Instructions to use jersobh/jeff-2b-gguf with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use jersobh/jeff-2b-gguf 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 jersobh/jeff-2b-gguf:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf jersobh/jeff-2b-gguf:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf jersobh/jeff-2b-gguf:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf jersobh/jeff-2b-gguf: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 jersobh/jeff-2b-gguf:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf jersobh/jeff-2b-gguf: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 jersobh/jeff-2b-gguf:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf jersobh/jeff-2b-gguf:Q4_K_M

Use Docker

docker model run hf.co/jersobh/jeff-2b-gguf:Q4_K_M

LM Studio
Jan
Ollama
How to use jersobh/jeff-2b-gguf with Ollama:
```
ollama run hf.co/jersobh/jeff-2b-gguf:Q4_K_M
```
Unsloth Desktop

How to use jersobh/jeff-2b-gguf with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf jersobh/jeff-2b-gguf:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "jersobh/jeff-2b-gguf:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use jersobh/jeff-2b-gguf with Docker Model Runner:
```
docker model run hf.co/jersobh/jeff-2b-gguf:Q4_K_M
```

Lemonade

How to use jersobh/jeff-2b-gguf with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull jersobh/jeff-2b-gguf:Q4_K_M

Run and chat with the model

lemonade run user.jeff-2b-gguf-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use jersobh/jeff-2b-gguf with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf jersobh/jeff-2b-gguf: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 jersobh/jeff-2b-gguf:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use jersobh/jeff-2b-gguf with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf jersobh/jeff-2b-gguf: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 "jersobh/jeff-2b-gguf: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"

Jeff-2B Decision Model (Optimized Dual Precision Benchmarks)

Fine-tuned on top of internlm/Intern-Decision-2B using Unsloth with extended 5-epoch training and hard-negative mining. Optimized for structured schema-guided routing and intent classification.

Benchmark Results Comparison

Benchmark / Task	Evaluation Slice	Unquantized (BF16)	Quantized (Q4_K_M GGUF)
BANKING77 Intent Classification	`test[:1000]`	95.80%	95.60%
MASSIVE Intent Classification	`test[:1000]`	87.10%	87.00%
CLINC150 Intent Classification	`test[:1000]`	95.20%	95.00%
Type-Decision Routing (Dolly)	`train[:1000] (Held-out)`	90.30%	90.10%

Training Configuration

Base Model: internlm/Intern-Decision-2B
Precision: Native BF16 (NVIDIA L4 24GB)
Method: LoRA (Rank: 32, Alpha: 32)
Epochs: 5 (Hard-negative mining enabled)
Export Precision: 4-bit Quantized GGUF (Q4_K_M)

Usage with llama-cli

./llama-cli -m jeff-2B.Q4_K_M.gguf -ngl 99 -n 4 -p "Task: routing\nState: Query user account balance\nOptions: [A: Execute tool, B: Chat directly]\n### Decision:\n"