Recipe
gemma-4-e4b-it-qat-q4-0-unquantized-q4-k-xl-rtx-3060-12gb-llama-cpp-tp1
gemma-4-e4b-it-qat-q4-0-unquantized-q4-k-xl-rtx-3060-12gb-llama-cpp-tp1Observed LocalMaxxing leaderboard run. Evidence for compatibility, not an executable launch contract.
Record
- Status
- candidate
- Source
- localmaxxing
- Engine
- llama.cpp
- Engine version
- b9775 / be4a6a63e
- Accelerators
- 1
- Tensor parallel
- 1
- Context tokens
- 4,608
- Max concurrency
- 1
- chat
- unknown
- reasoning
- unknown
- tools
- unknown
- vision
- unknown
Hugging Face model card
Identity
https://huggingface.co/google/gemma-4-E4B-it-qat-q4_0-gguf- Repository
- google/gemma-4-E4B-it-qat-q4_0-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
Observed source tokens
Mechanical split of the source command. Unverified against the engine CLI. The registry does not offer Run for this recipe.
/home/dario/llama.cpp/build/bin/llama-bench-m/mnt/hdd/models/gemma-4-E4B-it-qat-UD-Q4_K_XL.gguf-p4096-n512-b4096-ub1024-t16-ngl99-faon-r2-ojson--no-warmup
| Flag | Value |
|---|---|
-m | /mnt/hdd/models/gemma-4-E4B-it-qat-UD-Q4_K_XL.gguf |
-p | 4096 |
-n | 512 |
-b | 4096 |
-ub | 1024 |
-t | 16 |
-ngl | 99 |
-fa | on |
-r | 2 |
-o | json |
Measured speed
| Concurrency | Context | Prefill | Decode | TTFT ms | Status | Sweep |
|---|---|---|---|---|---|---|
| 1 | 4,608 | 3,012.7 | 88.1 | 1,370.9 | observed | gemma-4-e4b-it-qat-q4-0-unquantized-q4-k-xl-rtx-3060-12gb-llama-cpp-tp1-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-3060-12gb
- id
- gemma-4-e4b-it-qat-q4-0-unquantized-q4-k-xl-rtx-3060-12gb-llama-cpp-tp1
- model instance id
- google-gemma-4-e4b-it-qat-q4-0-gguf--q4-k-xl
- recipe source
- localmaxxing
- schema version
- local-ai-registry/v1
- speed sweep ids
- gemma-4-e4b-it-qat-q4-0-unquantized-q4-k-xl-rtx-3060-12gb-llama-cpp-tp1-sweep
- status
- candidate
capabilities
draft launch
- accelerator backend
- nvidia
- arguments
- -hf, google/gemma-4-E4B-it-qat-q4_0-gguf:Q4_K_XL, --n-gpu-layers, 999, --host, 0.0.0.0, --port, 8080, -c, 4608
- 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:b2497f8834f5ecb4e38530f6bf2734b8e0be107f0f0857e259672d1cb85b71c2
- 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 only | target |
|---|---|
| No | /root/.cache/huggingface |
engine
- name
- llama.cpp
- version
- b9775 / be4a6a63e
serving
- max concurrency
- 1
- max context tokens
- 4,608
- 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
- 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
- cuda
- localmaxxing · hardware label
- RTX 3060
- localmaxxing · notes
- 4096 prompt / 512 generation. Two throwaway llama-bench runs preceded the recorded sweep. Winner selected by minimum measured TTFT across (b2048/ub512/t16), (b4096/ub1024/t16), and (b4096/ub512/t32), with exact command preserved. Local file: /mnt/hdd/models/gemma-4-E4B-it-qat-UD-Q4_K_XL.gguf. Temperature 0. Peak VRAM and maximum sampled sustained GPU power were measured during the winner run.
- localmaxxing · observed command
- /home/dario/llama.cpp/build/bin/llama-bench -m /mnt/hdd/models/gemma-4-E4B-it-qat-UD-Q4_K_XL.gguf -p 4096 -n 512 -b 4096 -ub 1024 -t 16 -ngl 99 -fa on -r 2 -o json --no-warmup
- localmaxxing · run id
- cmrz00bvh05h6o401em8c6unn
- localmaxxing · tokenized · arguments
- /home/dario/llama.cpp/build/bin/llama-bench, -m, /mnt/hdd/models/gemma-4-E4B-it-qat-UD-Q4_K_XL.gguf, -p, 4096, -n, 512, -b, 4096, -ub, 1024, -t, 16, -ngl, 99, -fa, on, -r, 2, -o, json, --no-warmup
- localmaxxing · tokenized · fidelity
- faithful
provenance
- captured at
- 2026-08-30T09:10:02Z
sources
| captured at | kind | url |
|---|---|---|
| 2026-08-30T09:10:02Z | normalized-recipe | www.localmaxxing.com/en/runs/cmrz00bvh05h6o401em8c6unn ↗ |