Speed sweep

pg-019f2c34-e743-7c8c-949c-8d11635e8eac-sweep

pg-019f2c34-e743-7c8c-949c-8d11635e8eac-sweep

JSON API Back to speed sweep search

Record

Recipe: pg-mlx-community-qwen3-5-9b-optiq-4bit-4bit-apple-m5-pro-64gb-mlx-88845590ed
Measured: 2026-07-04T08:18:12.415Z
Points: 4

Measured speed

Concurrency	Context	Prefill	Decode	TTFT ms	Status
1	8,256	—	40.2	—	observed
1	32,832	—	35.8	—	observed
1	258,111	—	17.6	—	observed
1	32,768	—	—	31,796.9	observed

Remaining fields

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

id: pg-019f2c34-e743-7c8c-949c-8d11635e8eac-sweep
measured at: 2026-07-04T08:18:12.415Z
recipe id: pg-mlx-community-qwen3-5-9b-optiq-4bit-4bit-apple-m5-pro-64gb-mlx-88845590ed
schema version: local-ai-registry/v1

metrics

base memory bytes: 7534259858
base memory context tokens: 192
concurrency: 1
decode32k context tokens: 32,832
decode32k tps: 35.791
decode8k context tokens: 8,256
decode8k tps: 40.168
decode max context tokens: 258,111
decode max context tps: 17.636
decode mode: non-mtp
inference engine version: unknown
latest point at: 2026-07-04T08:18:12.415Z
max context tokens: 258,111
max prompt tokens: 258,047
memory8k bytes: 9578171038
memory8k context tokens: 8,256
memory max context bytes: 42521971378
memory max context tokens: 258,111
peak generation tps: 35.791
peak memory bytes: 42521971378
peak prompt tps: 1,030.54
point count: 223
ttft32k cached prompt tokens: 28,736
ttft32k context tokens: 32,768
ttft32k seconds: 31.797

rows

concurrency	context tokens	decode tok s	decode tok s per stream	output tokens	peak vram gb	prefill tok s	samples
1	8,256	40.168	40.168	Unknown	8.92	Unknown	1
1	32,832	35.791	35.791	Unknown	39.602	Unknown	1
1	258,111	17.636	17.636	Unknown	39.602	Unknown	1
1	32,768	Unknown	Unknown	Unknown	Unknown	Unknown	1

Provenance & metadata (1)

source

paths: publication:pg-20260827T060320709Z, run:019f2c34-e743-7c8c-949c-8d11635e8eac
repository: local.ai ↗