Home › Evidence › Records › cfg-0023

cfg-0023

backfilled
citable URL: https://halobench.com/records/cfg-0023/ — this address never moves; the anchor /records/#cfg-0023 keeps resolving
modelQwen3.6-35B-A3B-UD-Q4_K_XL.gguf · UD-Q4_K_XL
engineggml-org/llama.cpp 3653e6d · rocm · host aihydra (igpu)
flags-ngl 999 -fa 1 -b 2048 -ub 512 -ctk f16 -ctv f16 -t 16
templatenot recorded at test time
treeupstream — 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.

Capability basis

unmeasured — performance numbers on this config stand on a guard alone, not an established capability

Runs on this config (6)

rundatekindsuitemetrics
run-00862026-08-08performancellama-bench@1prefill_tps=1073.910307
run-00872026-08-08performancellama-bench@1decode_tps=51.124388
run-00882026-08-08performancellama-bench@1prefill_tps=600.414968
run-00892026-08-08performancellama-bench@1decode_tps=42.452831
run-00902026-08-08performancellama-bench@1prefill_tps=257.519599
run-00912026-08-08performancellama-bench@1decode_tps=28.464561

Cited by — computed at build time, never stored

model pages qwen36-35b