doandune/LexGuard-llama3-Risk-Adapter

TEXT GENERATIONConcurrency Cost:1Model Size:8BQuant:FP8Ctx Length:8kPublished:Mar 14, 2026License:apache-2.0Architecture:Transformer Open Weights Cold

The doandune/LexGuard-llama3-Risk-Adapter is an 8 billion parameter language model, finetuned from unsloth/llama-3-8b-Instruct-bnb-4bit. Developed by doandune, this model was trained using Unsloth and Huggingface's TRL library for accelerated finetuning. It is designed to adapt the Llama 3 architecture for specific risk-related applications, leveraging its 8192 token context length.

Loading preview...

Model Overview

The doandune/LexGuard-llama3-Risk-Adapter is an 8 billion parameter language model, developed by doandune. It is finetuned from the unsloth/llama-3-8b-Instruct-bnb-4bit base model, leveraging the Llama 3 architecture. This model was specifically trained using the Unsloth library, which enables significantly faster finetuning, alongside Huggingface's TRL library.

Key Capabilities

  • Efficient Finetuning: Utilizes Unsloth for accelerated training, making it resource-efficient for adaptation.
  • Llama 3 Foundation: Benefits from the robust capabilities and performance of the Llama 3 instruction-tuned base model.
  • Risk-Specific Adaptation: Designed as an "adapter" model, suggesting its purpose is to specialize the Llama 3 base for particular risk-related use cases.

Good For

  • Developers looking for a Llama 3-based model optimized for specific risk assessment or compliance tasks.
  • Applications requiring a performant 8B parameter model with an 8192 token context length, finetuned for specialized domains.
  • Users interested in models trained with Unsloth for efficiency and speed.