medical-llm-mistral

Medical LLM - Mistral 7B (LoRA Fine-tuned)

Fine-tuned Mistral-7B on 5000+ medical QA samples using LoRA (Low-Rank Adaptation) and 4-bit quantization. Built end-to-end pipeline with Streamlit deployment and Medical Agent with conversation memory.

Model Details

• Base Model: Mistral-7B
• Fine-tuning: LoRA (only 0.02% parameters trained)
• Quantization: 4-bit (memory efficient)
• Dataset: 5000+ Medical QA samples
• Training Platform: Kaggle (GPU T4)

Datasets Used

Evaluation Results

How to Use

from transformers import pipeline

pipe = pipeline( "text-generation", model="Anjaligiriraj/medical-llm-mistral" )

result = pipe("What are symptoms of diabetes?", max_new_tokens=200) print(result[0]['generated_text'])

GitHub Repository

https://github.com/Anjli-cdhry/Healthcare-LLM-LoRA

Disclaimer

For educational purposes only. Always consult a qualified medical professional.