Speed sweep

qwen3-6-27b-ud-q8-k-xl-ryzen-ai-max-plus-395-128gb-llama-cpp-tp1-f4cq90a4-sweep

qwen3-6-27b-ud-q8-k-xl-ryzen-ai-max-plus-395-128gb-llama-cpp-tp1-f4cq90a4-sweep

Record

Recipe
qwen3-6-27b-ud-q8-k-xl-ryzen-ai-max-plus-395-128gb-llama-cpp-tp1-f4cq90a4
Measured
2026-04-26T05:57:47.419Z
Points
1

Measured speed

ConcurrencyContextPrefillDecodeTTFT msStatus
132,7686.12,548.9observed

Remaining fields

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

id
qwen3-6-27b-ud-q8-k-xl-ryzen-ai-max-plus-395-128gb-llama-cpp-tp1-f4cq90a4-sweep
measured at
2026-04-26T05:57:47.419Z
recipe id
qwen3-6-27b-ud-q8-k-xl-ryzen-ai-max-plus-395-128gb-llama-cpp-tp1-f4cq90a4
schema version
local-ai-registry/v1

metrics

concurrency
1
inference engine version
b8870
latest point at
2026-04-26T05:57:47.419Z
max context tokens
32,768
peak generation tps
6.07
point count
1

rows

concurrencycontext tokensdecode tok sdecode tok s per streamoutput tokenspeak vram gbprefill tok ssamples
132,7686.076.0751233Unknown1
Provenance & metadata (1)