Speed sweep

pg-01a011ee-c97b-78c2-8e3e-f83afc0954a6-sweep

pg-01a011ee-c97b-78c2-8e3e-f83afc0954a6-sweep

Record

Recipe
pg-unsloth-nvidia-nemotron-3-super-120b-a12b-gguf-ud-q3-k-m-apple-m3-ultra-96gb-80c-llama-cpp-f6e7687db1
Measured
2026-08-17T22:54:16.951Z
Points
4

Measured speed

ConcurrencyContextPrefillDecodeTTFT msStatus
18,25637.5observed
132,83236.5observed
1262,08028.2observed
132,76856,572.5observed

Remaining fields

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

id
pg-01a011ee-c97b-78c2-8e3e-f83afc0954a6-sweep
measured at
2026-08-17T22:54:16.951Z
recipe id
pg-unsloth-nvidia-nemotron-3-super-120b-a12b-gguf-ud-q3-k-m-apple-m3-ultra-96gb-80c-llama-cpp-f6e7687db1
schema version
local-ai-registry/v1

metrics

base memory bytes
63388762112
base memory context tokens
192
concurrency
1
decode32k context tokens
32,832
decode32k tps
36.481
decode8k context tokens
8,256
decode8k tps
37.527
decode max context tokens
262,080
decode max context tps
28.152
decode mode
non-mtp
inference engine version
unknown
latest point at
2026-08-17T22:54:16.951Z
max context tokens
262,080
max prompt tokens
262,016
memory8k bytes
64664174592
memory8k context tokens
8,256
memory max context bytes
70170624000
memory max context tokens
262,080
peak generation tps
36.481
peak memory bytes
70170624000
peak prompt tps
579.221
point count
224
ttft32k cached prompt tokens
28,668
ttft32k context tokens
32,768
ttft32k seconds
56.573

rows

concurrencycontext tokensdecode tok sdecode tok s per streamoutput tokenspeak vram gbprefill tok ssamples
18,25637.52737.527Unknown60.223Unknown1
132,83236.48136.481Unknown65.351Unknown1
1262,08028.15228.152Unknown65.351Unknown1
132,768UnknownUnknownUnknownUnknownUnknown1
Provenance & metadata (1)

source

paths
publication:pg-20260827T060320709Z, run:01a011ee-c97b-78c2-8e3e-f83afc0954a6
repository
local.ai