Whisper Psychology Chatbot Overview

Whisper is a specialized mental health chatbot developed by DeepFinders at SLTC Research University. It is fine-tuned from the google/gemma-3-1b-it base model, focusing on empathetic and supportive responses within mental health conversations.

Key Capabilities

Specialized Mental Health Support: Designed to engage in psychology-focused dialogues, offering supportive interactions.
Empathetic Communication: Fine-tuned on a dedicated psychology dataset (jkhedri/psychology-dataset) to foster empathetic responses.
Efficient Fine-tuning: Utilizes LoRA (Low-Rank Adaptation) with specific configurations (r=8, lora_alpha=16) targeting key attention modules for efficient adaptation.
Gemma-3-1B-IT Base: Built upon a robust 1 billion parameter instruction-tuned model, providing a strong foundation for conversational AI.

Good For

Educational and Research Purposes: Ideal for academic studies and research into AI applications in mental health.
Developing Supportive Chat Interfaces: Suitable for building prototypes or components of mental health support systems.
Exploring Empathetic AI: Useful for developers interested in creating AI that can provide nuanced and understanding conversational support.

Limitations

It is crucial to note that Whisper is not a substitute for professional mental health care. It is intended for educational and research use, and users should implement appropriate safeguards due to the potential for incorrect or inappropriate responses.