Instructions to use VoiceScribe/voicescribe-corrector-gemma3-1b-q4-gguf-metal with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use VoiceScribe/voicescribe-corrector-gemma3-1b-q4-gguf-metal with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf VoiceScribe/voicescribe-corrector-gemma3-1b-q4-gguf-metal:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf VoiceScribe/voicescribe-corrector-gemma3-1b-q4-gguf-metal:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf VoiceScribe/voicescribe-corrector-gemma3-1b-q4-gguf-metal:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf VoiceScribe/voicescribe-corrector-gemma3-1b-q4-gguf-metal:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf VoiceScribe/voicescribe-corrector-gemma3-1b-q4-gguf-metal:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf VoiceScribe/voicescribe-corrector-gemma3-1b-q4-gguf-metal:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf VoiceScribe/voicescribe-corrector-gemma3-1b-q4-gguf-metal:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf VoiceScribe/voicescribe-corrector-gemma3-1b-q4-gguf-metal:Q4_K_M

Use Docker

docker model run hf.co/VoiceScribe/voicescribe-corrector-gemma3-1b-q4-gguf-metal:Q4_K_M

LM Studio
Jan
Ollama
How to use VoiceScribe/voicescribe-corrector-gemma3-1b-q4-gguf-metal with Ollama:
```
ollama run hf.co/VoiceScribe/voicescribe-corrector-gemma3-1b-q4-gguf-metal:Q4_K_M
```
Unsloth Desktop
Docker Model Runner
How to use VoiceScribe/voicescribe-corrector-gemma3-1b-q4-gguf-metal with Docker Model Runner:
```
docker model run hf.co/VoiceScribe/voicescribe-corrector-gemma3-1b-q4-gguf-metal:Q4_K_M
```

Lemonade

How to use VoiceScribe/voicescribe-corrector-gemma3-1b-q4-gguf-metal with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull VoiceScribe/voicescribe-corrector-gemma3-1b-q4-gguf-metal:Q4_K_M

Run and chat with the model

lemonade run user.voicescribe-corrector-gemma3-1b-q4-gguf-metal-Q4_K_M

List all available models

lemonade list

Atomic Chat

voicescribe-corrector-gemma3-1b-q4-gguf-metal

Voice Scribe mirror for a local macOS corrector candidate.

Recommended profile

tiny fallback, weaker than Qwen2.5 1.5B

Benchmark snapshot

Teacher benchmark WP#529, 100 Russian correction + RU->EN translation rows:

Correction: 7.53/10
Translation: 7.49/10
Combined: 7.51/10
Observed max RSS on dev M2 Max: 934 MB

Hardware note: M2 Max measurements are a development baseline. Real M1 8GB, M3 Air 16GB, and M4 Mac mini 16GB should be validated before product default changes.

Files

gemma-3-1b-it-Q4_K_M.gguf
UPSTREAM_SOURCE.md
voicescribe-model-layout.json

Original upstream license and terms remain governed by the upstream repository.

Downloads last month: 9

GGUF

Model size

1.0B params

Architecture

gemma3

Hardware compatibility

4-bit