WeMusic-Agent: Efficient Conversational Music Recommendation via Knowledge Internalization and Agentic Boundary Learning

TL;DR

WeMusic-Agent is a novel framework that combines knowledge internalization and agentic boundary learning to improve conversational music recommendation, enabling efficient use of internal knowledge and external tools for personalized, relevant, and diverse music suggestions.

Contribution

The paper introduces WeMusic-Agent, a training framework that integrates knowledge internalization and boundary learning, and constructs a new benchmark for evaluating conversational music recommendation.

Findings

WeMusic-Agent outperforms existing models on real-world data.

The framework effectively balances internal knowledge and external tool invocation.

The new benchmark enables comprehensive evaluation of personalized music recommendations.

Abstract

Personalized music recommendation in conversational scenarios usually requires a deep understanding of user preferences and nuanced musical context, yet existing methods often struggle with balancing specialized domain knowledge and flexible tool integration. This paper proposes WeMusic-Agent, a training framework for efficient LLM-based conversational music recommendation. By integrating the knowledge internalization and agentic boundary learning, the framework aims to teach the model to intelligently decide when to leverage internalized knowledge and when to call specialized tools (e.g., music retrieval APIs, music recommendation systems). Under this framework, we present WeMusic-Agent-M1, an agentic model that internalizes extensive musical knowledge via continued pretraining on 50B music-related corpus while acquiring the ability to invoke external tools when necessary. Additionally, considering the lack of open-source benchmarks for conversational music recommendation, we also construct a benchmark for personalized music recommendations derived from real-world data in WeChat Listen. This benchmark enables comprehensive evaluation across multiple dimensions, including relevance, personalization, and diversity of the recommendations. Experiments on real-world data demonstrate that WeMusic-Agent achieves significant improvements over existing models.