SeongryongJung/Qwen3-8B-Tooluse-RLSD-TR

TEXT GENERATIONConcurrent Unit Cost:1Model Size:8BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Jul 3, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

SeongryongJung/Qwen3-8B-Tooluse-RLSD-TR is an 8 billion parameter Qwen3-based language model specifically fine-tuned for tool use capabilities. This model leverages RLSD_TR training with a batch size of 32 to enhance its ability to interact with and utilize external tools. It achieves a validation mean@16 score of 63.42% on tool-use tasks, making it suitable for applications requiring robust function calling and external API interaction.

Loading preview...

Model Overview

SeongryongJung/Qwen3-8B-Tooluse-RLSD-TR is an 8 billion parameter model built upon the Qwen3 architecture, specifically optimized for tool use. This model has undergone specialized training using the RLSD_TR (Reinforcement Learning from Simulated Demonstrations with Trajectory Rejection) method, with a batch size of 32, to significantly improve its performance in scenarios requiring interaction with external tools or APIs.

Key Capabilities

  • Enhanced Tool Use: Fine-tuned to understand and execute tool-related instructions, enabling it to effectively call functions and interact with external systems.
  • RLSD_TR Training: Utilizes a reinforcement learning approach to refine its tool-use capabilities, focusing on robust and reliable performance.
  • Performance on Tool-use Benchmarks: Achieves a 63.42% validation mean@16 on the dedicated tool-use dataset, indicating strong performance in this domain.

Good For

  • Applications requiring function calling or API interaction.
  • Developing agents that need to leverage external tools to complete tasks.
  • Scenarios where a model's ability to accurately interpret and execute tool-related commands is critical.