liunanfu1992/Qwen3-8B-LOPD

TEXT GENERATIONPricing:Input $0.468 / Output $1.82Concurrent Unit Cost:1Model Size:8BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Aug 11, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Featherless Exclusive Cold

liunanfu1992/Qwen3-8B-LOPD is an 8 billion parameter causal language model fine-tuned from Qwen3-8B. It utilizes Latent On-Policy Self-Distillation (LOPD) to specialize in agentic tool-use tasks. This model is designed for applications requiring advanced tool interaction and decision-making capabilities.

Loading preview...

Model Overview

liunanfu1992/Qwen3-8B-LOPD is an 8 billion parameter language model built upon the Qwen3-8B architecture. Its key differentiator is the application of Latent On-Policy Self-Distillation (LOPD) during fine-tuning, a technique specifically aimed at enhancing performance in agentic tool-use scenarios.

Key Capabilities

  • Agentic Tool-Use Specialization: The model is optimized for tasks that involve interacting with and utilizing external tools, making it suitable for complex automation and agent-based systems.
  • LOPD Fine-tuning: Leverages a specialized distillation method to improve its ability to reason and act within tool-augmented environments.
  • Qwen3-8B Foundation: Benefits from the robust base capabilities of the Qwen3-8B model, providing a strong general language understanding foundation.

When to Use This Model

This model is particularly well-suited for:

  • Developing AI agents that need to interact with APIs or other tools.
  • Applications requiring advanced decision-making in environments where external functions are available.
  • Research into self-distillation techniques and their impact on agentic capabilities.

For more in-depth technical details on the LOPD method, refer to the GitHub repository and the associated research paper.