Recipe

qwen3-8-27b-nvfp4-dgx-spark-gb10-128gb-sglang-tp1-qh8u9odj

qwen3-8-27b-nvfp4-dgx-spark-gb10-128gb-sglang-tp1-qh8u9odj

Observed LocalMaxxing result. It is evidence for compatibility, not an executable launch contract.

Record

Status
candidate
Source
localmaxxing
Engine
sglang
Engine version
0.0.0.dev1+g5f55db35e
Accelerators
1
Tensor parallel
1
Context tokens
131,072
Max concurrency
1
chat
unknown
reasoning
unknown
tools
unknown
vision
unknown

Hugging Face model card

Identity

https://huggingface.co/RadixArk/Qwen3.8-27B-NVFP4
Repository
RadixArk/Qwen3.8-27B-NVFP4
Status
known
Link type
Exact Hub repository

Public Hugging Face repository confirmed by the Hub API.

Candidate: useful compatibility or speed evidence. The registry does not offer Run until promotion requirements are met.

Observed configuration

sglang

Evidence only · candidate · reference

Candidate evidence — not a Run contract

Source
https://www.localmaxxing.com/en/runs/cmtackra701gkli01qh8u9odj

No tokenized launch fields. This is measured or documented compatibility, not a Docker launch.

Measured speed

ConcurrencyContextPrefillDecodeTTFT msStatusSweep
1131,07212.4180.3observedqwen3-8-27b-nvfp4-dgx-spark-gb10-128gb-sglang-tp1-qh8u9odj-sweep

Remaining fields

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

hardware count
1
hardware id
dgx-spark-gb10-128gb
id
qwen3-8-27b-nvfp4-dgx-spark-gb10-128gb-sglang-tp1-qh8u9odj
model instance id
radixark-qwen3-8-27b-nvfp4--nvfp4
recipe source
localmaxxing
schema version
local-ai-registry/v1
speed sweep ids
qwen3-8-27b-nvfp4-dgx-spark-gb10-128gb-sglang-tp1-qh8u9odj-sweep
status
candidate

capabilities

draft launch

accelerator backend
nvidia
arguments
python3, -m, sglang.launch_server, --model-path, RadixArk/Qwen3.8-27B-NVFP4, --tp, 1, --host, 0.0.0.0, --port, 30000, --context-length, 131072
container port
30,000
host port
30,000
image
lmsysorg/sglang:dev-cu13@sha256:6cd4635214f279e0a43019f88e3120d407567640a58aa7dcc0085e3d91402cc4
ipc
host
kind
docker
shm size
16g
synthesized · generated at
2026-08-31T22:12:17Z
synthesized · image provenance
gemma-4-12b-it-nvfp4-rtxpro6000-sglang-tp1
synthesized · template
sglang-launch-server-v1

mounts

read onlytarget
No/root/.cache/huggingface

engine

name
sglang
version
0.0.0.dev1+g5f55db35e

serving

max concurrency
1
max context tokens
131,072
tensor parallel
1
Provenance & metadata (3)

facts

capabilities.chat · provenance · captured at
2026-08-27T06:04:13.773Z
capabilities.chat · reason capability-not-verified
capabilities.chat · state unknown
capabilities.reasoning · provenance · captured at
2026-08-27T06:04:13.773Z
capabilities.reasoning · reason capability-not-verified
capabilities.reasoning · state unknown
capabilities.tools · provenance · captured at
2026-08-27T06:04:13.773Z
capabilities.tools · reason capability-not-verified
capabilities.tools · state unknown
capabilities.vision · provenance · captured at
2026-08-27T06:04:13.773Z
capabilities.vision · reason capability-not-verified
capabilities.vision · state unknown
engine.graph mode · provenance · captured at
2026-08-27T06:04:13.773Z
engine.graph mode · reason runtime-detail-not-published
engine.graph mode · state unknown
serving.kv cache tokens · provenance · captured at
2026-08-27T06:04:13.773Z
serving.kv cache tokens · reason kv-cache-capacity-not-published
serving.kv cache tokens · state unknown

metadata

localmaxxing · hardware label
GB10 Grace Blackwell
localmaxxing · notes
Single-stream decode, batch 1, one concurrent request. n=9 generations (3 prompt kinds x 3 repeats), 512 output tokens each, streaming, temperature 0.60 top_p 0.95 top_k 20. Served context length 131072 (this is the CONFIGURED window, not the prompt depth; prompts here are short). mem-fraction-static 0.80, kv-cache-dtype fp8_e4m3, attention backend flashinfer, tp 1, max-running-requests 32. Idle box, one model loaded at a time, all six cells of the battery ran back to back in one session on one machine. Speculative decoding OFF (no draft model loaded). Per-workload decode: code 12.38, reasoning 12.37, prose 12.38 tok/s; the headline figure is the mean of all 9. Quantization detail: RadixArk NVFP4 build of the same base model. Served from a local copy; the hub revision was not recorded, so this run pins the checkpoint by its config.json sha256 instead (see config-hashes.txt in the artifact directory). Measured with a 3-prompt-kind by 3-repeat battery; per-run rows, the resolved server args and the launch script are retained locally as speed-rarkfp4-off.json, serverinfo-rarkfp4-off.json and battery.log. Happy to share the raw rows on request.
localmaxxing · revision unpinned
Yes
localmaxxing · run id
cmtackra701gkli01qh8u9odj

provenance

captured at
2026-08-27T06:04:13.773Z

sources

captured atkindurl
2026-08-27T06:04:13.773Znormalized-recipewww.localmaxxing.com/en/runs/cmtackra701gkli01qh8u9odj