Alfikrah/qwen2.5-3b-legal-id-grpo

TEXT GENERATIONPricing:Input $0.32 / Cached $0.064 / Output $1.6Concurrent Unit Cost:1Model Size:3.1BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Aug 9, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

Alfikrah/qwen2.5-3b-legal-id-grpo is a 3.1 billion parameter Qwen2.5 model developed by Alfikrah, fine-tuned for legal applications in Indonesian. This model was further optimized using Unsloth and Huggingface's TRL library for faster training. It is designed to process legal-specific text in Indonesian, building upon its base as a legal-focused model.

Loading preview...

Model Overview

Alfikrah/qwen2.5-3b-legal-id-grpo is a specialized large language model developed by Alfikrah. It is based on the Qwen2.5 architecture with 3.1 billion parameters and a context length of 32768 tokens. This model is a further fine-tuned version of Alfikrah/qwen2.5-3b-legal-id-sft, indicating its focus on legal domain understanding.

Key Characteristics

  • Domain-Specific: Primarily designed for legal applications, specifically within the Indonesian context.
  • Optimized Training: The model was trained with performance enhancements using Unsloth and Huggingface's TRL library, resulting in a 2x faster training process.
  • Base Model: It builds upon a previously fine-tuned legal Indonesian model, suggesting a deeper specialization in this area.

Use Cases

This model is particularly well-suited for tasks requiring an understanding of Indonesian legal texts. Potential applications include:

  • Legal document analysis and summarization.
  • Answering questions related to Indonesian law.
  • Assisting with legal research in the Indonesian language.

Finetunes
1 models