sigilant-runner · Phi-3.5-mini-instruct · L4 · vllm · 4 configs

Config | TPS | TPS p95 | TTFT | TTFT p95 | ITL | PPL | TPS% | TTFT% | PPL% | Score
FP16_BASELINE · ctx:16384 · kv:k8v8 · default  <- best | 58.4 | n/a | 143.1 | n/a | 17.19 | 3.03 | 100.0 | 100.0 | 100.0 | 100
FP16_BASELINE · ctx:32768 · kv:k8v8 · long | 58.4 | n/a | 144.1 | n/a | 17.19 | 3.03 | 100.0 | 99.3 | 100.0 | 100
FP16_BASELINE · ctx:16384 · kv:k16v16 · long | 58.4 | n/a | 146.7 | n/a | 17.19 | 3.03 | 100.0 | 97.5 | 100.0 | 100
FP16_BASELINE · ctx:8192 · kv:k16v16 · default | 58.4 | n/a | 149.0 | n/a | 17.18 | 3.03 | 100.0 | 96.0 | 100.0 | 99

Best config: FP16_BASELINE · ctx:16384 · kv:k8v8 · default
Auto baseline compare: score Δ=0.00 TPS Δ=0.00 TTFT Δ=0.0ms PPL Δ=0.00
Confidence: target=medium gap_before=0.00% var_before=n/a% replay=False(disabled) gap_after=0.00%

PPL is a quality proxy, not production validation.
Full production safety and long-context certification require Sigilant Optimizer.
