Salesforce AI Released APIGen-MT and xLAM-2-fc-r Model Series: Advancing Multi-Turn Agent Training with Verified Data Pipelines and Scalable LLM Architectures

AI agents are essential for complex human interactions in business settings, requiring multi-turn conversation management and precise tool usage. Traditional methods struggle with data scarcity and context tracking, leading to errors. Salesforce AI Research introduced APIGen-MT and xLAM-2-fc-r models, generating high-quality, multi-turn interaction datasets and outperforming existing models on key benchmarks. Their open-source framework aims to enhance robustness and accessibility in AI agent training.