Commit graph

7 commits

Author SHA1 Message Date
PeterChrz
9bf17dba6f
Make label part of band identity; correct ASUS GPU attribution
Labels now define distinct sessions — different power-cap configs
(asus_210w, sapphire_210w, asus_uncapped) render as separate bands
instead of merging into one. The label badge displays alongside
quant/engine/config badges rather than replacing them.

Scenario drilldown filters by label to prevent cross-band mixing.

Data fix: runs 22-32 (asus_210w) and 43-46 (asus_uncapped) reassigned
from Sapphire (gpu_id=11) to ASUS (gpu_id=12) — they were run on the
ASUS card but imported with the Sapphire GPU identity.
2026-10-04 23:18:21 -04:00
PeterChrz
f96fd1ad77
Exclude scenariobench from band total geo mean; label tok/s 2026-10-01 14:53:59 -04:00
PeterChrz
8c9e512fc2
Add geometric-mean band total (tok/s + tokens) under each band header 2026-09-30 15:47:01 -04:00
PeterChrz
f334538866
Surface thinking level for scenariobench bands; backfill Ollama Cloud runs to low 2026-09-30 10:30:55 -04:00
PeterChrz
06908e252d
v2.4: per-band scenariobench aggregation, band scenario detail page, nickname badges, orderBy fix 2026-09-20 22:49:23 -04:00
PeterChrz
9f9327bb27
v2.3: group runs by actual harness config, per-level depthsweep, scenario gen speed, named server configs 2026-09-19 23:45:04 -04:00
PeterChrz
2e205ef792
v2.1: session-grouped summary home page (GPU/model/quant/runner bands, per-suite pass/fail, avg + weighted tok/s), flat task table moved to /results, summary command + tables on schedule and after import 2026-09-18 20:12:30 -04:00