keerthi4/phi1_5-lahacks

Model Card for Model ID

This model outputs first aid instructions based on your needs.

Model Details

Model Description

Our LA Hacks 2024 project is a first-aid handbook designed to provide immediate first aid guidance in emergency situations. Users can submit a text description or a photo of their health emergency, and the system will generate tailored first aid responses. This model is convenient, user-friendly, and an essential tool for non-critical emergencies.

This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.

• Developed by: Keerthi Nalabotu, Diana Vins, David Wang, Rohan Nair

• Shared by [optional]:

• Model type: Large Language Model

• Language(s) (NLP): English

• License: MIT License

• Finetuned from model: microsoft/phi-1_5

• Repository: badri55/First_aid__dataset

• Paper [optional]:

• Demo [optional]: [More Information Needed]

Uses

This model is specifically designed to provide first aid instructions for emergency and non-emergency situations based on user inputs. It aims to make first aid knowledge readily accessible to everyone, anywhere.

Direct Use

The model can be directly interacted with through a user interface where users can input symptoms or describe an emergency to receive immediate guidance.

Downstream Use [optional]

Future features include pasting a dataset name into the user interface and creating your own fine tuned model without any hastle. Additionally, future developments includeintegrating the model into mobile apps and health platforms, enabling users to receive personalized first aid guidance on the go.

Out-of-Scope Use

Misuse and uses that the model will not work well for would include anything non-medical related. Malicious uses include using the application with intent of violence, harm of any kind, or any illegal activity. The model is not intended to replace professional medical advice or emergency services. Its use should be limited to non-critical first aid situations.

Bias, Risks, and Limitations

Users should be aware that while the model provides first aid assistance, it is not a substitute for professional medical advice or emergency services. The model's suggestions should be used as a preliminary step or in situations where professional medical help is not immediately available.

Recommendations

We recommend users to always seek professional medical advice when possible. The model is designed as an aid, not a replacement for human medical professionals.

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

from transformers import AutoModelForCausalLM, AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained("keerthi4/phi1_5-lahacks") model = AutoModelForCausalLM.from_pretrained("keerthi4/phi1_5-lahacks")

Example usage

inputs = tokenizer("Describe your emergency situation here", return_tensors="pt") outputs = model.generate(inputs["input_ids"]) print(tokenizer.decode(outputs[0]))

Training Details

Training Data

Link: link text

The model was trained on all 44 rows of this dataset. The model was trained on a diverse dataset of first aid scenarios and medical emergencies sourced from public health databases and manuals.

Training Procedure

The model was finetuned on the microsoft/phi-1_5 model using a custom dataset that includes structured first aid steps and responses to a wide variety of health emergencies.

Training Hyperparameters

• Training regime: [More Information Needed]

Evaluation

Performance was measured using accuracy of the first aid instructions and user feedback on the utility and clarity of the instructions provided.

Testing Data, Factors & Metrics

Testing Data

Testing Data: The model was tested using the creator's input to the fine-tuned model.

Factors

Metrics

Results

Summary

Environmental Impact

260 grams of CO2eq total emissions. Efforts were made to minimize the carbon footprint during training by utilizing efficient hardware and optimizing compute time.

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

• Hardware Type: [More Information Needed]
• Hours used: 10 hours
• Cloud Provider: Intel Developer Cloud
• Compute Region: [More Information Needed]
• Carbon Emitted: 260 grams CO2eq

Technical Specifications [optional]

Model Architecture and Objective

[More Information Needed]

Compute Infrastructure

Hardware

[More Information Needed]

Software

[More Information Needed]

Model Card Contact

Keerthi:

Diana Vins: diana.v.vins@gmail.com

David Wang:

Rohan Nair: