Ridhsaid3/Llama-3-8B-Indo-Legal-GRPO

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:8BQuant:FP8Ctx Length:8kPublished:May 22, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Warm

Ridhsaid3/Llama-3-8B-Indo-Legal-GRPO is an 8 billion parameter Llama 3 model developed by Ridhsaid3, fine-tuned for Indonesian legal applications. This model was trained using Unsloth and Huggingface's TRL library, enabling faster fine-tuning. It is specifically designed to excel in tasks related to Indonesian legal contexts, building upon its predecessor, Ridhsaid3/Llama-3-8B-Indo-Legal-SFT. With an 8192 token context length, it offers robust performance for specialized legal text processing.

Loading preview...

Ridhsaid3/Llama-3-8B-Indo-Legal-GRPO Overview

This model is an 8 billion parameter Llama 3 variant, developed by Ridhsaid3, specifically fine-tuned for applications within the Indonesian legal domain. It builds upon the previously fine-tuned Ridhsaid3/Llama-3-8B-Indo-Legal-SFT model, indicating a specialized focus on legal language and concepts relevant to Indonesia.

Key Characteristics

  • Architecture: Based on the Llama 3 family, providing a strong foundation for language understanding and generation.
  • Parameter Count: Features 8 billion parameters, balancing performance with computational efficiency.
  • Context Length: Supports an 8192 token context window, suitable for processing moderately long legal documents or queries.
  • Training Efficiency: The model was fine-tuned using Unsloth and Huggingface's TRL library, which facilitated a 2x faster training process.

Primary Use Case

This model is primarily intended for tasks requiring deep understanding and generation of text within the Indonesian legal framework. Its specialized fine-tuning makes it particularly suitable for:

  • Analyzing Indonesian legal documents.
  • Answering questions related to Indonesian law.
  • Assisting in legal research specific to Indonesia.
  • Processing and generating legal texts in the Indonesian language.