diansm/legal-llm-indonesia-qwen-finetuned

TEXT GENERATIONConcurrency Cost:1Model Size:1.5BQuant:BF16Ctx Length:32kTool Calling:SupportedPublished:Jun 4, 2026License:apache-2.0Architecture:Transformer Open Weights Cold

The diansm/legal-llm-indonesia-qwen-finetuned is a 1.5 billion parameter Qwen2.5-Instruct model, developed by diansm, specifically fine-tuned for legal applications in Indonesia. This model leverages Unsloth and Huggingface's TRL library for efficient training, offering a specialized solution for processing and generating Indonesian legal text. With a context length of 32768 tokens, it is designed to handle extensive legal documents and queries.

Loading preview...

Model Overview

The dianm/legal-llm-indonesia-qwen-finetuned is a specialized language model based on the Qwen2.5-1.5B-Instruct architecture, developed by diansm. It has been fine-tuned specifically for legal applications within Indonesia, building upon the unsloth/Qwen2.5-1.5B-Instruct-bnb-4bit base model.

Key Capabilities

  • Indonesian Legal Domain Specialization: The primary differentiator of this model is its fine-tuning for the Indonesian legal context, making it suitable for tasks requiring an understanding of local legal nuances.
  • Efficient Training: The model was trained using Unsloth and Huggingface's TRL library, indicating an optimized and potentially faster training process.
  • Qwen2.5 Architecture: Inherits the capabilities of the Qwen2.5-Instruct family, known for its general language understanding and instruction-following abilities.
  • 1.5 Billion Parameters: A compact yet capable model size, balancing performance with computational efficiency.

Good For

  • Indonesian Legal Text Processing: Ideal for tasks such as legal document analysis, summarization, question answering, or generation within the Indonesian legal framework.
  • Developers requiring specialized LLMs: Suitable for applications where a general-purpose LLM might lack the specific domain knowledge for Indonesian legal matters.