Speed sweep

pg-019f3295-e938-783d-9362-13a656efec9f-sweep

pg-019f3295-e938-783d-9362-13a656efec9f-sweep

JSON API Back to speed sweep search

Record

Recipe: pg-mlx-community-gemma-4-31b-it-bf16-bf16-apple-m3-ultra-96gb-60c-mlx-0cede0a37a
Measured: 2026-07-05T14:01:53.204Z
Points: 4

Measured speed

Concurrency	Context	Prefill	Decode	TTFT ms	Status
1	8,256	—	10.6	—	observed
1	32,832	—	9.8	—	observed
1	61,504	—	9	—	observed
1	32,768	—	—	166,733.1	observed

Remaining fields

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

id: pg-019f3295-e938-783d-9362-13a656efec9f-sweep
measured at: 2026-07-05T14:01:53.204Z
recipe id: pg-mlx-community-gemma-4-31b-it-bf16-bf16-apple-m3-ultra-96gb-60c-mlx-0cede0a37a
schema version: local-ai-registry/v1

metrics

base memory bytes: 61874156328
base memory context tokens: 192
concurrency: 1
decode32k context tokens: 32,832
decode32k tps: 9.819
decode8k context tokens: 8,256
decode8k tps: 10.621
decode max context tokens: 61,504
decode max context tps: 8.993
decode mode: non-mtp
inference engine version: 0.31.3
latest point at: 2026-07-05T14:01:53.204Z
max context tokens: 61,504
max prompt tokens: 61,440
memory8k bytes: 67444059382
memory8k context tokens: 8,256
memory max context bytes: 82316881138
memory max context tokens: 61,504
peak generation tps: 9.819
peak memory bytes: 82316881138
peak prompt tps: 196.53
point count: 55
ttft32k cached prompt tokens: 28,736
ttft32k context tokens: 32,768
ttft32k seconds: 166.733

rows

concurrency	context tokens	decode tok s	decode tok s per stream	output tokens	peak vram gb	prefill tok s	samples
1	8,256	10.621	10.621	Unknown	62.812	Unknown	1
1	32,832	9.819	9.819	Unknown	76.664	Unknown	1
1	61,504	8.993	8.993	Unknown	76.664	Unknown	1
1	32,768	Unknown	Unknown	Unknown	Unknown	Unknown	1

Provenance & metadata (1)

source

paths: publication:pg-20260827T060320709Z, run:019f3295-e938-783d-9362-13a656efec9f
repository: local.ai ↗