mayukareddy/lexis-qwen25-7b-obligation-generator
The mayukareddy/lexis-qwen25-7b-obligation-generator is a 7.6 billion parameter Qwen2.5-7B-Instruct model, fine-tuned using QLoRA for specialized legal obligation extraction. This model excels at generating structured JSON outputs detailing actions, deontic types, owners, deadlines, and risk levels from legal texts. It is specifically optimized for legal document analysis, leveraging a 32768 token context length for comprehensive understanding.
Loading preview...
LEXIS Qwen2.5-7B-Instruct: Obligation Generator
This model, mayukareddy/lexis-qwen25-7b-obligation-generator, is a specialized version of the Qwen2.5-7B-Instruct architecture, fine-tuned using QLoRA for the specific task of extracting legal obligations. It leverages the CUAD QA dataset to achieve high accuracy in identifying and structuring key contractual elements.
Key Capabilities
- Structured JSON Output: Generates detailed JSON objects for each identified obligation, including:
action: The specific action required.deontic_type: The nature of the obligation (e.g., 'shall', 'may').owner: The party responsible for the obligation.deadline: The timeframe for completion.risk_level: An assessment of the associated risk.
- Legal Document Analysis: Optimized for processing and understanding complex legal texts.
- Efficient Fine-tuning: Utilizes QLoRA with 4-bit NF4 quantization, r=64, and alpha=128 for efficient training.
Good For
- Automated extraction of obligations from contracts and legal documents.
- Legal tech applications requiring structured data from unstructured legal text.
- Enhancing legal compliance and risk management processes.