Speed sweep

pg-01a02e18-a23a-7978-aeac-30afe0a8b365-sweep

pg-01a02e18-a23a-7978-aeac-30afe0a8b365-sweep

JSON API Back to speed sweep search

Record

Recipe: pg-qwen3-6-35b-a3b-q4kxl-unknown-apple-m4-max-36gb-llama-cpp-0fa7e74f87
Measured: 2026-08-23T10:09:21.452Z
Points: 4

Measured speed

Concurrency	Context	Prefill	Decode	TTFT ms	Status
1	8,256	—	62.6	—	observed
1	32,832	—	55.1	—	observed
1	212,928	—	16.2	—	observed
1	32,768	—	—	43,906.7	observed

Remaining fields

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

id: pg-01a02e18-a23a-7978-aeac-30afe0a8b365-sweep
measured at: 2026-08-23T10:09:21.452Z
recipe id: pg-qwen3-6-35b-a3b-q4kxl-unknown-apple-m4-max-36gb-llama-cpp-0fa7e74f87
schema version: local-ai-registry/v1

metrics

base memory bytes: 27063304192
base memory context tokens: 192
concurrency: 1
decode32k context tokens: 32,832
decode32k tps: 55.113
decode8k context tokens: 8,256
decode8k tps: 62.599
decode max context tokens: 212,928
decode max context tps: 16.182
decode mode: non-mtp
inference engine version: unknown
latest point at: 2026-08-23T10:09:21.452Z
max context tokens: 212,928
max prompt tokens: 212,864
memory8k bytes: 27575369728
memory8k context tokens: 8,256
memory max context bytes: 30198546432
memory max context tokens: 212,928
peak generation tps: 55.113
peak memory bytes: 30198546432
peak prompt tps: 746.31
point count: 182
ttft32k cached prompt tokens: 28,668
ttft32k context tokens: 32,768
ttft32k seconds: 43.907

rows

concurrency	context tokens	decode tok s	decode tok s per stream	output tokens	peak vram gb	prefill tok s	samples
1	8,256	62.599	62.599	Unknown	25.682	Unknown	1
1	32,832	55.113	55.113	Unknown	28.125	Unknown	1
1	212,928	16.182	16.182	Unknown	28.125	Unknown	1
1	32,768	Unknown	Unknown	Unknown	Unknown	Unknown	1

Provenance & metadata (1)

source

paths: publication:pg-20260827T060320709Z, run:01a02e18-a23a-7978-aeac-30afe0a8b365
repository: local.ai ↗