Still need to change a lot. Has some bugs in the UI.

Skin Disease and Skincare Chatbot

This project is a chatbot specialized in diagnosing skin diseases and addressing skincare issues. The chatbot is powered by a fine-tuned version of OpenBioLLM, an expert language model in the medical domain. The model has been fine-tuned specifically for skin problems and skincare by training it on curated data from the internet. Additionally, the chatbot features a vision component that can identify diseases in images using a fine-tuned version of the Peli-Gemma vision model, trained on 22,000 skin disease images. The chatbot also maintains chat history to keep conversations in context.

Model Links

• Peli-Gemma Vision Model (fine-tuned for dermatology)
• OpenBioLLM-Derm (fine-tuned for dermatology)

Running the Chatbot Locally

To run the chatbot locally, remove the ngrok command line at the end of the main.py file and execute it.

Running the Chatbot on Google Colab (Free Version)

If you want to run the chatbot on Google Colab, follow these steps:

Step 1: Get an API Key from Ngrok

Since Colab does not have an inbuilt browser, you will need an API key from Ngrok.

Step 2: Upload static.zip to Colab

Upload the static.zip file to your Colab environment and run the following commands:

!unzip static.zip
!pip install bitsandbytes
!pip install fastapi
!pip install uvicorn
!pip install langchain
!pip install langchain_community
!pip install python-multipart
!pip install pyngrok
!ngrok config add-authtoken Your_ngrok_key
!python main.py

Screen Shots: