Speed sweep

pg-01a011ee-d3d7-7e2d-ac59-cc031f79b0e8-sweep

pg-01a011ee-d3d7-7e2d-ac59-cc031f79b0e8-sweep

JSON API Back to speed sweep search

Record

Recipe: pg-prism-ml-bonsai-27b-gguf-q1-0-apple-m3-ultra-96gb-60c-llama-cpp-a9a1aa53c1
Measured: 2026-08-17T22:54:19.603Z
Points: 4

Measured speed

Concurrency	Context	Prefill	Decode	TTFT ms	Status
1	8,256	—	47.9	—	observed
1	32,832	—	40.1	—	observed
1	262,080	—	10	—	observed
1	32,768	—	—	123,380.8	observed

Remaining fields

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

id: pg-01a011ee-d3d7-7e2d-ac59-cc031f79b0e8-sweep
measured at: 2026-08-17T22:54:19.603Z
recipe id: pg-prism-ml-bonsai-27b-gguf-q1-0-apple-m3-ultra-96gb-60c-llama-cpp-a9a1aa53c1
schema version: local-ai-registry/v1

metrics

base memory bytes: 21509324800
base memory context tokens: 192
concurrency: 1
decode32k context tokens: 32,832
decode32k tps: 40.063
decode8k context tokens: 8,256
decode8k tps: 47.949
decode max context tokens: 262,080
decode max context tps: 9.984
decode mode: non-mtp
inference engine version: unknown
latest point at: 2026-08-17T22:54:19.603Z
max context tokens: 262,080
max prompt tokens: 262,016
memory8k bytes: 22681944064
memory8k context tokens: 8,256
memory max context bytes: 27794472960
memory max context tokens: 262,080
peak generation tps: 40.063
peak memory bytes: 27794472960
peak prompt tps: 265.584
point count: 224
ttft32k cached prompt tokens: 28,668
ttft32k context tokens: 32,768
ttft32k seconds: 123.381

rows

concurrency	context tokens	decode tok s	decode tok s per stream	output tokens	peak vram gb	prefill tok s	samples
1	8,256	47.949	47.949	Unknown	21.124	Unknown	1
1	32,832	40.063	40.063	Unknown	25.886	Unknown	1
1	262,080	9.984	9.984	Unknown	25.886	Unknown	1
1	32,768	Unknown	Unknown	Unknown	Unknown	Unknown	1

Provenance & metadata (1)

source

paths: publication:pg-20260827T060320709Z, run:01a011ee-d3d7-7e2d-ac59-cc031f79b0e8
repository: local.ai ↗