Instructions to use tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-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 tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-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 tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-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 tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-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 tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-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 tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-GGUF:Q4_K_M

Use Docker

docker model run hf.co/tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-GGUF:Q4_K_M

LM Studio
Jan
Ollama
How to use tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-GGUF with Ollama:
```
ollama run hf.co/tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-GGUF:Q4_K_M
```
Unsloth Desktop

How to use tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-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": "tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-GGUF with Docker Model Runner:
```
docker model run hf.co/tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-GGUF:Q4_K_M
```

Lemonade

How to use tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Qwen3.8-2B-Function-Calling-xLAM-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-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 tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-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 tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-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 "tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-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"

Qwen3.8-2B-Function-Calling-xLAM GGUF

GGUF quantizations of ermiaazarkhalili/Qwen3.8-2B-Function-Calling-xLAM-Unsloth, a 2B parameter model fine-tuned for function calling and tool use via the xLAM pipeline.

Model Architecture

Property	Value
Architecture	Qwen3.5 (SSM + Attention hybrid)
Parameters	2B
Context Length	262144 tokens
Base Model	Qwen3.8-2B

Quantizations

Quantization	File Size	Description
Q8_0	2.08 GB	Near-lossless, maximum quality
Q6_K	1.61 GB	Very high quality
Q5_K_M	1.45 GB	High quality
Q5_K_S	1.42 GB	High quality, smaller
Q4_K_M	1.31 GB	Recommended, good balance
Q4_K_S	1.25 GB	Good quality, smaller
IQ4_XS	1.23 GB	Good quality, imatrix calibrated
Q3_K_L	1.20 GB	Medium quality
Q3_K_M	1.13 GB	Medium quality
Q3_K_S	1.05 GB	Below average quality
Q2_K	990 MB	Low quality, smallest

Q4_K_M is recommended for most use cases.

Usage

Download a single file with huggingface-cli:

huggingface-cli download tinyopsec/Qwen3.8-2B-Function-Calling-xLAM-GGUF --include "Q4_K_M.gguf" --local-dir ./

Run with llama.cpp:

llama-cli -m Q4_K_M.gguf -p "Your prompt here" -n 512

Run as a server:

llama-server -m Q4_K_M.gguf --port 8080

Chat Template

This model uses the Qwen3 chat template:

<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant

Conversion Details

Quantizations were produced using llama.cpp. The intermediate F16 GGUF was created with convert_hf_to_gguf.py and K-quants were produced with llama-quantize. IQ4_XS was produced using an imatrix calibration file.