Recipe

qwen38-awq-int4-rtxpro4500-vllm-tp1

qwen38-awq-int4-rtxpro4500-vllm-tp1

Candidate derived from qwen38-awq-int4-rtx3090-vllm-tp1 for rtx-pro-4500-blackwell-32gb; the launch contract is copied, the evidence is not. Pending acceptance on the target card.

Record

Status
candidate
Source
0xsero
Engine
vllm
Engine version
0.27.1
Graph
full-and-piecewise
Accelerators
1
Tensor parallel
1
Context tokens
98,304
Max concurrency
4
chat
yes
reasoning
no
tools
no
vision
no

Hugging Face model card

Identity

https://huggingface.co/cyankiwi/Qwen3.8-27B-AWQ-INT4
Repository
cyankiwi/Qwen3.8-27B-AWQ-INT4
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

vllm

Evidence only · candidate · reference

Candidate evidence — not a Run contract

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

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
rtx-pro-4500-blackwell-32gb
id
qwen38-awq-int4-rtxpro4500-vllm-tp1
model instance id
cyankiwi-qwen3-8-27b-awq-int4--int4-w4a16
recipe source
0xsero
schema version
local-ai-registry/v1
status
candidate

capabilities

chat
Yes
reasoning
No
tools
No
vision
No

draft launch

accelerator backend
nvidia
arguments
--model, cyankiwi/Qwen3.8-27B-AWQ-INT4, --revision, 63768c10df38c0395e12ef49edac1bd539eaeeea, --served-model-name, Qwen3.8-27B, --tensor-parallel-size, 1, --max-model-len, 98304, --max-num-seqs, 2, --kv-cache-dtype, fp8, --gpu-memory-utilization, 0.97, --enable-auto-tool-choice, --tool-call-parser, qwen3_xml, --reasoning-parser, qwen3
container port
8,000
environment · VLLM MEMORY PROFILER ESTIMATE CUDAGRAPHS
0
host port
12,434
image
vllm/vllm-openai@sha256:0a51ea5b4ae2dc5d81890e5173f54203d2a3ae0cfffe51b8fd2afd4391bfd967
kind
docker
shm size
16g
synthesized · generated at
2026-09-02T16:35:13Z
synthesized · image provenance
qwen38-awq-int4-rtx3090-vllm-tp1
synthesized · template
clone-candidate-v1

engine

graph mode
full-and-piecewise
name
vllm
version
0.27.1

serving

max concurrency
4
max context tokens
98,304
tensor parallel
1
Provenance & metadata (3)

facts

metadata

derived from
qwen38-awq-int4-rtx3090-vllm-tp1

provenance

captured at
2026-09-02T16:35:13Z

sources

captured atkindurl
2026-09-02T16:35:13Znormalized-recipegithub.com/0xSero/local-ai-registry