Speed sweep

pg-019fd7b6-180b-75dc-a6ea-74cafdfeee55-sweep

pg-019fd7b6-180b-75dc-a6ea-74cafdfeee55-sweep

JSON API Back to speed sweep search

Record

Recipe: pg-unsloth-qwen3-6-27b-mtp-gguf-qwen3-6-27b-q4-k-m-gguf-q4-k-m-apple-m4-max-36gb-llama-cpp-f21a9cc5ac
Measured: 2026-08-06T15:34:22.986Z
Points: 4

Measured speed

Concurrency	Context	Prefill	Decode	TTFT ms	Status
1	8,256	—	17.9	—	observed
1	16,320	—	17.4	—	observed
1	16,320	—	17.4	—	observed
1	16,256	—	—	91,922.2	observed

Remaining fields

Identity, launch, related records, and measured speed are shown above. This is the rest of the normalized record.

id: pg-019fd7b6-180b-75dc-a6ea-74cafdfeee55-sweep
measured at: 2026-08-06T15:34:22.986Z
recipe id: pg-unsloth-qwen3-6-27b-mtp-gguf-qwen3-6-27b-q4-k-m-gguf-q4-k-m-apple-m4-max-36gb-llama-cpp-f21a9cc5ac
schema version: local-ai-registry/v1

metrics

base memory bytes: 18421972992
base memory context tokens: 192
concurrency: 1
decode32k context tokens: 16,320
decode32k tps: 17.361
decode8k context tokens: 8,256
decode8k tps: 17.95
decode max context tokens: 16,320
decode max context tps: 17.361
decode mode: non-mtp
inference engine version: unknown
latest point at: 2026-08-06T15:34:22.986Z
max context tokens: 16,320
max prompt tokens: 16,256
memory8k bytes: 19594379264
memory8k context tokens: 8,256
memory max context bytes: 20728692736
memory max context tokens: 16,320
peak generation tps: 17.95
peak memory bytes: 20728692736
peak prompt tps: 188.607
point count: 14
ttft32k cached prompt tokens: 12,284
ttft32k context tokens: 16,256
ttft32k seconds: 91.922

rows

concurrency	context tokens	decode tok s	decode tok s per stream	output tokens	peak vram gb	prefill tok s	samples
1	8,256	17.95	17.95	Unknown	18.249	Unknown	1
1	16,320	17.361	17.361	Unknown	19.305	Unknown	1
1	16,320	17.361	17.361	Unknown	19.305	Unknown	1
1	16,256	Unknown	Unknown	Unknown	Unknown	Unknown	1

Provenance & metadata (1)

source

paths: publication:pg-20260827T060320709Z, run:019fd7b6-180b-75dc-a6ea-74cafdfeee55
repository: local.ai ↗