Recipe
gemma-4-12b-it-qat-q4-0-unquantized-q4-0-rtx-3060-ti-8gb-llama-cpp-tp1
gemma-4-12b-it-qat-q4-0-unquantized-q4-0-rtx-3060-ti-8gb-llama-cpp-tp1Observed LocalMaxxing leaderboard run. Evidence for compatibility, not an executable launch contract.
Record
- Status
- candidate
- Source
- localmaxxing
- Engine
- llama.cpp
- Engine version
- b9653 / 9dbc6621a
- Accelerators
- 1
- Tensor parallel
- 1
- Context tokens
- 2,048
- Max concurrency
- 2,048
- chat
- unknown
- reasoning
- unknown
- tools
- unknown
- vision
- unknown
Hugging Face model card
Identity
https://huggingface.co/google/gemma-4-12B-it-qat-q4_0-gguf- Repository
- google/gemma-4-12B-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.
llama-bench-hfgoogle/gemma-4-12B-it-qat-q4_0-gguf-hffgemma-4-12b-it-qat-q4_0.gguf-ngl99-p512-n128-r10-omd
| Flag | Value |
|---|---|
-hf | google/gemma-4-12B-it-qat-q4_0-gguf |
-hff | gemma-4-12b-it-qat-q4_0.gguf |
-ngl | 99 |
-p | 512 |
-n | 128 |
-r | 10 |
-o | md |
Measured speed
| Concurrency | Context | Prefill | Decode | TTFT ms | Status | Sweep |
|---|---|---|---|---|---|---|
| 2048 | 2,048 | 35.1 | 17.4 | — | observed | gemma-4-12b-it-qat-q4-0-unquantized-q4-0-rtx-3060-ti-8gb-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-ti-8gb
- id
- gemma-4-12b-it-qat-q4-0-unquantized-q4-0-rtx-3060-ti-8gb-llama-cpp-tp1
- model instance id
- google-gemma-4-12b-it-qat-q4-0-gguf--q4-0
- recipe source
- localmaxxing
- schema version
- local-ai-registry/v1
- speed sweep ids
- gemma-4-12b-it-qat-q4-0-unquantized-q4-0-rtx-3060-ti-8gb-llama-cpp-tp1-sweep
- status
- candidate
capabilities
draft launch
- accelerator backend
- nvidia
- arguments
- -hf, google/gemma-4-12B-it-qat-q4_0-gguf, --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: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
- b9653 / 9dbc6621a
serving
- max concurrency
- 2,048
- 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
- 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 Ti
- localmaxxing · notes
- llama.cpp b9653 CUDA backend on Windows 10, RTX 3060 Ti 8GB. User requested google/gemma-4-12B-it; llama.cpp requires GGUF, so this run used Google's official GGUF repo google/gemma-4-12B-it-qat-q4_0-gguf with file gemma-4-12b-it-qat-q4_0.gguf. Q4_0 quantization, all layers offloaded with -ngl 99. Model size reported by llama-bench was 6.48 GiB / 11.91B params, close to available VRAM on this 8GB GPU; benchmark showed high run-to-run variance. Benchmark used -p 512 -n 128 -r 10 and llama-bench defaults unless specified: context 2048, batch 2048, ubatch 512, flash-attn auto, mmap enabled. TTFT and peak VRAM were not measured. Initial Hugging Face cache symlink warning on Windows was harmless; inference completed normally with CUDA.
- localmaxxing · observed command
- llama-bench -hf google/gemma-4-12B-it-qat-q4_0-gguf -hff gemma-4-12b-it-qat-q4_0.gguf -ngl 99 -p 512 -n 128 -r 10 -o md
- localmaxxing · run id
- cmqfmh4bg00cmmt01qujsk0yl
- localmaxxing · tokenized · arguments
- llama-bench, -hf, google/gemma-4-12B-it-qat-q4_0-gguf, -hff, gemma-4-12b-it-qat-q4_0.gguf, -ngl, 99, -p, 512, -n, 128, -r, 10, -o, md
- 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/cmqfmh4bg00cmmt01qujsk0yl ↗ |