syaefur/pgabl-wafa-legal-assistant-grpo

TEXT GENERATIONConcurrent Unit Cost:1Model Size:1.5BQuant:BF16Context Size:32kTool Calling:SupportedPublished:Jul 11, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

The syaefur/pgabl-wafa-legal-assistant-grpo is a 1.5 billion parameter Qwen2 model developed by syaefur, fine-tuned from syaefur/pgabl-wafa-legal-assistant-sft-exp1. This model was trained using Unsloth and Huggingface's TRL library, enabling 2x faster training. With a context length of 32768 tokens, it is designed for legal assistant applications.

Loading preview...

Overview

The syaefur/pgabl-wafa-legal-assistant-grpo is a 1.5 billion parameter Qwen2 model, developed by syaefur. It has been fine-tuned from the syaefur/pgabl-wafa-legal-assistant-sft-exp1 model, indicating a specialization in legal assistant functionalities. The model leverages a substantial context length of 32768 tokens, allowing it to process extensive legal documents and queries.

Training Methodology

This model's training process utilized Unsloth and Huggingface's TRL library, which significantly accelerated the fine-tuning, achieving 2x faster training speeds. This efficient training approach suggests a focus on optimizing development cycles while maintaining performance.

Key Characteristics

  • Model Family: Qwen2 architecture
  • Parameter Count: 1.5 billion parameters
  • Context Length: 32768 tokens
  • Training Tools: Unsloth and Huggingface's TRL library for accelerated fine-tuning
  • License: Apache-2.0

Potential Use Cases

Given its fine-tuning history and name, this model is likely optimized for tasks related to legal assistance, such as:

  • Answering legal questions
  • Summarizing legal documents
  • Assisting with legal research
  • Generating legal-themed text

Finetunes
1 models