metadata

license: apache-2.0
datasets:
  - FreedomIntelligence/medical-o1-reasoning-SFT
language:
  - en
metrics:
  - accuracy
base_model:
  - mistralai/Mistral-7B-Instruct-v0.3
pipeline_tag: text-generation
library_name: transformers
tags:
  - chemistry
  - medical
  - Doctor
  - AI_Doctor
  - Illness
  - MedicalAI
  - MBBS
  - AI_AGENT

Model Card for Model ID

🩺 Medical Diagnosis AI Model - Powered by Mistral-7B & LoRA 🚀 🔹 Model Overview: Base Model: Mistral-7B (7.7 billion parameters) Fine-Tuning Method: LoRA (Low-Rank Adaptation) Quantization: bnb_4bit (reduces memory footprint while retaining performance) 🔹 Parameter Details: Original Mistral-7B Parameters: 7.7 billion LoRA Fine-Tuned Parameters: ~~4.48% of total model parameters (~~340 million) Final Merged Model Size (bnb_4bit Quantized): ~4.5GB 🔹 Key Features: ✅ Accurate Diagnoses for symptoms like chest pain, dizziness, and breathlessness ✅ Step-by-Step Medical Reasoning using Chain-of-Thought (CoT) prompting ✅ Efficient Inference with reduced VRAM usage (ideal for GPUs with limited memory)

🔹 Use Case: Designed to assist healthcare professionals by offering clear, evidence-backed insights for improved clinical decision-making. 📝 Note: While this model offers valuable insights, it's intended to support — not replace — professional medical judgment.

This modelcard aims to be a base template for new models. It has been generated using this raw template.

Model Details

Original Mistral-7B Parameters: 7.7 billion LoRA Fine-Tuned Parameters: 4.48% of total model parameters (~340 million) Final Merged Model Size (bnb_4bit Quantized): ~4.5GB 🔹 Key Features: ✅ Accurate Diagnoses for symptoms like chest pain, dizziness, and breathlessness ✅ Step-by-Step Medical Reasoning using Chain-of-Thought (CoT) prompting ✅ Efficient Inference with reduced VRAM usage (ideal for GPUs with limited memory)

Model Description

This model leverages the powerful Mistral-7B language model, known for its strong reasoning capabilities and deep language understanding. Through LoRA fine-tuning, the model now excels in medical-specific tasks like: ✅ Diagnosing conditions from symptoms such as chest pain, dizziness, and shortness of breath ✅ Providing detailed, step-by-step medical reasoning using Chain-of-Thought (CoT) prompting ✅ Generating confident, evidence-backed answers with improved precision

Developed by: [Ritvik Gaur]
Funded by [optional]: [More Information Needed]
Shared by [optional]: [More Information Needed]
Model type: [Medical LLM]
Language(s) (NLP): [More Information Needed]
License: [More Information Needed]
Finetuned from model [optional]: [Mistral-7B-Instruct-v3]

Model Sources [optional]

Repository: [More Information Needed]
Paper [optional]: [More Information Needed]
Demo [optional]: [More Information Needed]

Uses

Direct Use

[More Information Needed]

Downstream Use [optional]

[More Information Needed]

Out-of-Scope Use

[More Information Needed]

Bias, Risks, and Limitations

Please dont fully rely on this model for real life illness, this model is just for support of real verifies health applications that requires LLM.

[More Information Needed]

Recommendations

Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.

How to Get Started with the Model

Use the code below to get started with the model.

!pip install -q -U bitsandbytes

!pip install -q -U peft

!pip install -q -U trl

!pip install -q -U tensorboardX

!pip install -q wandb

from transformers import AutoModelForCausalLM, AutoTokenizer

✅ Load the uploaded model

model = AutoModelForCausalLM.from_pretrained("ritvik77/Medical_Doctor_AI_LoRA-Mistral-7B-Instruct_FullModel") tokenizer = AutoTokenizer.from_pretrained("ritvik77/Medical_Doctor_AI_LoRA-Mistral-7B-Instruct_FullModel")

✅ Sample inference

prompt = "Patient reports chest pain and dizziness with nose bleeding, What’s the likely diagnosis is it cancer ?" inputs = tokenizer(prompt, return_tensors="pt").to(model.device)

outputs = model.generate(**inputs, max_new_tokens=300) print(tokenizer.decode(outputs[0], skip_special_tokens=True)) Python code for usage: from transformers import AutoModelForCausalLM, AutoTokenizer

✅ Load the uploaded model

✅ Sample inference

prompt = "Patient reports chest pain and dizziness. What’s the likely diagnosis?" inputs = tokenizer(prompt, return_tensors="pt").to(model.device) outputs = model.generate(**inputs, max_new_tokens=300) print(tokenizer.decode(outputs[0], skip_special_tokens=True))

[More Information Needed]

Training Details

Training Data

[More Information Needed]

Training Procedure

Preprocessing [optional]

[More Information Needed]

Training Hyperparameters

Training regime: [More Information Needed]
Parameter Value Description

Base Model mistralai/Mistral-7B-Instruct Chosen for its strong reasoning capabilities.

Fine-Tuning Framework LoRA (Low-Rank Adaptation) Efficiently fine-tuned only ~4.48% of total parameters.

Quantization bnb_4bit Enabled for reduced VRAM consumption.

Train Batch Size 12 Optimized to balance GPU utilization and convergence.

Eval Batch Size 12 Matches training batch size to ensure stable evaluation.

Gradient Accumulation Steps 3 Effective batch size = 36 for improved stability.

Learning Rate 3e-5 Lowered to ensure smoother convergence

Warmup Ratio 0.2 Gradual learning rate ramp-up for improved stability

Scheduler Type Cosine Ensures smooth and controlled learning rate decay

Number of Epochs 5 Balanced to ensure convergence without overfitting

Max Gradient Norm 0.5 Prevents exploding gradients

Weight Decay 0.08 Regularization for improved generalization

bf16 Precision True Maximizes GPU utilization and precision

Gradient Checkpointing Enabled Reduces memory usage during training

🔎 LoRA Configuration Parameter Value Description Rank Dimension 128 Balanced for strong expressiveness without excessive memory overhead LoRA Alpha 128 Ensures stable gradient updates LoRA Dropout 0.1 Helps prevent overfitting

Speeds, Sizes, Times [optional]

[More Information Needed]

Evaluation

Testing Data, Factors & Metrics

Testing Data

[More Information Needed]

Factors

[More Information Needed]

Metrics

[More Information Needed]

Results

[More Information Needed]

Summary

Model Examination [optional]

[More Information Needed]

Environmental Impact

Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).

Hardware Type: [More Information Needed]
Hours used: [More Information Needed]
Cloud Provider: [More Information Needed]
Compute Region: [More Information Needed]
Carbon Emitted: [More Information Needed]

Technical Specifications [optional]

Model Architecture and Objective

[More Information Needed]

Compute Infrastructure

[More Information Needed]

Hardware

[More Information Needed]

Software

[More Information Needed]

Citation [optional]

BibTeX:

[More Information Needed]

APA:

[More Information Needed]

Glossary [optional]

[More Information Needed]

More Information [optional]

[More Information Needed]

Model Card Authors [optional]

[More Information Needed]

Model Card Contact

[More Information Needed]