Speed sweep

pg-01a03076-eefb-7bab-98d9-40a058b3f772-sweep

pg-01a03076-eefb-7bab-98d9-40a058b3f772-sweep

JSON API Back to speed sweep search

Record

Recipe: pg-lfm2-5-8b-a1b-q8-unknown-apple-m4-pro-48gb-llama-cpp-3bab535682
Measured: 2026-08-23T21:11:35.911Z
Points: 4

Measured speed

Concurrency	Context	Prefill	Decode	TTFT ms	Status
1	8,256	—	107.2	—	observed
1	32,832	—	94.4	—	observed
1	127,936	—	63.7	—	observed
1	32,768	—	—	29,160.5	observed

Remaining fields

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

id: pg-01a03076-eefb-7bab-98d9-40a058b3f772-sweep
measured at: 2026-08-23T21:11:35.911Z
recipe id: pg-lfm2-5-8b-a1b-q8-unknown-apple-m4-pro-48gb-llama-cpp-3bab535682
schema version: local-ai-registry/v1

metrics

base memory bytes: 10724720640
base memory context tokens: 192
concurrency: 1
decode32k context tokens: 32,832
decode32k tps: 94.433
decode8k context tokens: 8,256
decode8k tps: 107.23
decode max context tokens: 127,936
decode max context tps: 63.676
decode mode: non-mtp
inference engine version: unknown
latest point at: 2026-08-23T21:11:35.911Z
max context tokens: 127,936
max prompt tokens: 127,872
memory8k bytes: 10775183360
memory8k context tokens: 8,256
memory max context bytes: 11291656192
memory max context tokens: 127,936
peak generation tps: 94.433
peak memory bytes: 11291656192
peak prompt tps: 1,123.711
point count: 112
ttft32k cached prompt tokens: 28,668
ttft32k context tokens: 32,768
ttft32k seconds: 29.161

rows

concurrency	context tokens	decode tok s	decode tok s per stream	output tokens	peak vram gb	prefill tok s	samples
1	8,256	107.23	107.23	Unknown	10.035	Unknown	1
1	32,832	94.433	94.433	Unknown	10.516	Unknown	1
1	127,936	63.676	63.676	Unknown	10.516	Unknown	1
1	32,768	Unknown	Unknown	Unknown	Unknown	Unknown	1

Provenance & metadata (1)

source

paths: publication:pg-20260827T060320709Z, run:01a03076-eefb-7bab-98d9-40a058b3f772
repository: local.ai ↗