Model Overview

lgaalves/mistral-7b_open_platypus is a 7 billion parameter instruction-tuned model built upon the Mistral-7B transformer architecture. Developed by Luiz G A Alves, this model was fine-tuned for 11 hours using LoRA on a Tesla V100-SXM2-16GB GPU.

Key Capabilities

• Instruction Following: Optimized for understanding and executing instructions.
• STEM and Logic Reasoning: Fine-tuned on the garage-bAInd/Open-Platypus dataset, which focuses on STEM and logic-based content, enhancing its performance in these areas.
• Benchmark Performance: Shows improved MMLU scores (59.76) compared to the base Mistral-7B-v0.1 (64.16) and Platypus2-7B (49.83) in the provided benchmark table, indicating strong reasoning abilities.

Use Cases

This model is well-suited for applications requiring robust analytical and reasoning capabilities, particularly in scientific, technical, engineering, and mathematical domains. Its instruction-tuned nature makes it effective for tasks where precise command execution and logical inference are crucial. Developers should perform safety testing and tuning for specific applications due to inherent LLM limitations.