Home › Evidence › Records › cfg-0144
⚠ This notebook has stopped taking notes: newest record is 23 days old (eng-0280), against an expected cadence of 14 days.

cfg-0144

backfilled
citable URL: https://halobench.com/records/cfg-0144/ — this address never moves; the anchor /records/#cfg-0144 keeps resolving
modelQwen3.6-35B-A3B-MTP-UD-Q4_K_M.gguf · UD-Q4_K_M · rev 0b21525e972670ed59e1812e170b27c26355381f0656ecc4e25617ece7dac58b
engineggml-org/llama.cpp 7077abbe14c510cb829c93a1328c2815b5805ebd · ROCm0/gfx1151 · host aihydra (igpu)
flagsHO-009-AB PLAIN (reasoning-OFF) capability fingerprint: same model/runtime/ backend as cfg-0132 (f16/f16 KV, -ngl 999 -fa on -c 32768 --load-mode none --parallel 1 --jinja -b 2048 -ub 512 -t 16, no spec/MTP flags) PLUS the tau2-protocol agent settings and -rea off. Reasoning disabled on the tau2 path to remove the llm_utils.generate() reasoning_content-drop confound (DIAG verdict, HO-009-DIAG t_fa64ef03). temp 0, seed 42, max_tokens 4096.
templatenot recorded at test time
treeupstream — stock
aged evidence — reconstructed from the archive. The -rea off delta is the deliberate DIAG-recommended fix, not a deviation. This cfg is scoped to the plain-vs-n2 capability comparison on build 7077abb only; no production/quality/throughput equivalence claim.

Capability basis

measured on this config

Runs on this config (2)

rundatekindsuitemetrics
run-05682026-08-22guardguard-c32768-depth8000tasks_passed=4 · tasks_total=4
run-05692026-08-22capabilitytau2-bench-airline@668d3bctasks_passed=6 · tasks_total=26 · mean_score=0.2308