ia-espirita/riv-ai-v2
ia-espirita/riv-ai-v2 is an 8 billion parameter instruction-tuned causal language model developed by IA.Espirita, based on Llama 3.1. Fine-tuned via QLoRA, it specializes in responding to questions about Spiritism (Doutrina Espírita) with doctrinal fidelity and source citations. This model is optimized for studying and disseminating Allan Kardec's complete works in Brazilian Portuguese, offering precise and didactic explanations.
Loading preview...
Overview
ia-espirita/riv-ai-v2 is an 8 billion parameter instruction-tuned model, developed by IA.Espirita, specifically designed for the Spiritist Doctrine. It is fine-tuned using QLoRA on the Llama 3.1 8B Instruct base model. This version significantly expands upon its predecessor (v1) by incorporating a dataset four times larger, comprising 4,896 question/answer pairs extracted from the complete works of Allan Kardec, including the five canonical works, complementary texts, and the Revista Espírita (1858–1869).
Key Capabilities
- Doctrinal Precision: Provides answers grounded in Allan Kardec's integral work.
- Source Citation: Automatically references the book, question, and/or chapter for each response.
- Didactic Tone: Explains complex Spiritist concepts in an accessible and friendly manner.
- Modern Analogies: Translates Spiritist concepts into contemporary language when appropriate.
- Brazilian Portuguese: Natively trained and optimized for PT-BR.
Good For
- Study and Research: Ideal for individuals studying Spiritism who require accurate, sourced information.
- Dissemination: Useful for creating educational content or tools about the Spiritist Doctrine.
- Offline Deployment: Available in GGUF format for local inference with tools like Ollama, LM Studio, and llama.cpp.
- Developer Integration: Can be integrated into Python applications using the Transformers library or deployed as an OpenAI-compatible server with vLLM.