Hermes Agent vs ml-intern

Hermes Agent is NousResearch's open-source AI assistant built around a closed-loop learning architecture — the agent doesn't just execute tasks, it synthesizes new skills from complex interactions, self-improves those skills during use, and maintains a deepening model of the user across sessions. With 115,000+ GitHub stars, it has become one of the most-adopted autonomous agent projects in the open-source ecosystem. The system runs on 200+ models via OpenRouter, Nous Portal, NVIDIA NIM, and others, with tool-based provider switching that requires zero code changes. Users can interact via a terminal interface or through Telegram, Discord, Slack, WhatsApp, or Signal — all from a single gateway process. Built-in cron scheduling enables fully unattended workflows, and the agent can spawn isolated subagents for parallel workstreams. What sets Hermes apart from typical agent frameworks is the memory layer: it captures observations via five session hooks, stores them in SQLite with FTS5 search, and uses a Chroma vector database for semantic retrieval — cutting context costs by ~10x versus naive approaches. The result is an agent that genuinely accumulates expertise over time rather than starting from scratch each session.

ml-intern is Hugging Face's own open-source autonomous ML engineering agent. Given a task description, it reads relevant papers, writes training code, executes it in a sandboxed environment, evaluates the results, iterates, and ultimately uploads a trained model to the Hugging Face Hub — with no human in the loop beyond the initial prompt. Under the hood, the agent runs an agentic loop of up to 300 iterations, using Claude as its reasoning backbone alongside smolagents. It has integrated access to HF documentation search, paper retrieval, GitHub code search, and sandboxed Python execution. When the context window fills (at 170k tokens), it auto-compacts rather than failing, and full sessions are uploaded to HF for inspection and reproducibility. What's notable here isn't just the capability — it's the source. Hugging Face is essentially shipping a proof-of-concept that the job of "write the ML training script, run it, fix it until it works, upload the result" can now be delegated to an agent. With 688 stars and active development as of this week, ml-intern is HF eating its own dog food on autonomous AI engineering. The "doom loop detector" that flags repetitive tool-use patterns is a candid acknowledgment of how agentic loops fail in practice.