cfg-0154
backfilled
citable URL: https://halobench.com/records/cfg-0154/ — this address never moves; the anchor /records/#cfg-0154 keeps resolving
| model | Qwen3.5-122B-A10B-UD-Q4_K_M-00001-of-00003.gguf · UD-Q4_K_M |
| engine | ggml-org/llama.cpp ce7689f · ROCm0/gfx1151 · host aihydra (igpu) |
| flags | HO-012 Cell-1 batch/ubatch sweep arm b512/ub256: llama-server build ce7689f (llama.cpp-kvfix), -ngl 999 -fa 1 -c 32768 --parallel 1 --load-mode none --jinja -rea off -ctk q8_0 -ctv q8_0 -b 512 -ub 256 -t 16 -n 4096, ROCm0 Radeon 8060S gfx1151 122880 MiB. Identical to cfg-0151 EXCEPT -b 512 -ub 256. From server-c1-b512-ub256.command.txt. |
| template | not recorded at test time |
| tree | upstream — stock |
aged evidence — reconstructed from the archive. From reviewer-admitted HO-012 evidence. batch/ubatch is a perf lever; throughput/matrix arm only, no capability claim. q8_0/q8_0 KV. Intra-suite comparison only.
Capability basis
unmeasured — performance numbers on this config stand on a guard alone, not an established capability
Runs on this config (1)
| run | date | kind | suite | metrics |
|---|---|---|---|---|
| run-0582 | 2026-08-23 | performance | ho012-122b-prod-matrix-r2.cell1 | decode_tps=21.8341 · ttft_ms=1682.65 |
Cited by — computed at build time, never stored
claims clm-0111