Speed sweep

pg-01a026ab-6b48-78fd-abcb-a9a11e676cc2-sweep

pg-01a026ab-6b48-78fd-abcb-a9a11e676cc2-sweep

JSON API Back to speed sweep search

Record

Recipe: pg-gemma-4-12b-q4km-unknown-apple-m4-pro-48gb-llama-cpp-70f4dea6fe
Measured: 2026-08-21T23:32:43.448Z
Points: 4

Measured speed

Concurrency	Context	Prefill	Decode	TTFT ms	Status
1	8,256	—	26.8	—	observed
1	32,832	—	23.4	—	observed
1	262,080	—	6.8	—	observed
1	32,768	—	—	206,749.1	observed

Remaining fields

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

id: pg-01a026ab-6b48-78fd-abcb-a9a11e676cc2-sweep
measured at: 2026-08-21T23:32:43.448Z
recipe id: pg-gemma-4-12b-q4km-unknown-apple-m4-pro-48gb-llama-cpp-70f4dea6fe
schema version: local-ai-registry/v1

metrics

base memory bytes: 12324192256
base memory context tokens: 192
concurrency: 1
decode32k context tokens: 32,832
decode32k tps: 23.44
decode8k context tokens: 8,256
decode8k tps: 26.765
decode max context tokens: 262,080
decode max context tps: 6.761
decode mode: non-mtp
inference engine version: unknown
latest point at: 2026-08-21T23:32:43.448Z
max context tokens: 262,080
max prompt tokens: 262,016
memory8k bytes: 14575288320
memory8k context tokens: 8,256
memory max context bytes: 24408850432
memory max context tokens: 262,080
peak generation tps: 23.44
peak memory bytes: 24468455424
peak prompt tps: 158.492
point count: 224
ttft32k cached prompt tokens: 28,672
ttft32k context tokens: 32,768
ttft32k seconds: 206.749

rows

concurrency	context tokens	decode tok s	decode tok s per stream	output tokens	peak vram gb	prefill tok s	samples
1	8,256	26.765	26.765	Unknown	13.574	Unknown	1
1	32,832	23.44	23.44	Unknown	22.788	Unknown	1
1	262,080	6.761	6.761	Unknown	22.733	Unknown	1
1	32,768	Unknown	Unknown	Unknown	Unknown	Unknown	1

Provenance & metadata (1)

source

paths: publication:pg-20260827T060320709Z, run:01a026ab-6b48-78fd-abcb-a9a11e676cc2
repository: local.ai ↗