Instructions to use josephmayo/Qwen2.5-agentic-7B-SLM-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 josephmayo/Qwen2.5-agentic-7B-SLM-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 josephmayo/Qwen2.5-agentic-7B-SLM-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf josephmayo/Qwen2.5-agentic-7B-SLM-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 josephmayo/Qwen2.5-agentic-7B-SLM-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf josephmayo/Qwen2.5-agentic-7B-SLM-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 josephmayo/Qwen2.5-agentic-7B-SLM-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf josephmayo/Qwen2.5-agentic-7B-SLM-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 josephmayo/Qwen2.5-agentic-7B-SLM-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf josephmayo/Qwen2.5-agentic-7B-SLM-GGUF:Q4_K_M

Use Docker

docker model run hf.co/josephmayo/Qwen2.5-agentic-7B-SLM-GGUF:Q4_K_M

LM Studio
Jan
Ollama
How to use josephmayo/Qwen2.5-agentic-7B-SLM-GGUF with Ollama:
```
ollama run hf.co/josephmayo/Qwen2.5-agentic-7B-SLM-GGUF:Q4_K_M
```

Unsloth Studio

How to use josephmayo/Qwen2.5-agentic-7B-SLM-GGUF 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 josephmayo/Qwen2.5-agentic-7B-SLM-GGUF 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 josephmayo/Qwen2.5-agentic-7B-SLM-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for josephmayo/Qwen2.5-agentic-7B-SLM-GGUF to start chatting

How to use josephmayo/Qwen2.5-agentic-7B-SLM-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf josephmayo/Qwen2.5-agentic-7B-SLM-GGUF: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": "josephmayo/Qwen2.5-agentic-7B-SLM-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use josephmayo/Qwen2.5-agentic-7B-SLM-GGUF with Docker Model Runner:
```
docker model run hf.co/josephmayo/Qwen2.5-agentic-7B-SLM-GGUF:Q4_K_M
```

Lemonade

How to use josephmayo/Qwen2.5-agentic-7B-SLM-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull josephmayo/Qwen2.5-agentic-7B-SLM-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Qwen2.5-agentic-7B-SLM-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use josephmayo/Qwen2.5-agentic-7B-SLM-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 josephmayo/Qwen2.5-agentic-7B-SLM-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 josephmayo/Qwen2.5-agentic-7B-SLM-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use josephmayo/Qwen2.5-agentic-7B-SLM-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf josephmayo/Qwen2.5-agentic-7B-SLM-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 "josephmayo/Qwen2.5-agentic-7B-SLM-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"

Qwen2.5-Coder-7B Agentic SLM v5 GGUF

This repository contains GGUF quantizations of the merged v5 7B model.

Base merged model: josephmayo/Qwen2.5-Coder-7B-agentic-SLM

Available quantizations:

qwen25-coder-agentic-slm-v5-q8_0.gguf
qwen25-coder-agentic-slm-v5-q6_k.gguf
qwen25-coder-agentic-slm-v5-q4_k_m.gguf

These files are intended for local inference with llama.cpp-compatible runtimes.

Current Proof Gate

The proof gate was run on the LoRA/merged model inside a verifier-rescue system.

Kaggle proof kernel: holykeys/qwen25-coder-agentic-slm-v5-rescue

Phase	Greedy pass@1	Coverage@K	Selected@K	Repair	Final
Qwen2.5-Coder-7B reference harness	37/50	40/50	40/50	2/50	42/50
v5 7B model primary	37/50	42/50	42/50	2/50	44/50
14B rescue on primary misses	1/6	3/6	3/6	1/6	4/6
v5 combined rescue system	38/50	45/50	45/50	3/50	48/50

Lift Summary

Against the 42/50 Qwen2.5-Coder-7B reference harness:

7B model primary: 44/50, +2/50, +4.76% relative.
Full v5 rescue system: 48/50, +6/50, +14.29% relative.
Failure reduction: 8 misses to 2 misses, 75% fewer failures.

Quantization Notes

The GGUF files were produced from the merged v5 7B model.

Recommended use:

Q8_0: highest quality among these quants, larger file.
Q6_K: good quality/size tradeoff.
Q4_K_M: smaller local deployment option, expected to lose some accuracy.

The published proof numbers were not rerun separately for each quant. Quantized evaluations should be run before making claims about exact Q8/Q6/Q4 performance.

Required Quant Eval

Before ranking these quants, run:

HumanEval/MBPP fast gate for all three quants.
LiveCodeBench recent slice.
BigCodeBench small then full split.
Latency per task on local CPU/GPU.
Tokens/sec, memory usage, and load time.
Invalid output rate: markdown leakage, syntax error, missing entrypoint.
Abstention/no-answer rate.

No Frontier Claim

This repository does not claim to beat Claude Sonnet 4.5.

The goal of this release is to provide deployable local artifacts for the current v5 agentic coding system while preserving exact benchmark provenance.

Example llama.cpp Usage

llama-cli \
  -m qwen25-coder-agentic-slm-v5-q6_k.gguf \
  -p "Return code only. Write a Python function add(a, b)." \
  -n 256

Use a verifier/test harness for coding tasks. Single-shot chat usage is not the intended evaluation mode.

Downloads last month: 79

GGUF

Model size

8B params

Architecture

qwen2

Hardware compatibility

4-bit

6-bit

8-bit

Inference Providers NEW

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

Model tree for josephmayo/Qwen2.5-agentic-7B-SLM-GGUF

Base model

Qwen/Qwen2.5-7B

Finetuned

Qwen/Qwen2.5-Coder-7B

Finetuned

Qwen/Qwen2.5-Coder-7B-Instruct

Finetuned

josephmayo/Qwen2.5-agentic-7B-SLM

Quantized

(2)

this model

Collection including josephmayo/Qwen2.5-agentic-7B-SLM-GGUF

SLMs / ELMs & fine tuned

Collection

these are models fine tuned to be used for specific usecase - mainly coding & agentic tasks • 9 items • Updated Jun 20