cfg-0102
citable URL: https://halobench.com/records/cfg-0102/ — this address never moves; the anchor /records/#cfg-0102 keeps resolving
| model | always42-universal.gguf · F16 · rev main |
| engine | ggml-org/llama.cpp 3653e6d6d · vulkan · host aihydra (igpu) |
| flags | -ngl 999 -fa 1 -b 2048 -ub 512 -ctk f16 -ctv f16 -t 16 |
| template | not recorded at test time |
| tree | upstream — stock |
Capability basis
unmeasured — performance numbers on this config stand on a guard alone, not an established capability
Runs on this config (4)
| run | date | kind | suite | metrics |
|---|---|---|---|---|
| run-0368 | 2026-08-17 | performance | llama-bench@1 | prefill_tps=12914.102 |
| run-0369 | 2026-08-17 | performance | llama-bench@1 | decode_tps=245.225 |
| run-0370 | 2026-08-17 | performance | llama-bench@1 | prefill_tps=6948.317 |
| run-0371 | 2026-08-17 | performance | llama-bench@1 | decode_tps=193.421 |
Cited by — computed at build time, never stored
model pages deep-thought-posttrain