Jarrodbarnes/Qwen3-4B-tau2-grpo-v1
TEXT GENERATIONConcurrency Cost:1Model Size:4BQuant:BF16Ctx Length:32kPublished:Jan 16, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Warm

Jarrodbarnes/Qwen3-4B-tau2-grpo-v1 is a 4 billion parameter Qwen3-based language model, fine-tuned specifically for multi-turn tool-use tasks. It achieves 59% Pass@4 on the tau2-bench test split, representing a significant improvement over its base model. This model excels at complex agentic workflows requiring sequential tool interactions, making it suitable for applications needing robust function calling capabilities.

Loading preview...