做智能体训练或微调的开发者终于有了大规模合成轨迹数据源——SynthTraces 用双模型对话自动生成 2000+ 条真实代码库交互轨迹，比手动标注高效太多，做 LLM 对齐或 Pi 优化的团队可以直接用。
Julien C 发布新项目 SynthTraces，这是一个最小化代码库，用于生成合成编程智能体会话轨迹。它让一个开放模型（通过 Hugging Face Inference Providers 服务）扮演编程智能体，另一个小模型（llama.cpp）扮演人类用户，在真实开源代码库（HuggingFace OSS 项目）上进行交互。最终生成了超过 2000 条 Pi 会话轨迹，可用于训练或微调 LLM，并针对 Pi 进行优化。所有数据已发布在 Hugging Face 上。
Today I'm launching a new project called SynthTraces 🔥 It is a minimal codebase to generate synthe...
Today I'm launching a new project called SynthTraces 🔥 It is a minimal codebase to generate synthetic coding agent session traces using Pi (from @badlogicgames ) I wanted a large number of coding-agent traces, so I built a tiny harness where two models talk to each other: - an open model (served via HF Inference Providers) plays the coding agent. It gets read + bash access to a real open source codebase (the huggingface OSS projects) - a small local model (llama.cpp) plays the human user, asking simple questions like "how do I run this?" or "how is CI set up?" The result is more than 2,000 Pi session traces which can be used to train or fine-tune LLMs, and optimize them for Pi 🤯 And ofc everything is published on @huggingface ✅ 💬 0 🔄 0 ❤️ 1 👀 52 ⚡