Learning Task Embeddings for Teamwork Adaptation in Multi-Agent Reinforcement Learning