Speed sweep

pg-01a02992-e986-78c0-864c-2be48d47db59-sweep

pg-01a02992-e986-78c0-864c-2be48d47db59-sweep

JSON API Back to speed sweep search

Record

Recipe: pg-qwen3-6-35b-a3b-ud-iq3xxs-unknown-apple-m4-pro-48gb-llama-cpp-8a36778063
Measured: 2026-08-22T13:04:49.018Z
Points: 4

Measured speed

Concurrency	Context	Prefill	Decode	TTFT ms	Status
1	8,256	—	50.4	—	observed
1	32,832	—	39.5	—	observed
1	262,080	—	7.4	—	observed
1	32,768	—	—	87,044.8	observed

Remaining fields

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

id: pg-01a02992-e986-78c0-864c-2be48d47db59-sweep
measured at: 2026-08-22T13:04:49.018Z
recipe id: pg-qwen3-6-35b-a3b-ud-iq3xxs-unknown-apple-m4-pro-48gb-llama-cpp-8a36778063
schema version: local-ai-registry/v1

metrics

base memory bytes: 18925060096
base memory context tokens: 192
concurrency: 1
decode32k context tokens: 32,832
decode32k tps: 39.519
decode8k context tokens: 8,256
decode8k tps: 50.367
decode max context tokens: 262,080
decode max context tps: 7.401
decode mode: non-mtp
inference engine version: unknown
latest point at: 2026-08-22T13:04:49.018Z
max context tokens: 262,080
max prompt tokens: 262,016
memory8k bytes: 19436994560
memory8k context tokens: 8,256
memory max context bytes: 22302212096
memory max context tokens: 262,080
peak generation tps: 39.519
peak memory bytes: 22302212096
peak prompt tps: 376.45
point count: 224
ttft32k cached prompt tokens: 28,668
ttft32k context tokens: 32,768
ttft32k seconds: 87.045

rows

concurrency	context tokens	decode tok s	decode tok s per stream	output tokens	peak vram gb	prefill tok s	samples
1	8,256	50.367	50.367	Unknown	18.102	Unknown	1
1	32,832	39.519	39.519	Unknown	20.771	Unknown	1
1	262,080	7.401	7.401	Unknown	20.771	Unknown	1
1	32,768	Unknown	Unknown	Unknown	Unknown	Unknown	1

Provenance & metadata (1)

source

paths: publication:pg-20260827T060320709Z, run:01a02992-e986-78c0-864c-2be48d47db59
repository: local.ai ↗