Speed sweep

pg-01a03076-f8b1-76c6-93c8-d5e711803da6-sweep

pg-01a03076-f8b1-76c6-93c8-d5e711803da6-sweep

JSON API Back to speed sweep search

Record

Recipe: pg-hcompany-holo-3-1-35b-a3b-gguf-q4-k-m-gguf-unknown-apple-m5-pro-64gb-llama-cpp-47e5f015b5
Measured: 2026-08-23T21:11:38.398Z
Points: 4

Measured speed

Concurrency	Context	Prefill	Decode	TTFT ms	Status
1	8,256	—	71	—	observed
1	32,832	—	56.7	—	observed
1	262,080	—	19.5	—	observed
1	32,768	—	—	48,589.6	observed

Remaining fields

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

id: pg-01a03076-f8b1-76c6-93c8-d5e711803da6-sweep
measured at: 2026-08-23T21:11:38.398Z
recipe id: pg-hcompany-holo-3-1-35b-a3b-gguf-q4-k-m-gguf-unknown-apple-m5-pro-64gb-llama-cpp-47e5f015b5
schema version: local-ai-registry/v1

metrics

base memory bytes: 26992689152
base memory context tokens: 192
concurrency: 1
decode32k context tokens: 32,832
decode32k tps: 56.681
decode8k context tokens: 8,256
decode8k tps: 71.007
decode max context tokens: 262,080
decode max context tps: 19.541
decode mode: non-mtp
inference engine version: unknown
latest point at: 2026-08-23T21:11:38.398Z
max context tokens: 262,080
max prompt tokens: 262,016
memory8k bytes: 27505770496
memory8k context tokens: 8,256
memory max context bytes: 30387863552
memory max context tokens: 262,080
peak generation tps: 56.681
peak memory bytes: 30387863552
peak prompt tps: 674.382
point count: 224
ttft32k cached prompt tokens: 28,668
ttft32k context tokens: 32,768
ttft32k seconds: 48.59

rows

concurrency	context tokens	decode tok s	decode tok s per stream	output tokens	peak vram gb	prefill tok s	samples
1	8,256	71.007	71.007	Unknown	25.617	Unknown	1
1	32,832	56.681	56.681	Unknown	28.301	Unknown	1
1	262,080	19.541	19.541	Unknown	28.301	Unknown	1
1	32,768	Unknown	Unknown	Unknown	Unknown	Unknown	1

Provenance & metadata (1)

source

paths: publication:pg-20260827T060320709Z, run:01a03076-f8b1-76c6-93c8-d5e711803da6
repository: local.ai ↗