Speed sweep

pg-019f7583-1d95-7432-a42c-e46186b79083-sweep

pg-019f7583-1d95-7432-a42c-e46186b79083-sweep

JSON API Back to speed sweep search

Record

Recipe: pg-mlx-community-qwen3-6-27b-optiq-4bit-4bit-apple-m5-max-128gb-mlx-4ab9ff13b4
Measured: 2026-07-18T13:55:54.853Z
Points: 4

Measured speed

Concurrency	Context	Prefill	Decode	TTFT ms	Status
1	8,256	—	25.5	—	observed
1	32,832	—	23.6	—	observed
1	65,600	—	21	—	observed
1	32,768	—	—	58,425.2	observed

Remaining fields

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

id: pg-019f7583-1d95-7432-a42c-e46186b79083-sweep
measured at: 2026-07-18T13:55:54.853Z
recipe id: pg-mlx-community-qwen3-6-27b-optiq-4bit-4bit-apple-m5-max-128gb-mlx-4ab9ff13b4
schema version: local-ai-registry/v1

metrics

base memory bytes: 19457777831
base memory context tokens: 192
concurrency: 1
decode32k context tokens: 32,832
decode32k tps: 23.593
decode8k context tokens: 8,256
decode8k tps: 25.485
decode max context tokens: 65,600
decode max context tps: 20.95
decode mode: non-mtp
inference engine version: 0.31.3
latest point at: 2026-07-18T13:55:54.853Z
max context tokens: 65,600
max prompt tokens: 65,536
memory8k bytes: 23153321214
memory8k context tokens: 8,256
memory max context bytes: 36241057030
memory max context tokens: 65,600
peak generation tps: 23.593
peak memory bytes: 36241057030
peak prompt tps: 560.854
point count: 56
ttft32k cached prompt tokens: 28,736
ttft32k context tokens: 32,768
ttft32k seconds: 58.425

rows

concurrency	context tokens	decode tok s	decode tok s per stream	output tokens	peak vram gb	prefill tok s	samples
1	8,256	25.485	25.485	Unknown	21.563	Unknown	1
1	32,832	23.593	23.593	Unknown	33.752	Unknown	1
1	65,600	20.95	20.95	Unknown	33.752	Unknown	1
1	32,768	Unknown	Unknown	Unknown	Unknown	Unknown	1

Provenance & metadata (1)

source

paths: publication:pg-20260827T060320709Z, run:019f7583-1d95-7432-a42c-e46186b79083
repository: local.ai ↗