Recipe

qwen3-8-27b-q6-k-rtx-5090-32gb-llama-cpp-tp1-l7yvwzwd

qwen3-8-27b-q6-k-rtx-5090-32gb-llama-cpp-tp1-l7yvwzwd

Observed LocalMaxxing leaderboard run. Evidence for compatibility, not an executable launch contract.

Record

Status
candidate
Source
localmaxxing
Engine
llama.cpp
Accelerators
1
Tensor parallel
1
Context tokens
2,048
chat
unknown
reasoning
unknown
tools
unknown
vision
unknown

Hugging Face model card

Identity

https://huggingface.co/unsloth/Qwen3.8-27B-GGUF
Repository
unsloth/Qwen3.8-27B-GGUF
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

llama.cpp

Evidence only · candidate · reference

Candidate evidence — not a Run contract

Source
https://www.localmaxxing.com/en/runs/cmt05msex0eitms01l7yvwzwd

Observed source tokens

Mechanical split of the source command. Unverified against the engine CLI. The registry does not offer Run for this recipe.

  1. llama-server
  2. --model
  3. Qwen3.8-27B-Q6_K.gguf
  4. --spec-draft-model
  5. Qwen3.8-27B-DFlash2-Q8_0.gguf
  6. --spec-type
  7. draft-dflash
  8. --spec-draft-n-max
  9. 3
  10. --spec-draft-ngl
  11. 999
  12. --ctx-size
  13. 8192
  14. --n-gpu-layers
  15. 999
  16. --flash-attn
  17. on
  18. --jinja
  19. --threads
  20. 8
FlagValue
--modelQwen3.8-27B-Q6_K.gguf
--spec-draft-modelQwen3.8-27B-DFlash2-Q8_0.gguf
--spec-typedraft-dflash
--spec-draft-n-max3
--spec-draft-ngl999
--ctx-size8192
--n-gpu-layers999
--flash-attnon
--threads8

Measured speed

ConcurrencyContextPrefillDecodeTTFT msStatusSweep
12,0481,043.197.770.9observedqwen3-8-27b-q6-k-rtx-5090-32gb-llama-cpp-tp1-l7yvwzwd-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
rtx-5090-32gb
id
qwen3-8-27b-q6-k-rtx-5090-32gb-llama-cpp-tp1-l7yvwzwd
model instance id
unsloth-qwen3-8-27b-gguf--q6-k
recipe source
localmaxxing
schema version
local-ai-registry/v1
speed sweep ids
qwen3-8-27b-q6-k-rtx-5090-32gb-llama-cpp-tp1-l7yvwzwd-sweep
status
candidate

capabilities

draft launch

accelerator backend
nvidia
arguments
-hf, unsloth/Qwen3.8-27B-GGUF:Q6_K, --n-gpu-layers, 999, --host, 0.0.0.0, --port, 8080, -c, 2048
container port
8,080
environment · LLAMA CACHE
/root/.cache/huggingface
host port
8,080
image
ghcr.io/ggml-org/llama.cpp:server-cuda12-b10481@sha256:b2497f8834f5ecb4e38530f6bf2734b8e0be107ff48e4720145911c86930f2ce
ipc
host
kind
docker
shm size
16g
synthesized · generated at
2026-08-31T22:12:17Z
synthesized · image provenance
gemma-4-12b-q4-k-m-rtx-3060-12gb-llama-cpp-tp1
synthesized · template
llama-cpp-server-v1

mounts

read onlytarget
No/root/.cache/huggingface

engine

name
llama.cpp

serving

max context tokens
2,048
tensor parallel
1
Provenance & metadata (3)

facts

capabilities.chat · provenance · captured at
2026-08-30T09:10:02Z
capabilities.chat · reason capability-not-verified
capabilities.chat · state unknown
capabilities.reasoning · provenance · captured at
2026-08-30T09:10:02Z
capabilities.reasoning · reason capability-not-verified
capabilities.reasoning · state unknown
capabilities.tools · provenance · captured at
2026-08-30T09:10:02Z
capabilities.tools · reason capability-not-verified
capabilities.tools · state unknown
capabilities.vision · provenance · captured at
2026-08-30T09:10:02Z
capabilities.vision · reason capability-not-verified
capabilities.vision · state unknown
draft launch.entrypoint · provenance · captured at
2026-09-01T01:41:26Z
draft launch.entrypoint · reason linux-amd64-container-config-entrypoint
draft launch.entrypoint · state known
engine.graph mode · provenance · captured at
2026-08-30T09:10:02Z
engine.graph mode · reason runtime-detail-not-published
engine.graph mode · state unknown
serving.kv cache tokens · provenance · captured at
2026-08-30T09:10:02Z
serving.kv cache tokens · reason kv-cache-capacity-not-published
serving.kv cache tokens · state unknown
serving.max concurrency · provenance · captured at
2026-08-31T23:03:15Z
serving.max concurrency · reason server-capacity-not-evidenced
serving.max concurrency · state unknown

metadata

localmaxxing · backend
cuda
localmaxxing · batch size
1
localmaxxing · hardware label
RTX 5090
localmaxxing · notes
llama.cpp built from PR #27342 with CUDA (sm_120), target Q6_K fully GPU-resident on the RTX 5090 with the 2 GB z-lab/Qwen3.8-27B-DFlash2-GGUF Q8_0 drafter alongside it (--spec-type draft-dflash --spec-draft-ngl 999), 8k context, flash attention on, greedy decode. Key tuning: --spec-draft-n-max 3 instead of the README's default 7 — draft acceptance on this workload ran ~2.4 tokens per block (vs ~5.3 on GSM8K), and sweeping n-max 2→7 gave 87.1 / 97.7 / 87.8 / 84.3 / 73.6 tok/s
localmaxxing · observed command
llama-server --model Qwen3.8-27B-Q6_K.gguf --spec-draft-model Qwen3.8-27B-DFlash2-Q8_0.gguf --spec-type draft-dflash --spec-draft-n-max 3 --spec-draft-ngl 999 --ctx-size 8192 --n-gpu-layers 999 --flash-attn on --jinja --threads 8
localmaxxing · run id
cmt05msex0eitms01l7yvwzwd
localmaxxing · tokenized · arguments
llama-server, --model, Qwen3.8-27B-Q6_K.gguf, --spec-draft-model, Qwen3.8-27B-DFlash2-Q8_0.gguf, --spec-type, draft-dflash, --spec-draft-n-max, 3, --spec-draft-ngl, 999, --ctx-size, 8192, --n-gpu-layers, 999, --flash-attn, on, --jinja, --threads, 8
localmaxxing · tokenized · fidelity
faithful

provenance

captured at
2026-08-30T09:10:02Z

sources

captured atkindurl
2026-08-30T09:10:02Znormalized-recipewww.localmaxxing.com/en/runs/cmt05msex0eitms01l7yvwzwd