Speed sweep

pg-01a0139f-eb24-7317-9f60-50616ed2716f-sweep

pg-01a0139f-eb24-7317-9f60-50616ed2716f-sweep

JSON API Back to speed sweep search

Record

Recipe: pg-mlx-community-qwen3-6-35b-a3b-4bit-4bit-apple-m3-ultra-96gb-80c-mlx-5c105cf937
Measured: 2026-08-18T06:47:22.515Z
Points: 4

Measured speed

Concurrency	Context	Prefill	Decode	TTFT ms	Status
1	8,256	—	86.7	—	observed
1	32,832	—	74.9	—	observed
1	262,079	—	34.4	—	observed
1	32,768	—	—	19,906.8	observed

Remaining fields

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

id: pg-01a0139f-eb24-7317-9f60-50616ed2716f-sweep
measured at: 2026-08-18T06:47:22.515Z
recipe id: pg-mlx-community-qwen3-6-35b-a3b-4bit-4bit-apple-m3-ultra-96gb-80c-mlx-5c105cf937
schema version: local-ai-registry/v1

metrics

base memory bytes: 19870893778
base memory context tokens: 192
concurrency: 1
decode32k context tokens: 32,832
decode32k tps: 74.876
decode8k context tokens: 8,256
decode8k tps: 86.701
decode max context tokens: 262,079
decode max context tps: 34.354
decode mode: non-mtp
inference engine version: 0.31.3
latest point at: 2026-08-18T06:47:22.515Z
max context tokens: 262,079
max prompt tokens: 262,015
memory8k bytes: 21935995182
memory8k context tokens: 8,256
memory max context bytes: 48839785100
memory max context tokens: 262,079
peak generation tps: 74.876
peak memory bytes: 48839785100
peak prompt tps: 1,646.073
point count: 224
ttft32k cached prompt tokens: 28,736
ttft32k context tokens: 32,768
ttft32k seconds: 19.907

rows

concurrency	context tokens	decode tok s	decode tok s per stream	output tokens	peak vram gb	prefill tok s	samples
1	8,256	86.701	86.701	Unknown	20.429	Unknown	1
1	32,832	74.876	74.876	Unknown	45.486	Unknown	1
1	262,079	34.354	34.354	Unknown	45.486	Unknown	1
1	32,768	Unknown	Unknown	Unknown	Unknown	Unknown	1

Provenance & metadata (1)

source

paths: publication:pg-20260827T060320709Z, run:01a0139f-eb24-7317-9f60-50616ed2716f
repository: local.ai ↗