Overview

MaziyarPanahi/Llama-3-8B-Instruct-v0.3 is an 8 billion parameter instruction-tuned language model. It is a continuation of the MaziyarPanahi/Llama-3-8B-Instruct-DPO series, indicating a focus on instruction following and dialogue optimization through techniques like Direct Preference Optimization (DPO).

Key Features

• Base Model: Developed from the Llama-3-8B-Instruct-DPO series.
• Instruction Tuning: Optimized for understanding and executing user instructions.
• Prompt Template: Utilizes the ChatML format for structured input and output, ensuring consistent conversational flow.
• Quantized Versions: GGUF quantized models are available for efficient deployment and inference on various hardware.

Usage

This model is suitable for applications requiring a capable instruction-following chatbot. Developers can integrate it using the Hugging Face transformers library, leveraging its AutoModelForCausalLM and AutoTokenizer classes. The provided Python example demonstrates how to set up a text generation pipeline with specific ChatML formatting and termination tokens for robust conversational AI.