Early access · runs on your machine

Ask your GPU clusteranything.

DeepLM Ops is a local, read-only assistant for GPU and Slurm questions. See what's free, why a job is waiting, and what will fit, in plain English. Run it with Docker in minutes.

Try itOpen-source tools

Install on Windows · macOS · Ubuntu

127.0.0.1:8090
The DeepLM Ops chat interface: a workspace sidebar, a chat panel and a side panel showing the connected model and recent sessions.

Install

One command. Running in minutes.

DeepLM Ops runs on your own machine. It is read-only, so it can answer questions about your cluster but never change it. Paste this into a terminal on macOS or Ubuntu:

macOS · Ubuntu
curl -fsSL https://deeplm.ai/dlm-ops-dev/install.sh | bash
  • Installs Docker if you do not have it
  • Picks the best way to run the model: your Mac's GPU, an NVIDIA GPU, or the CPU
  • Downloads DeepLM Ops (the 2.55 GiB model comes down on the first start)
  • Starts it and opens it in your browser

Safe to re-run: the same command upgrades or restarts it. This is 0.2.0-rc.4, a release candidate for testing. Want to see what it does first? Read the script.

On Windows? The one-line installer supports macOS and Ubuntu. Follow the Windows guide.

Options: add -s -- --my-machine to use only this machine with no sample cluster, or -s -- --no-open to print the sign-in link instead of opening a browser.

The sign-in link it prints is a password and works until the container restarts. Don't paste it into chats, issues or screen shares.

Then ask it:

  • Why is my job 5102 still pending?
  • Why did 5098 fail?
  • Is anyone wasting a GPU on gpu03?
  • Will Llama 3.1 8B fit on my GPU for inference with 8k context?
Prefer to do it by hand?Windows guidemacOS guideUbuntu guideDocker commandsAll docs

Products

Built for people who run GPU clusters.

DeepLM Ops answers your questions. Insights and Baseline show you the numbers. All of them run on your own machines.

Release candidate

DeepLM Ops

A local, read-only assistant for GPU and Slurm questions. Ask why a job is pending, who is using a GPU, or whether a model fits. Runs on your own machine with Docker.

Read the docs
Free & open source

DeepLM Insights

Real-time Grafana dashboards and Prometheus metrics for HPC and Slurm GPU clusters. Track job performance, GPU utilization, power and checkpoint efficiency.

Insights docs
Free & open source

DeepLM Baseline

Benchmark your GPU cluster's real performance in one run: compute, interconnect and network scaling, checked against vendor specifications.

Baseline docs

More:Product overviewGPU cost calculatorAbout DeepLM