MatrixCoder03/gemma-3-270m-med-lora

Gemma-3-270M Medical Reminder LoRA Adapter

This repository contains a LoRA (Low-Rank Adaptation) adapter fine-tuned on structured medical reminder dialogues.
The adapter extends google/gemma-3-270m to generate concise, structured responses for medication reminders and scheduling tasks.

Model Details

• Base model: google/gemma-3-270m
• Fine-tuning method: LoRA (PEFT)
• Task: Instruction-following / reminder generation
• Language: English

Intended Use

This adapter is intended for:

• Educational demonstrations of PEFT / LoRA fine-tuning
• Research prototypes involving medical reminder dialogue
• Cost- and memory-constrained deployments

Not intended for real clinical decision-making.

How to Use

from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel

base_model = "google/gemma-3-270m"
adapter = "MatrixCoder03/gemma-3-270m-med-lora"

model = AutoModelForCausalLM.from_pretrained(base_model)
model = PeftModel.from_pretrained(model, adapter)

tokenizer = AutoTokenizer.from_pretrained(base_model)

inputs = tokenizer(
    "Remind the patient to take 200mg paracetamol at 5pm.",
    return_tensors="pt"
)

outputs = model.generate(**inputs, max_new_tokens=50)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))