Llama 3 Smaug 8B

🚀 Llama-3-Smaug-8B

This model is built using the Smaug recipe to enhance performance in real - world multi - turn conversations, applied to Meta Llama 3.

🚀 Quick Start

This section will provide a quick overview of the model and its purpose. The Llama-3-Smaug-8B model is constructed by applying the Smaug recipe to meta-llama/Meta-Llama-3-8B-Instruct to improve its performance in real - world multi - turn conversations.

✨ Features

• Built with Meta Llama 3, leveraging its powerful architecture.
• Uses the Smaug recipe to enhance performance in multi - turn conversations.

📚 Documentation

Model Description

• Developed by: Abacus.AI
• License: https://llama.meta.com/llama3/license/
• Finetuned from model: meta-llama/Meta-Llama-3-8B-Instruct.

Evaluation

MT - Bench

########## First turn ##########
                   score
model             turn
Llama-3-Smaug-8B 1   8.77500
Meta-Llama-3-8B-Instruct 1   8.31250
########## Second turn ##########
                   score
model             turn
Meta-Llama-3-8B-Instruct 2   7.8875 
Llama-3-Smaug-8B 2   7.8875
########## Average ##########
                 score
model
Llama-3-Smaug-8B  8.331250
Meta-Llama-3-8B-Instruct 8.10

This version of Smaug uses new techniques and new data compared to Smaug-72B, and more information will be released later on. For now, see the previous Smaug paper: https://arxiv.org/abs/2402.13228.

📄 License

The model is under the Llama 2 license. The specific license details can be found at: https://llama.meta.com/llama3/license/