.

References

A. Khan, Y. Wang, W. Zhang, Y. Tian, and M. T. Özsu, “LLM + Vector Data: Coupling of Large Language Models with Vector Data Management for Enhancing Data Science,” Proceedings - 2025 IEEE 41st International Conference on Data Engineering Workshops, ICDEW 2025, pp. 93–96, 2025, doi: 10.1109/ICDEW67478.2025.00018.

P. Lewis et al., “Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks,” Apr. 2021, [Online]. Available: http://arxiv.org/abs/2005.11401

Y. V Patel, V. Salvia, and I. Kumar, “A Survey on Retrieval-Augmented Generation: From Naive to Adaptive Approaches with Financial Insights.”

A. Kotiyal, J. Praveen Gujjar, M. S. Guru Prasad, R. M. Devadas, V. Hiremani, and P. Tangade, “Chat with PDF using LangChain Model,” in 2nd IEEE International Conference on Advances in Information Technology, ICAIT 2024 - Proceedings, Institute of Electrical and Electronics Engineers Inc., 2024. doi: 10.1109/ICAIT61638.2024.10690817.

K. R. Ong and W. P. Wong, “Optimizing Information Retrieval in RAG through Intelligent Reranking and Follow-Up Query Predictions,” Sep. 03, 2025. doi: 10.21203/rs.3.rs-7506627/v1.

S. Mathur and A. Chhabra, “Vector Search Algorithms: A Brief Survey,” Proceedings of the 4th International Conference on Ubiquitous Computing and Intelligent Information Systems, ICUIS 2024, pp. 365–371, 2024, doi: 10.1109/ICUIS64676.2024.10866377.

C.-Y. Lin, “ROUGE: A Package for Automatic Evaluation of Summaries,” 2004. Accessed: Oct. 25, 2025. [Online]. Available: https://aclanthology.org/W04-1013/

H. Rachmat, H. Riza, and T. F. Abidin, “Fine-Tuning Large Language Model (LLM) to Answer Basic Questions for Prospective New Students at Syiah Kuala University Using the Retrieval-Augmented Generation (RAG) Method,” 2024 9th International Conference on Informatics and Computing, ICIC 2024, 2024, doi: 10.1109/ICIC64337.2024.10956296.

K. Papineni, S. Roukos, T. Ward, and W.-J. Zhu, “BLEU: a Method for Automatic Evaluation of Machine Translation."

J. Gohil, H. L. Shifare, and M. Shukla, “Developing a User-Friendly Conversational AI Assistant for University Using Ollama and LLama3,” 2025 International Conference on Data Science, Agents and Artificial Intelligence, ICDSAAI 2025, 2025, doi: 10.1109/ICDSAAI65575.2025.11011878.

AI@Meta, “Llama 3 Model Card,” 2024. [Online]. Available: https://github.com/meta-llama/llama3/blob/main/MODEL_CARD.md

A. Sonkar, S. P. Singh, K. Sahu, A. Sahu, and S. Mishra, “Dynamic Query Handling with RAG Fusion for PDF-Based Knowledge Retrieval Systems,” 2025 4th OPJU International Technology Conference on Smart Computing for Innovation and Advancement in Industry 5.0, OTCON 2025, 2025, doi: 10.1109/OTCON65728.2025.11070378.

X. Xie, H. Liu, W. Hou, and H. Huang, “A Brief Survey of Vector Databases,” 2023 9th International Conference on Big Data and Information Analytics, BigDIA 2023 - Proceedings, pp. 364–371, 2023, doi: 10.1109/BIGDIA60676.2023.10429609.

S. Vakayil, D. Sujitha Juliet, J. Anitha, and S. Vakayil, “RAG-Based LLM Chatbot Using Llama-2,” ICDCS 2024 - 2024 7th International Conference on Devices, Circuits and Systems, pp. 195–199, 2024, doi: 10.1109/ICDCS59278.2024.10561020.

M. R. Putri, A. Y. Husodo, and B. Irmawati, “Simplification of Embedding Process in Retrieval Augmented Generation for Optimizing Question Answering Chatbot Model,” COMNETSAT 2024 - IEEE International Conference on Communication, Networks and Satellite, pp. 665–670, 2024, doi: 10.1109/COMNETSAT63286.2024.10862926.

U. Hasanah and B. P. Hartato, “Assessing Short Answers in Indonesian Using Semantic Text Similarity Method and Dynamic Corpus,” ICITEE 2020 - Proceedings of the 12th International Conference on Information Technology and Electrical Engineering, pp. 312–316, Oct. 2020, doi: 10.1109/ICITEE49829.2020.9271696.

N. Mardiana, R. D. Dana, Faisal, I. Farida, A. G. Azwar, and Nurwathi, “Similarity Measures Implementation on Face Authentication using Indonesian Citizen ID Card,” Proceeding of 2023 17th International Conference on Telecommunication Systems, Services, and Applications, TSSA 2023, 2023, doi: 10.1109/TSSA59948.2023.10366880.

G. Putro Dirgantoro, M. A. Soeleman, and C. Supriyanto, “Smoothing Weight Distance to Solve Euclidean Distance Measurement Problems in K-Nearest Neighbor Algorithm,” Proceedings - 2021 IEEE 5th International Conference on Information Technology, Information Systems and Electrical Engineering: Applying Data Science and Artificial Intelligence Technologies for Global Challenges During Pandemic Era, ICITISEE 2021, pp. 294–298, 2021, doi: 10.1109/ICITISEE53823.2021.9655820.

D. S. Sisodia, I. Nehapriyanka, and P. Amulya, “Longest common subsequence based multistage collaborative filtering for recommender systems,” Proceedings - 2020 21st International Arab Conference on Information Technology, ACIT 2020, Nov. 2020, doi: 10.1109/ACIT50332.2020.9300068.

P. Malik and A. S. Baghel, “An improvement in BLEU metric for English-Hindi machine translation evaluation,” Proceeding - IEEE International Conference on Computing, Communication and Automation, ICCCA 2016, pp. 331–336, Jan. 2017, doi: 10.1109/CCAA.2016.7813740.

J. Zhang, “Improving Genetic Algorithm-Based Automatic Machine Translation Models with Gated Recurrent Units and Bilingual Evaluation Understudy,” pp. 12–16, Sep. 2025, doi: 10.1109/ICICR65456.2025.00010.

Submitted

2025-10-29

Accepted

2025-12-18

Published

2025-12-28

How to Cite

[1]
N. Desyana and A. Luthfiarta, “Analisis Komparatif Kinerja Llama Murni, Rag Native, dan Rag Fine-Tuning”, TEKNOSI, vol. 11, no. 3, pp. 263–272, Dec. 2025.

Similar Articles

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 > >> 

You may also start an advanced similarity search for this article.