Particularly, financial named‑entity recognition (NER) is one of the many important approaches to translate unformatted reports and news into structured knowledge graphs. However, free, easy‑to‑use large language models (LLMs) often fail to differentiate organisations as people, or disregard an actual monetary amount entirely. This paper takes Meta's Llama 3 8B and applies it to financial NER by making a combination of instruction fine‑tuning and Low‑Rank Adaptation (LoRA). Each of our annotated sentences is converted into an instruction–input–output triple with the model learning to understand task descriptions and finetune with small low‑rank matrices instead of all weights. Using a corpus of 1 693 sentences our method can obtain a micro‑F1 of 0.894 in comparison to Qwen3-8B, Baichuan2-7B, T5 and BERT‑Base. We present dataset statistics, describe training hyper‑parameters and perform visualization on entity density, learning curve and evaluation metrics. Our results show that instruction tuning along with parameter‑efficient fine‑tuning is conducive to state‑of‑the‑art performance on domain‑sensitive NER.