Notes on LoRA Fine-Tuning and Local LLM Inference
Jul 8, 2025 • 8 min read
What I learned running open-source LLMs locally with Ollama and llama.cpp, with LoRA adapters on top.
Category
What I learned running open-source LLMs locally with Ollama and llama.cpp, with LoRA adapters on top.
Sharing the encoder across Paraphrase, STS and Sentiment — what helped and what hurt.