Speed sweep

pg-01a013a0-0479-7a3a-bbf9-f895235f4e1d-sweep

pg-01a013a0-0479-7a3a-bbf9-f895235f4e1d-sweep

JSON API Back to speed sweep search

Record

Recipe: pg-mlx-community-nvidia-nemotron-3-nano-30b-a3b-nvfp4-nvfp4-apple-m4-pro-48gb-mlx-269a632afc
Measured: 2026-08-18T06:47:29.134Z
Points: 4

Measured speed

Concurrency	Context	Prefill	Decode	TTFT ms	Status
1	8,256	—	58.2	—	observed
1	32,832	—	51.5	—	observed
1	262,079	—	25	—	observed
1	32,768	—	—	59,753.3	observed

Remaining fields

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

id: pg-01a013a0-0479-7a3a-bbf9-f895235f4e1d-sweep
measured at: 2026-08-18T06:47:29.134Z
recipe id: pg-mlx-community-nvidia-nemotron-3-nano-30b-a3b-nvfp4-nvfp4-apple-m4-pro-48gb-mlx-269a632afc
schema version: local-ai-registry/v1

metrics

base memory bytes: 19934518702
base memory context tokens: 192
concurrency: 1
decode32k context tokens: 32,832
decode32k tps: 51.456
decode8k context tokens: 8,256
decode8k tps: 58.235
decode max context tokens: 262,079
decode max context tps: 25.019
decode mode: non-mtp
inference engine version: 0.31.3
latest point at: 2026-08-18T06:47:29.134Z
max context tokens: 262,079
max prompt tokens: 262,015
memory8k bytes: 22882311290
memory8k context tokens: 8,256
memory max context bytes: 26131420470
memory max context tokens: 262,079
peak generation tps: 51.456
peak memory bytes: 26131420470
peak prompt tps: 548.388
point count: 224
ttft32k cached prompt tokens: 28,736
ttft32k context tokens: 32,768
ttft32k seconds: 59.753

rows

concurrency	context tokens	decode tok s	decode tok s per stream	output tokens	peak vram gb	prefill tok s	samples
1	8,256	58.235	58.235	Unknown	21.311	Unknown	1
1	32,832	51.456	51.456	Unknown	24.337	Unknown	1
1	262,079	25.019	25.019	Unknown	24.337	Unknown	1
1	32,768	Unknown	Unknown	Unknown	Unknown	Unknown	1

Provenance & metadata (1)

source

paths: publication:pg-20260827T060320709Z, run:01a013a0-0479-7a3a-bbf9-f895235f4e1d
repository: local.ai ↗