Answers are grounded in this paper and cite their sources.
Pleco Studio 1.0 is a local-first, open-source desktop application for building, fine-tuning, evaluating and deploying custom language models on an individual workstation. This white paper documents its design and implementation status as at 5 November 2026. It covers the architecture, which keeps the desktop application separate from the machine-learning runtime. It also covers project-scoped dataset and model management with integrity checks, LoRA fine-tuning with Transformers and PEFT, evaluation and experiment tracking, a local authenticated OpenAI-compatible API, and optional Open WebUI and Ollama integrations. The paper reports the repository’s test results and states its limitations plainly: QLoRA and real Ollama import are unvalidated, evaluation metrics are lexical, and hardware validation is limited. It does not claim production readiness or model quality. It has not been peer-reviewed.
No takes yet. Share an insight, caveat, or question.
Pushpitha Sandaruwan Weerasooriya Weerasooriya Mudiyanselage (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: