cfg-0064
backfilled
citable URL: https://halobench.com/records/cfg-0064/ — this address never moves; the anchor /records/#cfg-0064 keeps resolving
| model | Qwen3.8-27B-Q8_0.gguf · Q8_0 |
| 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 --load-mode none |
| template | not recorded at test time |
| tree | upstream — stock |
aged evidence — reconstructed from the archive. Ingested from llama-bench, which measures throughput and not footprint: the memory block, chat-template hash and model revision are absent by construction. Peak memory comes from a Tier 0 capability probe, not from this tool. Everything present here — build commit, backend, KV quant, batch sizes, flash attention — is machine-reported at launch. --load-mode none comes from the queue invocation itself (bench/queue-qwen38-screen.sh passes -lm none on every rep), added per clm-0053's rule that benchmark flags must match the serving standard on unified memory.
Capability basis
unmeasured — performance numbers on this config stand on a guard alone, not an established capability
Runs on this config (5)
| run | date | kind | suite | metrics |
|---|---|---|---|---|
| run-0257 | 2026-08-14 | performance | llama-bench@1 | prefill_tps=249.625 |
| run-0258 | 2026-08-14 | performance | llama-bench@1 | decode_tps=7.86291 |
| run-0259 | 2026-08-14 | performance | llama-bench@1 | prefill_tps=87.8246 |
| run-0260 | 2026-08-14 | performance | llama-bench@1 | decode_tps=7.27308 |
| run-0265 | 2026-08-14 | performance | llama-bench@1 | device_lost=true |
Cited by — computed at build time, never stored
model pages qwen38-27b