- Add the [Tool-Integrated Reasoning (TIR) agent](https://github.com/QwenLM/Qwen-Agent/blob/main/examples/tir_math.py) for Qwen2.5-Math.
- Add support for [caching the LLM's responses](https://github.com/QwenLM/Qwen-Agent/blob/main/qwen_agent/llm/base.py#L65). LLM caching can be enabled by providing `cache_dir` in the LLM's config, i.e.,`llm = get_chat_model(cfg={..., "cache_dir": "/tmp/directory_to_save_llm_cache"})`.
- Add [self-play agents](https://github.com/QwenLM/Qwen-Agent/blob/main/qwen_agent/agents/dialogue_simulator.py) to simulate a dialogue between a user and an AI agent, mainly for data synthesis and testing purposes.