Instructions to use guildlm/go-test with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use guildlm/go-test with MLX:

# Make sure mlx-lm is installed
# pip install --upgrade mlx-lm

# Generate text with mlx-lm
from mlx_lm import load, generate

model, tokenizer = load("guildlm/go-test")

prompt = "Write a story about Einstein"
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
    messages, add_generation_prompt=True
)

text = generate(model, tokenizer, prompt=prompt, verbose=True)

llama-cpp-python

How to use guildlm/go-test with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="guildlm/go-test",
	filename="go-test.Q4_K_M.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use guildlm/go-test 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 guildlm/go-test:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf guildlm/go-test:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf guildlm/go-test:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf guildlm/go-test: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 guildlm/go-test:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf guildlm/go-test: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 guildlm/go-test:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf guildlm/go-test:Q4_K_M

Use Docker

docker model run hf.co/guildlm/go-test:Q4_K_M

LM Studio
Jan

vLLM

How to use guildlm/go-test with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "guildlm/go-test"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "guildlm/go-test",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/guildlm/go-test:Q4_K_M

Ollama
How to use guildlm/go-test with Ollama:
```
ollama run hf.co/guildlm/go-test:Q4_K_M
```

Unsloth Studio

How to use guildlm/go-test 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 guildlm/go-test 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 guildlm/go-test to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for guildlm/go-test to start chatting

How to use guildlm/go-test with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "guildlm/go-test"

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "guildlm/go-test"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use guildlm/go-test with Hermes Agent:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "guildlm/go-test"

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 guildlm/go-test

Run Hermes

hermes

Atomic Chat new

MLX LM

How to use guildlm/go-test with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "guildlm/go-test"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "guildlm/go-test"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "guildlm/go-test",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

Docker Model Runner
How to use guildlm/go-test with Docker Model Runner:
```
docker model run hf.co/guildlm/go-test:Q4_K_M
```

Lemonade

How to use guildlm/go-test with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull guildlm/go-test:Q4_K_M

Run and chat with the model

lemonade run user.go-test-Q4_K_M

List all available models

lemonade list

GuildLM · go-test

A small, sharp Go testing specialist from the GuildLM Code Guild.

go-test writes thorough, table-driven Go tests with real assertions and meaningful edge cases — the kind of tests that actually fail when the code is wrong. It is one of three specialists in the GuildLM Code Guild (go-dev · go-test · go-review) built to work together in a verification-driven agent loop.

Why a dedicated test model? Across every GuildLM experiment, writing tests is the one job where targeted fine-tuning clearly beats the base. On a mutation benchmark — inject a bug, does the generated test catch it? — go-test catches 14/18 (78%) versus 7/18 (39%) for the untuned base. It catches twice as many real bugs as the model it's built on. Test-writing is genuinely where specialization pays. See the research log, Report #2.

Why this isn't "just Qwen with a name"

go-test is a fused, standalone model (no separate adapter) with its own identity — ask who it is and it answers GuildLM go-test. It is an honest Apache-2.0 derivative of Qwen2.5-Coder-7B-Instruct, fine-tuned for one craft: testing Go.

What it's for

Generating *_test.go files: table-driven cases, sub-tests, httptest, sync/atomic for concurrency checks.
Covering the edges a tired human skips: empty input, zero values, negatives, inverted bounds, error paths.
Acting as the testing role inside the GuildLM Builder, so the code that go-dev writes is held to a real, executable contract.

Benchmarks

Measured locally with the real Go toolchain — tests are scored by whether they compile, assert, and catch injected mutations, not by an LLM judge.

Benchmark	Metric	go-test	base 7B
crucible `go_test_bench` (18 mutations)	bug-catch@1	14/18 · 78%	7/18 · 39%

This is the GuildLM specialist where the fine-tuning win is real, large, and repeatable — go-test catches 2× the injected bugs of its base. A guard in the Builder also rejects trivially-passing tests (no assertions), so its output is held to a real bar in the loop.

Quickstart

Apple Silicon (MLX)

pip install mlx-lm
python -m mlx_lm generate --model guildlm/go-test \
  --prompt "Write table-driven Go tests for func Clamp(x, lo, hi int) int covering below/inside/above range and inverted bounds." \
  --max-tokens 400

Ollama (GGUF)

ollama run guildlm/go-test "Write httptest-based tests for a POST /echo JSON endpoint built with newMux()."

Inside the agent loop (recommended)

python -m mlx_lm server --model guildlm/go-test --port 8081
guildlm-build --spec specs/myservice.yaml --out ./out \
  --base-url http://localhost:8080/v1 \
  --test-model guildlm/go-test --test-base-url http://localhost:8081/v1 \
  --candidates 3

Prompting

Trained with the system prompt:

You are GuildLM go-test, a Go testing specialist from the GuildLM Code Guild.

Give it a function signature or an implementation and ask for tests. It defaults to table-driven style with explicit t.Errorf/t.Fatalf assertions.

The Guild

Specialist	Job
go-dev	writes the implementation
go-test	writes thorough table-driven tests
go-review	audits for bugs a green build hides

Agent loop: https://github.com/guildlm/builder
Research log: https://guildlm.github.io/research/

License & attribution

Apache-2.0, inherited from Qwen2.5-Coder-7B-Instruct (© Alibaba Cloud). GuildLM fine-tuning, identity, and packaging under the same license. Trained locally on Apple Silicon with MLX — total cloud spend: $0.