Llama-3.1-8B-ArtTherapy Overview
mariadelcarmenramirez/Llama-3.1-8B-ArtTherapy is an 8 billion parameter language model, fine-tuned from Meta's Llama-3.1-8B-Instruct. Its core specialization lies in facilitating structured art therapy dialogues, generating responses that align with specific therapeutic phases. This model was created by merging a QLoRA adapter into the base Llama-3.1-8B-Instruct weights.
Key Capabilities
- Phase-aware Therapeutic Dialogue: Supports structured conversations across four distinct art therapy phases (Intake, Exploration, Reflection, Closure) using dedicated
<FASE_N> tags in system prompts. - Art Therapy Specialization: Guides users to interact with artwork, explore relationships within it, and engage in symbolic dialogue, focusing on the user's experience rather than interpretation.
- Spanish Language Focus: Trained exclusively on Spanish-language art therapy dialogues, ensuring high performance in this domain.
Good for
- Art Therapy Assistance: Ideal for applications requiring an AI assistant to facilitate art therapy sessions, providing structured prompts and maintaining a therapeutic arc.
- Dialogue-Centric Art Exploration: Useful for guiding users through reflective and interactive processes with their artwork.
- Research and Development: Can serve as a base for further research into AI-assisted art therapy methodologies, particularly in Spanish-speaking contexts.
Limitations
It is crucial to note that this model is not intended to replace licensed art therapists or mental health professionals. It should not be used for clinical diagnosis, crisis intervention, or as a standalone therapeutic agent. Its responses are not clinically validated and performance in languages other than Spanish or outside the art therapy domain is not guaranteed. Always include a <FASE_N> tag for optimal results.