Inspiration

The inspiration for this project came from witnessing the profound impact mental health issues have on individuals and communities. Personally experiencing moments of grief and anxiety, I felt a strong need to create a tool that could act as a bridge between users and timely emotional support. The loss of my mother, Pauly Augustine, further reinforced the significance of accessible mental health resources. I wanted to design a solution that could cater to users at any time, empowering them with coping mechanisms when they need them most.

What it does

The Gemini Mental Health Assistant is a versatile mental health support app powered by AI. It provides users with personalized emotional support and tools for managing mental well-being

How we built it

The primary language for building the app is Python.We used PyTorch for model fine-tuning and building the core AI functionalities.The conversational AI aspect of the app was developed using pre-trained language models like GEMMA and Gemini, which were further fine-tuned on mental health counseling conversations.

Challenges we ran into

Fine-tuning for Emotional Sensitivity.Maintaining Data Privacy.Optimizing AI Model.Ensuring Ethical AI Use.Diverse User Needs.Handling Ambiguous User Inpu

Accomplishments that we're proud of

What we learned

What's next for Gemini Mental Health Assistant

Built With

  • adaptation)peft
  • cloud
  • face
  • face)lora
  • fine-tuninggoogle
  • gcp)hugging
  • hugging
  • low-rank
  • parameter-efficient
  • platform
  • python
  • pytorch
  • transformers
Share this project:

Updates

Submission history