Local AI Registry

Recipe

muse-glimmer-30b-ud-q8-k-xl-radeon-ai-pro-r9700-32gb-llama-cpp-tp3

muse-glimmer-30b-ud-q8-k-xl-radeon-ai-pro-r9700-32gb-llama-cpp-tp3
JSON APIFind compatible recipes

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

Record

Status
candidate
Source
localmaxxing
Engine
llama.cpp
Engine version
b10434
Accelerators
3
Tensor parallel
3
Context tokens
834
Max concurrency
1
chat
unknown
reasoning
unknown
tools
unknown
vision
unknown

Hugging Face model card

Identity

https://huggingface.co/meta-models/Muse-Glimmer-30B
Repository
meta-models/Muse-Glimmer-30B
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/cmsvu44yg08dnms015vw6cpw6

Observed source tokens

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

  1. /home/mikekey/.config/llama-swap/gpu-order.sh
  2. exec
  3. --compute-only
  4. /home/mikekey/Experiments/llama.cpp-rdna/build-rocm-tweaked714/bin/llama-server
  5. --port
  6. ${PORT}
  7. --host
  8. 0.0.0.0
  9. --no-webui
  10. -ngl
  11. 99
  12. --jinja
  13. --flash-attn
  14. auto
  15. --cache-reuse
  16. 256
  17. --prio
  18. 3
  19. -sm
  20. tensor
  21. -m
  22. /home/mikekey/models/gguf/meta/Muse-Glimmer-30B-UD-Q8_K_XL.gguf
  23. --mmproj
  24. /home/mikekey/models/gguf/meta/mmproj-Muse-Glimmer-30B-Q8_0.gguf
  25. -md
  26. /home/mikekey/models/gguf/meta/Muse-Glimmer-30B-dflash-draft-Q8_0.gguf
  27. --spec-type
  28. draft-dflash
  29. --spec-draft-p-min
  30. 0.3
  31. --spec-draft-n-max
  32. 16
  33. --chat-template-kwargs
  34. {"reasoning_strength":"zero"}
  35. --reasoning-format
  36. auto
  37. --temp
  38. 0.6
  39. --top-p
  40. 0.95
  41. --top-k
  42. 20
  43. --min-p
  44. 0.0
  45. --repeat-penalty
  46. 1.0
  47. -ctk
  48. f16
  49. -ctv
  50. f16
  51. -ts
  52. 1/1
  53. -np
  54. 1
  55. -c
  56. 131072
FlagValue
--compute-only/home/mikekey/Experiments/llama.cpp-rdna/build-rocm-tweaked714/bin/llama-server
--port${PORT}
--host0.0.0.0
-ngl99
--flash-attnauto
--cache-reuse256
--prio3
-smtensor
-m/home/mikekey/models/gguf/meta/Muse-Glimmer-30B-UD-Q8_K_XL.gguf
--mmproj/home/mikekey/models/gguf/meta/mmproj-Muse-Glimmer-30B-Q8_0.gguf
-md/home/mikekey/models/gguf/meta/Muse-Glimmer-30B-dflash-draft-Q8_0.gguf
--spec-typedraft-dflash
--spec-draft-p-min0.3
--spec-draft-n-max16
--chat-template-kwargs{"reasoning_strength":"zero"}
--reasoning-formatauto
--temp0.6
--top-p0.95
--top-k20
--min-p0.0
--repeat-penalty1.0
-ctkf16
-ctvf16
-ts1/1
-np1
-c131072

Measured speed

ConcurrencyContextPrefillDecodeTTFT msStatusSweep
1834787.547.8734observedmuse-glimmer-30b-ud-q8-k-xl-radeon-ai-pro-r9700-32gb-llama-cpp-tp3-sweep

Remaining fields

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

hardware count
3
hardware id
radeon-ai-pro-r9700-32gb
id
muse-glimmer-30b-ud-q8-k-xl-radeon-ai-pro-r9700-32gb-llama-cpp-tp3
model instance id
meta-models-muse-glimmer-30b--ud-q8-k-xl
recipe source
localmaxxing
schema version
local-ai-registry/v1
speed sweep ids
muse-glimmer-30b-ud-q8-k-xl-radeon-ai-pro-r9700-32gb-llama-cpp-tp3-sweep
status
candidate

capabilities

engine

name
llama.cpp
version
b10434

serving

max concurrency
1
max context tokens
834
tensor parallel
3
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
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

metadata

localmaxxing · backend
rocm
localmaxxing · hardware label
Radeon AI Pro R9700
localmaxxing · notes
stew675 rdna-boosts fork + ROCm 7.14 TheRock, 2x R9700 -sm tensor, DFlash draft n-max 16, GGML_CUDA_ALLREDUCE=internal over GPU P2P (kernel 7.1.8). Was 34 layer-split.
localmaxxing · run id
cmsvu44yg08dnms015vw6cpw6
localmaxxing · tokenized · arguments
/home/mikekey/.config/llama-swap/gpu-order.sh, exec, --compute-only, /home/mikekey/Experiments/llama.cpp-rdna/build-rocm-tweaked714/bin/llama-server, --port, ${PORT}, --host, 0.0.0.0, --no-webui, -ngl, 99, --jinja, --flash-attn, auto, --cache-reuse, 256, --prio, 3, -sm, tensor, -m, /home/mikekey/models/gguf/meta/Muse-Glimmer-30B-UD-Q8_K_XL.gguf, --mmproj, /home/mikekey/models/gguf/meta/mmproj-Muse-Glimmer-30B-Q8_0.gguf, -md, /home/mikekey/models/gguf/meta/Muse-Glimmer-30B-dflash-draft-Q8_0.gguf, --spec-type, draft-dflash, --spec-draft-p-min, 0.3, --spec-draft-n-max, 16, --chat-template-kwargs, {"reasoning_strength":"zero"}, --reasoning-format, auto, --temp, 0.6, --top-p, 0.95, --top-k, 20, --min-p, 0.0, --repeat-penalty, 1.0, -ctk, f16, -ctv, f16, -ts, 1/1, -np, 1, -c, 131072
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/cmsvu44yg08dnms015vw6cpw6 ↗