mit-han-lab/EvolvingInfra
Agent
Can I trust it?
Be careful with this one.
Warnings Check
Very new: it was created 7 days ago. Wait a little before installing it.
Stars from real accounts No data
We haven't checked its stars for fakes yet.
Active OK
Last change today.
Licence Careful
No licence: legally you can't reuse the code, unless the README says otherwise.
Quick checks from public data. They help, but they are no guarantee.
What it is
An LLM serving system that adapts online to agent workloads without offline profiling. It separates compute allocation from KV cache management, predicts prefill and decode load with a cost model covering sub-agents and context management, and runs an agent-aware distributed KV…
Who it's for
Infrastructure engineers and researchers serving LLM agent workloads at scale on GPU clusters
How to use it
Ask your AI to set it up
Paste this into Claude, ChatGPT or your coding assistant.
Help me install and set up mit-han-lab/EvolvingInfra in my project. Follow its README: https://github.com/mit-han-lab/EvolvingInfra. Explain each step and check it is safe before running anything.
Star curve
Alternatives
Similar projects named in its profile. Each link searches the directory.
Get alerts for repos like this
We'll email you when something in AI agents starts rising. Free, once a week, only what matches.
What arrives: one email on Mondays with the repos that match, never more. See last week's. Privacy
For maintainers
Add this badge to your README. It shows the repo's place on the radar, updates every hour and links back to this page.
[](https://repoalerts.com/repo/mit-han-lab/EvolvingInfra?ref=badge)Similar repos
- openlake-project/openlake
A Rust storage engine built on io_uring for GPU workloads. It offers KV cache offload for LLM inference across GPU hosts, plus an S3-compatible object store for checkpoints and training data. The READ
- Generated-Computers/Locals
A local inference server for Apple Silicon Macs that runs many AI agents at once, with tool calls constrained to each tool's schema and prompt caching across turns. It works with coding agents such as
- smhanov/agentloop
A Go library that implements the LLM tool-calling loop: it sends prompts to a model, executes requested tool calls, feeds results back, and repeats until a final answer. It adds budgets, hooks, skills
- mw00/project-maya
Runs GLM-5.3-Flash and DeepSeek-V4.1-Flash-1017 locally on one or more NVIDIA or AMD GPUs, with an engine, server, browser dashboard, quantization tools and OpenAI- and Anthropic-compatible API. Exper
- NineAbyss/OR-License
A benchmark of 31 operations research tasks for evaluating LLM agents. Agents query an interactive business-information simulator, write optimization programs in an isolated environment, and submit so
- FutureMLS-Lab/FlashStates
Inference serving for hybrid attention models (attention plus linear-attention or state-space layers) that reuses cached prefixes for coding-agent sessions. It saves recurrent checkpoints at any token