Spaces:
Running
on
Zero
Running
on
Zero
File size: 5,420 Bytes
a578794 5ef3aa7 a578794 71f359b a578794 5ef3aa7 a578794 5ef3aa7 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 |
---
title: Psychology AI Assistant
emoji: π§
colorFrom: purple
colorTo: blue
sdk: gradio
sdk_version: 5.34.2
app_file: app.py
pinned: false
license: apache-2.0
models:
- Qwen/Qwen2.5-0.5B
- phxdev/psychology-qwen-0.5b
tags:
- psychology
- mental-health
- therapy
- counseling
- peft
- qwen
- fine-tuned
---
# π§ Psychology AI Assistant
An empathetic AI assistant fine-tuned for psychology and mental health support. Based on Qwen 2.5-0.5B with specialized training using PEFT/LoRA techniques.
## β¨ Features
- **Empathetic Responses**: Fine-tuned specifically for psychology and mental health conversations
- **Multiple Support Styles**: Choose from therapeutic, supportive friend, crisis support, anxiety coaching, or mindfulness guidance
- **Evidence-Based Approach**: Responses grounded in psychological principles and therapeutic techniques
- **Interactive Interface**: Modern, accessible design with conversation starters and customizable settings
- **Privacy-Focused**: No conversation data is stored or logged
## π― Model Details
- **Base Model**: [Qwen/Qwen2.5-0.5B](https://huggingface.co/Qwen/Qwen2.5-0.5B)
- **Fine-tuned Model**: [phxdev/psychology-qwen-0.5b](https://huggingface.co/phxdev/psychology-qwen-0.5b)
- **Training Method**: PEFT (Parameter Efficient Fine-Tuning) with LoRA
- **Specialization**: Psychology, mental health, therapeutic communication
- **Parameters**: ~500M parameters (0.5B)
- **Context Length**: 2048 tokens
## π How to Use
1. **Start a Conversation**: Type your thoughts, concerns, or questions in the chat box
2. **Choose Your Style**: Select the type of support you prefer from the dropdown menu
3. **Adjust Settings**: Customize response creativity and length to your preference
4. **Try Examples**: Click on any of the conversation starters for common scenarios
### Support Styles Available:
- **π¬ Therapeutic**: Professional therapist approach with validation and structured guidance
- **π₯ Supportive Friend**: Warm, caring friend who studied psychology
- **π Crisis Support**: Immediate support and grounding techniques for crisis situations
- **π° Anxiety Coach**: Specialized techniques for managing anxiety and panic
- **π§ Mindfulness Guide**: Present-moment awareness and self-compassion practices
## β οΈ Important Disclaimer
This AI assistant provides supportive guidance based on psychological principles, but it is **not a replacement for professional therapy or medical advice**.
If you're experiencing a mental health crisis, please contact:
- **988 Suicide & Crisis Lifeline**: Call or text 988
- **Crisis Text Line**: Text HOME to 741741
- **Emergency Services**: Call 911 (US) or your local emergency number
## π οΈ Technical Implementation
The model uses advanced prompt engineering and fine-tuning techniques to provide contextually appropriate responses:
- **PEFT Training**: Efficient fine-tuning preserving base model capabilities while adding psychology specialization
- **Prompt Templates**: Multiple specialized templates for different therapeutic approaches
- **Response Quality**: Built-in safeguards against generic or unhelpful responses
- **Mobile Optimized**: Responsive design works well on all devices
## π¨ Interface Features
- **Clean Design**: Modern, calming interface designed for sensitive conversations
- **Conversation History**: Full chat history maintained during your session
- **Example Prompts**: Curated conversation starters for common mental health topics
- **Accessibility**: Screen reader friendly and keyboard navigable
- **Privacy**: No data logging or conversation storage
## π§ Model Performance
The fine-tuned model shows significant improvements over the base Qwen model in:
- **Empathy**: Better emotional validation and understanding
- **Practical Advice**: More actionable, specific guidance
- **Psychological Awareness**: Understanding of mental health concepts and terminology
- **Conversational Flow**: Natural, supportive dialogue without generic responses
## π Example Use Cases
- **Stress Management**: Work stress, academic pressure, life transitions
- **Anxiety Support**: Panic attacks, social anxiety, general worry
- **Relationship Guidance**: Communication skills, boundary setting, conflict resolution
- **Self-Esteem**: Building confidence, overcoming negative self-talk
- **Mindfulness**: Present-moment awareness, meditation guidance, self-compassion
- **Grief Support**: Processing loss, coping with change
## π€ Contributing & Feedback
This model is part of ongoing research into AI applications for mental health support. If you have feedback or suggestions:
- **Model Issues**: Report on the [model repository](https://huggingface.co/phxdev/psychology-qwen-0.5b)
- **Interface Bugs**: Open an issue on this Space
- **Research Collaboration**: Contact [@phxdev](https://huggingface.co/phxdev)
## π Related Resources
- **Base Model**: [Qwen/Qwen2.5-0.5B](https://huggingface.co/Qwen/Qwen2.5-0.5B)
- **PEFT Library**: [huggingface/peft](https://github.com/huggingface/peft)
- **Mental Health Resources**: [National Alliance on Mental Illness](https://nami.org)
## π License
This Space and the fine-tuned model are released under the Apache 2.0 License. The base Qwen model follows its original licensing terms.
---
**Built with β€οΈ by [@phxdev](https://huggingface.co/phxdev) using Gradio and HuggingFace Transformers** |