Instructions to use RedTeamLab/Gemma-4-12B-Sol-Traces-v1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use RedTeamLab/Gemma-4-12B-Sol-Traces-v1 with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="RedTeamLab/Gemma-4-12B-Sol-Traces-v1",
	filename="gemma-4-12b-sol-traces-v1-Q4_K_M.gguf",
)

llm.create_chat_completion(
	messages = "No input example has been defined for this model task."
)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use RedTeamLab/Gemma-4-12B-Sol-Traces-v1 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 RedTeamLab/Gemma-4-12B-Sol-Traces-v1:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf RedTeamLab/Gemma-4-12B-Sol-Traces-v1:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf RedTeamLab/Gemma-4-12B-Sol-Traces-v1:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf RedTeamLab/Gemma-4-12B-Sol-Traces-v1: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 RedTeamLab/Gemma-4-12B-Sol-Traces-v1:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf RedTeamLab/Gemma-4-12B-Sol-Traces-v1: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 RedTeamLab/Gemma-4-12B-Sol-Traces-v1:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf RedTeamLab/Gemma-4-12B-Sol-Traces-v1:Q4_K_M

Use Docker

docker model run hf.co/RedTeamLab/Gemma-4-12B-Sol-Traces-v1:Q4_K_M

LM Studio
Jan
Ollama
How to use RedTeamLab/Gemma-4-12B-Sol-Traces-v1 with Ollama:
```
ollama run hf.co/RedTeamLab/Gemma-4-12B-Sol-Traces-v1:Q4_K_M
```

Unsloth Studio

How to use RedTeamLab/Gemma-4-12B-Sol-Traces-v1 with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for RedTeamLab/Gemma-4-12B-Sol-Traces-v1 to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for RedTeamLab/Gemma-4-12B-Sol-Traces-v1 to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for RedTeamLab/Gemma-4-12B-Sol-Traces-v1 to start chatting

How to use RedTeamLab/Gemma-4-12B-Sol-Traces-v1 with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf RedTeamLab/Gemma-4-12B-Sol-Traces-v1:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "RedTeamLab/Gemma-4-12B-Sol-Traces-v1:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use RedTeamLab/Gemma-4-12B-Sol-Traces-v1 with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf RedTeamLab/Gemma-4-12B-Sol-Traces-v1:Q4_K_M

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default RedTeamLab/Gemma-4-12B-Sol-Traces-v1:Q4_K_M

Run Hermes

hermes

Atomic Chat new

OpenClaw new

How to use RedTeamLab/Gemma-4-12B-Sol-Traces-v1 with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf RedTeamLab/Gemma-4-12B-Sol-Traces-v1:Q4_K_M

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "RedTeamLab/Gemma-4-12B-Sol-Traces-v1:Q4_K_M" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Docker Model Runner
How to use RedTeamLab/Gemma-4-12B-Sol-Traces-v1 with Docker Model Runner:
```
docker model run hf.co/RedTeamLab/Gemma-4-12B-Sol-Traces-v1:Q4_K_M
```

Lemonade

How to use RedTeamLab/Gemma-4-12B-Sol-Traces-v1 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull RedTeamLab/Gemma-4-12B-Sol-Traces-v1:Q4_K_M

Run and chat with the model

lemonade run user.Gemma-4-12B-Sol-Traces-v1-Q4_K_M

List all available models

lemonade list

Gemma-4-12B-Sol-Traces-v1

Agentic tool-use LLM fine-tuned from unsloth/gemma-4-12B-it using LoRA.
Trained on 25,000 verified coding-agent trajectories with the same Sol Traces methodology as the 26B variant, but on a smaller, faster base model.

Training Details

Parameter	Value
Base model	`unsloth/gemma-4-12B-it` (12B unified)
Fine-tuning	LoRA (r=16, alpha=16, dropout=0)
Dataset	21,174 train / 1,324 val (gemma-4-native-tools format)
Dataset provenance	GPT-5.6-Sol trajectories (original-synthetic, 32,560 attempts → 25,000 accepted)
Epochs	1
Learning rate	1e-4, cosine scheduler with 3% warmup
Batch size	8 (2 × 4 gradient accumulation)
Max sequence	8,192 tokens
GPU	Modal H100 80GB
Training time	~3h
Final train loss	0.080
Validation loss	0.0258
Peak VRAM	46.7 GiB / 80 GiB

Files

File	Size	Description
`gemma-4-12b-sol-traces-v1-f16.gguf`	22.2 GB	Full bf16 merged model
`training_stats.json`	—	Full training metrics

Usage (llama.cpp)

llama-cli \
  -m gemma-4-12b-sol-traces-v1-f16.gguf \
  -ngl 99 \
  --prompt "List the files in the repository matching *.py"

Data Generation

Trajectories generated by GPT-5.6-Sol (OpenAI) running as a repository coding agent through OpenCode CLI. Each trajectory captures a full agent session: code search, file reading, command execution, patch application, and verification — across 85+ repository families in Python, TypeScript, Rust, Go, Java, C++, and more.

Downloads last month: -

GGUF

Model size

12B params

Architecture

gemma4

Hardware compatibility

4-bit

16-bit

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support