Experience and Lessons Learned from Serving Multi-Stage Qwen3-Omni in vLLM-Omni
vLLM development signal - relevant to the current vLLM serving pin.
Independent publication // local-first AI // field reports from owned systems
Nodehome
A running publication about local models, private inference, self-hosted agents, weird hardware, research sweeps, and the builders wiring their own AI stack together.
AI is getting physical again. It shows up in terminals, racks, side projects, and ugly little workflows people actually control.
Latest
Daily Sweep - Jul 03, 2026adding MuonEq and Muon+, both seem to be very slight improvements at ...
Karpathy signal - watch for workflow shifts, repo movement, or vocabulary changes.
karpathy pushed nanochat
Karpathy signal - watch for workflow shifts, repo movement, or vocabulary changes.
llm-coding-agent 0.1a0
Simon Willison signal - surfaces practical patterns before they spread.
simonw released 0.1a0 at simonw/llm-coding-agent
Simon Willison tool release - usually practical and worth evaluating.
Using DSPy to evaluate and improve Datasette Agent's SQL system prompts
Simon Willison signal - surfaces practical patterns before they spread.
simonw pushed llm-coding-agent
Simon Willison signal - surfaces practical patterns before they spread.
simonw pushed research
Simon Willison signal - surfaces practical patterns before they spread.
Field Reports
builds, experiments, notesThree RTX 3090s, One 32B Model: A Pipeline-Parallel Canary
A current field note on why the 3x3090 serving path moved through pipeline parallelism, not tensor parallelism, for the tested 32B AWQ model.
Gemma 4 12B And The Sensory Agent Lane
A public-safe read on Gemma 4 12B as a local sensory preprocessor: useful for seeing, hearing, and structuring observations without turning into an action system.
Hardware
machines, thermals, economicsPower Caps On Three RTX 3090s: Bursts Versus Sustained Load
A measured note on 300W bursty inference, lower caps for sustained runs, and why power-cap sweet spots are workload-specific.
Parallel Agent Serving Is A Hardware Shape Now
A field-report read on 14x RTX 3090 agent serving, EXL3, FP8 KV cache, Aphrodite, and why concurrency is becoming the local hardware metric.