Speed sweep

pg-019f2c34-cfbe-75fd-8fce-bc6db05630d8-sweep

pg-019f2c34-cfbe-75fd-8fce-bc6db05630d8-sweep

Record

Recipe
pg-mlx-community-nvidia-nemotron-3-nano-30b-a3b-mlx-8bit-8bit-apple-m3-ultra-96gb-60c-mlx-402c5513ab
Measured
2026-07-04T08:18:06.394Z
Points
4

Measured speed

ConcurrencyContextPrefillDecodeTTFT msStatus
18,25691.9observed
132,83283.2observed
1258,11146observed
132,76822,549.1observed

Remaining fields

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

id
pg-019f2c34-cfbe-75fd-8fce-bc6db05630d8-sweep
measured at
2026-07-04T08:18:06.394Z
recipe id
pg-mlx-community-nvidia-nemotron-3-nano-30b-a3b-mlx-8bit-8bit-apple-m3-ultra-96gb-60c-mlx-402c5513ab
schema version
local-ai-registry/v1

metrics

base memory bytes
34158691592
base memory context tokens
192
concurrency
1
decode32k context tokens
32,832
decode32k tps
83.206
decode8k context tokens
8,256
decode8k tps
91.918
decode max context tokens
258,111
decode max context tps
46.02
decode mode
non-mtp
inference engine version
0.31.3
latest point at
2026-07-04T08:18:06.394Z
max context tokens
258,111
max prompt tokens
258,047
memory8k bytes
37118961530
memory8k context tokens
8,256
memory max context bytes
40347000710
memory max context tokens
258,111
peak generation tps
83.206
peak memory bytes
40347000710
peak prompt tps
1,453.186
point count
223
ttft32k cached prompt tokens
28,736
ttft32k context tokens
32,768
ttft32k seconds
22.549

rows

concurrencycontext tokensdecode tok sdecode tok s per streamoutput tokenspeak vram gbprefill tok ssamples
18,25691.91891.918Unknown34.57Unknown1
132,83283.20683.206Unknown37.576Unknown1
1258,11146.0246.02Unknown37.576Unknown1
132,768UnknownUnknownUnknownUnknownUnknown1
Provenance & metadata (1)

source

paths
publication:pg-20260827T060320709Z, run:019f2c34-cfbe-75fd-8fce-bc6db05630d8
repository
local.ai