Klear-AgentForge: Forging Agentic Intelligence through Posttraining Scaling
This addresses the lack of open-source post-training details for agentic models, benefiting the open-source community, though it is incremental as it builds on existing methods.
The paper tackles the problem of developing high-performance agentic models for interacting with tools and environments by presenting an open-source pipeline that trains from the Qwen3-8B base model, achieving state-of-the-art performance among similarly sized LLMs and competitiveness with larger models.
Despite the proliferation of powerful agentic models, the lack of critical post-training details hinders the development of strong counterparts in the open-source community. In this study, we present a comprehensive and fully open-source pipeline for training a high-performance agentic model for interacting with external tools and environments, named Klear-Qwen3-AgentForge, starting from the Qwen3-8B base model. We design effective supervised fine-tuning (SFT) with synthetic data followed by multi-turn reinforcement learning (RL) to unlock the potential for multiple diverse agentic tasks. We perform exclusive experiments on various agentic benchmarks in both tool use and coding domains. Klear-Qwen3-AgentForge-8B achieves state-of-the-art performance among LLMs of similar size and remains competitive with significantly larger models.