hesamaria/rag-chatbot
🧠 Build a RAG-driven chatbot to answer user questions using real-world knowledge from documents, focused on mental health and social policy in Indonesia.
What's novel
🧠 Build a RAG-driven chatbot to answer user questions using real-world knowledge from documents, focused on mental health and social policy in Indonesia.
Code Analysis
9 files read · 3 roundsA FastAPI RAG chatbot that retrieves Indonesian mental health documents from a FAISS index and generates answers using TinyLlama-1.1B with a LoRA adapter fine-tuned on 28 QA pairs.
Strengths
The code is straightforward and readable, with a clean separation between indexing, pipeline, and API layers. The domain focus on Indonesian mental health policy (pasung, BPJS, Puskesmas) is a legitimate and underrepresented use case.
Weaknesses
The README is entirely fabricated (describes a desktop app with installers that don't exist), the code contains critical bugs (invalid dtype parameter, pointless training loop), the training data is only 28 examples making the LoRA fine-tuning meaningless, and there are zero tests. The project also has leftover Ollama references and dependencies that are never used.
Score Breakdown
Signal breakdown
Innovation
Craft
Traction
Scope
Evidence
Commits
10
Contributors
2
Files
22
Active weeks
4
Repository
Language
Python
Stars
2
Forks
1
License
—