piik/Qwen-2.5-3B-Indonesian-Legal-SFT
The piik/Qwen-2.5-3B-Indonesian-Legal-SFT is a 3.1 billion parameter Qwen2.5-based language model, developed by piik. It is fine-tuned from unsloth/Qwen2.5-3B-Instruct-bnb-4bit, leveraging Unsloth and Huggingface's TRL library for faster training. This model is specifically optimized for Indonesian legal text generation and understanding, making it suitable for specialized applications in the legal domain.
Loading preview...
Model Overview
The piik/Qwen-2.5-3B-Indonesian-Legal-SFT is a 3.1 billion parameter language model developed by piik. It is built upon the Qwen2.5 architecture and was fine-tuned from the unsloth/Qwen2.5-3B-Instruct-bnb-4bit base model. The fine-tuning process utilized Unsloth and Huggingface's TRL library, which enabled a 2x faster training speed.
Key Capabilities
- Specialized Domain: This model is specifically fine-tuned for the Indonesian legal domain, suggesting enhanced performance on legal texts and queries in Indonesian.
- Efficient Training: Leverages Unsloth for accelerated training, indicating potential for further efficient fine-tuning or deployment.
- Qwen2.5 Architecture: Benefits from the robust capabilities of the Qwen2.5 model family.
Good For
- Indonesian Legal Applications: Ideal for tasks requiring understanding or generation of Indonesian legal documents, contracts, regulations, or legal research.
- Domain-Specific NLP: Suitable for developers building applications that require high accuracy in a specialized, low-resource language legal context.
- Efficient Deployment: Its 3.1 billion parameter size makes it a relatively efficient choice for deployment compared to larger models, especially when combined with its optimized training methodology.