MedMCQ — Subject Classifier (Qwen3-0.6B)

A small fine-tuned Qwen3 model that takes a medical multiple-choice question (MCQ) and predicts which medical subject the question belongs to.

This is the router model in the MedMCQ three-hop pipeline. It is the first model called when a new MCQ enters the system: its output decides which downstream subject-specific classifier and generator are used.

The MedMCQ pipeline

The MedMCQ project explores small, specialized models for medical MCQs. Instead of using one large model for everything, it splits the task across three hops:

Subject routing — this model. Given an MCQ, pick the medical subject (Anatomy, Pharmacology, etc.).
Topic classification — a per-subject classifier picks the topic within that subject.
Answer generation — a per-subject generator produces the answer / new MCQs in that subject.

Each hop is a separate, narrow model. They are all published under the MedMCQ Medical Models collection.

Quick start

from transformers import AutoTokenizer, AutoModelForCausalLM

repo = "stravoris/medmcq-subject-classifier-qwen3-0.6b"
tokenizer = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(repo)

prompt = """Classify the following medical MCQ by subject.

Question: Which artery supplies the head of the femur in adults?
A) Obturator artery
B) Medial circumflex femoral artery
C) Lateral circumflex femoral artery
D) Superior gluteal artery

Subject:"""

inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=8, do_sample=False)
print(tokenizer.decode(outputs[0][inputs.input_ids.shape[1]:], skip_special_tokens=True))

Prompt format

The model expects MCQs in this exact form:

Classify the following medical MCQ by subject.

Question: <question text>
A) <option A>
B) <option B>
C) <option C>
D) <option D>

Subject:

The model completes the prompt with one of the 15 subject labels:

Anaesthesia, Anatomy, Biochemistry, Cell Biology & Histology, ENT, Genetics, Microbiology, Obstetrics & Gynaecology, Ophthalmology, Orthopaedics, Pathology, Pharmacology, Physiology, Psychiatry, Radiology.

What this model is not

This is a sample model for demonstration. It is not a production-grade medical AI system:

It has not been formally evaluated against board-level benchmarks.
It should not be used to make clinical decisions or provide medical advice.
It was trained on a curated educational dataset and inherits any biases or gaps in that data.

The MedMCQ project exists to explore small-model pipeline architectures for medical reasoning, not to ship a medical product.