mirror of
https://github.com/NousResearch/hermes-agent.git
synced 2026-07-21 16:18:55 +00:00
Replaces the dozen ad-hoc measure-*/profile-* scripts (each reinventing the CDP client — 4 different copies — plus its own arg parsing, stats, output path, and none with a baseline) with one framework under scripts/perf/: - lib/cdp.mjs one CDP client + target discovery + typing + CPU-profile wrapper + DOM selectors - lib/stats.mjs percentiles, histograms, CPU-profile self-time ranking - lib/baseline.mjs load/compare/update baseline + regression gate (new capability) - lib/launch.mjs attach, OR spawn a fully ISOLATED instance - scenarios/* one module per measurement, registered in scenarios/index.mjs - run.mjs / serve.mjs, baseline.json, README.md Isolation solves the long-standing measurement blocker: a running `hgui` held the Electron single-instance lock, so a second instance quit. `--spawn` / `perf:serve` launch with their own --user-data-dir (separate lock scope), their own HERMES_HOME (separate backend/sessions, config seeded from ~/.hermes so it reaches a chat view without onboarding), and their own --remote-debugging-port. Synthetic scenarios drive $messages via window.__PERF_DRIVE__, so no LLM credits. Scenario -> sunset script mapping: stream <- measure-synthetic-stream, profile-synth-stream, profile-long-stream stream --real <- measure-real-stream, profile-real-stream keystroke <- measure-latency, profile-typing, leak-typing transcript <- (new: long-transcript mount cost) submit <- measure-submit, measure-jump session-switch <- profile-session-switch profile-switch <- measure-profile-switch CPU profiling is now a cross-cutting --cpuprofile flag, not 5 separate scripts. CI-tier scenarios (stream, keystroke, transcript) need no backend/credits and are gated against baseline.json (seed values; re-capture with --update-baseline on a reference device). Backend-tier scenarios are report-only. perf-probe.tsx gains loadTranscript() for the transcript scenario. No core files touched; isolation is via CLI args, not env-gated app changes. Verified: node --check all modules, tsc, eslint, and a unit smoke of the stats + regression-gate logic. The end-to-end GUI run (which opens a window) is left to run interactively via `npm run perf -- --spawn`.
87 lines
3.4 KiB
JavaScript
87 lines
3.4 KiB
JavaScript
// Submit (Enter) latency + scroll stability. Subsumes measure-submit and
|
|
// measure-jump. Backend tier: fires a REAL prompt, so run it on a throwaway
|
|
// session with a live backend. Report-only (no committed baseline — real
|
|
// round-trips are too environment-dependent to gate).
|
|
|
|
import { SELECTORS, sleep, typeIntoComposer } from '../lib/cdp.mjs'
|
|
import { summarize } from '../lib/stats.mjs'
|
|
|
|
const MEASURE = `
|
|
new Promise((resolve) => {
|
|
const composer = document.querySelector(${JSON.stringify(SELECTORS.composer)})
|
|
const thread = document.querySelector(${JSON.stringify(SELECTORS.threadContent)}) ||
|
|
document.querySelector(${JSON.stringify(SELECTORS.threadViewport)})
|
|
const viewport = document.querySelector(${JSON.stringify(SELECTORS.threadViewport)})
|
|
const startCount = thread ? thread.querySelectorAll(${JSON.stringify(SELECTORS.turnPair)}).length : 0
|
|
const startScroll = viewport ? viewport.scrollTop : 0
|
|
const m = { start: performance.now(), maxJumpPx: 0 }
|
|
let done = false
|
|
|
|
const finish = (reason) => {
|
|
if (done) return
|
|
done = true
|
|
clearTimeout(timer); composerObs.disconnect(); threadObs && threadObs.disconnect()
|
|
m.reason = reason
|
|
resolve(m)
|
|
}
|
|
|
|
const composerObs = new MutationObserver(() => {
|
|
if (!m.composerClearedMs && composer && composer.innerText.length === 0) {
|
|
m.composerClearedMs = performance.now() - m.start
|
|
}
|
|
})
|
|
composer && composerObs.observe(composer, { childList: true, subtree: true, characterData: true })
|
|
|
|
let threadObs = null
|
|
if (thread) {
|
|
threadObs = new MutationObserver(() => {
|
|
if (viewport) m.maxJumpPx = Math.max(m.maxJumpPx, Math.abs(viewport.scrollTop - startScroll))
|
|
const c = thread.querySelectorAll(${JSON.stringify(SELECTORS.turnPair)}).length
|
|
if (!m.userMsgRenderedMs && c > startCount) {
|
|
m.userMsgRenderedMs = performance.now() - m.start
|
|
requestAnimationFrame(() => { m.userMsgPaintMs = performance.now() - m.start; finish('paint') })
|
|
}
|
|
})
|
|
threadObs.observe(thread, { childList: true, subtree: true })
|
|
}
|
|
|
|
const timer = setTimeout(() => finish('timeout'), 5000)
|
|
composer && composer.dispatchEvent(new KeyboardEvent('keydown', { key: 'Enter', code: 'Enter', bubbles: true, cancelable: true }))
|
|
})
|
|
`
|
|
|
|
export default {
|
|
name: 'submit',
|
|
tier: 'backend',
|
|
description: 'Enter → composer cleared → user message painted, plus scroll jump.',
|
|
async run(cdp, opts = {}) {
|
|
const rounds = Number(opts.rounds ?? 3)
|
|
await cdp.send('Runtime.enable')
|
|
|
|
const clears = []
|
|
const paints = []
|
|
const jumps = []
|
|
|
|
for (let i = 0; i < rounds; i++) {
|
|
await typeIntoComposer(cdp, `perf submit round ${i} ${'x'.repeat(30)}`, { cps: 60 })
|
|
await sleep(250)
|
|
const m = await cdp.eval(MEASURE)
|
|
|
|
if (typeof m.composerClearedMs === 'number') clears.push(m.composerClearedMs)
|
|
if (typeof m.userMsgPaintMs === 'number') paints.push(m.userMsgPaintMs)
|
|
jumps.push(m.maxJumpPx ?? 0)
|
|
|
|
// Let the turn finish before the next round so they don't pile up.
|
|
await sleep(4000)
|
|
}
|
|
|
|
return {
|
|
metrics: {
|
|
submit_clear_p95_ms: summarize(clears).p95,
|
|
submit_paint_p95_ms: summarize(paints).p95,
|
|
submit_scroll_jump_max_px: Math.max(0, ...jumps)
|
|
},
|
|
detail: { rounds, clears: summarize(clears), paints: summarize(paints) }
|
|
}
|
|
}
|
|
}
|