AgentOhana: Design Unified Data and Training Pipeline for Effective Agent Learning
This work addresses the problem of inefficient agent training for researchers and developers by providing a comprehensive pipeline, though it appears incremental as it builds on existing methods for data handling.
The paper tackles the challenge of heterogeneous data sources in training autonomous agents with large language models by introducing AgentOhana, a unified data and training pipeline that standardizes multi-turn trajectories, and presents xLAM-v0.1, which demonstrates exceptional performance across various benchmarks.
Autonomous agents powered by large language models (LLMs) have garnered significant research attention. However, fully harnessing the potential of LLMs for agent-based tasks presents inherent challenges due to the heterogeneous nature of diverse data sources featuring multi-turn trajectories. In this paper, we introduce \textbf{AgentOhana} as a comprehensive solution to address these challenges. \textit{AgentOhana} aggregates agent trajectories from distinct environments, spanning a wide array of scenarios. It meticulously standardizes and unifies these trajectories into a consistent format, streamlining the creation of a generic data loader optimized for agent training. Leveraging the data unification, our training pipeline maintains equilibrium across different data sources and preserves independent randomness across devices during dataset partitioning and model training. Additionally, we present \textbf{xLAM-v0.1}, a large action model tailored for AI agents, which demonstrates exceptional performance across various benchmarks. Begin the exploration at \url{https://github.com/SalesforceAIResearch/xLAM}.