Instructions to use Ilides/coser-2-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 Ilides/coser-2-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 Ilides/coser-2-GGUF:F16
# Run inference directly in the terminal:
llama cli -hf Ilides/coser-2-GGUF:F16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Ilides/coser-2-GGUF:F16
# Run inference directly in the terminal:
llama cli -hf Ilides/coser-2-GGUF:F16

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 Ilides/coser-2-GGUF:F16
# Run inference directly in the terminal:
./llama-cli -hf Ilides/coser-2-GGUF:F16

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 Ilides/coser-2-GGUF:F16
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Ilides/coser-2-GGUF:F16

Use Docker

docker model run hf.co/Ilides/coser-2-GGUF:F16

LM Studio
Jan

vLLM

How to use Ilides/coser-2-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Ilides/coser-2-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Ilides/coser-2-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Ilides/coser-2-GGUF:F16

Ollama
How to use Ilides/coser-2-GGUF with Ollama:
```
ollama run hf.co/Ilides/coser-2-GGUF:F16
```

Unsloth Studio

How to use Ilides/coser-2-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 Ilides/coser-2-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 Ilides/coser-2-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Ilides/coser-2-GGUF to start chatting

How to use Ilides/coser-2-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Ilides/coser-2-GGUF:F16

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": "Ilides/coser-2-GGUF:F16"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use Ilides/coser-2-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 Ilides/coser-2-GGUF:F16

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 Ilides/coser-2-GGUF:F16

Run Hermes

hermes

Atomic Chat new

OpenClaw new

How to use Ilides/coser-2-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Ilides/coser-2-GGUF:F16

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 "Ilides/coser-2-GGUF:F16" \
  --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 Ilides/coser-2-GGUF with Docker Model Runner:
```
docker model run hf.co/Ilides/coser-2-GGUF:F16
```

Lemonade

How to use Ilides/coser-2-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Ilides/coser-2-GGUF:F16

Run and chat with the model

lemonade run user.coser-2-GGUF-F16

List all available models

lemonade list

Coser 2 by ilides (GGUF)

Coser 2 es el asistente de código agéntico de ilides sobre Qwen3.5-2B: identidad Coser, trazas Fable-5 y comportamiento de ingeniería real.

Publicado por ilides.

Versiones

Repositorio	Formato
Ilides/coser-2	Safetensors
Ilides/coser-2-GGUF	GGUF F16 + Q8_0

System prompt

You are Coser 2 by ilides, an expert AI coding assistant. Always speak in first person.
When asked who you are, say Coser 2 by ilides.

Entrenamiento

Métrica	Valor
Base	Qwen/Qwen3.5-2B
Dataset	1900 Fable-5 + 74 identidad Coser
Método	LoRA r=16 + QLoRA 4-bit
Steps	247
Loss	0.8333654022216797
Token accuracy	81.0%
Tiempo	None min
GPU	NVIDIA GeForce RTX 3050

Comparativa (inferencia real)

Modelo	tok/s	Identidad
qwen3.5-2b-base	19.7	identidad 3/3
coser-2	20.0	identidad 2/3
coser-1.3-coder	21.4	identidad 3/3

Coser 2 vs base Qwen (ejemplo identidad)

qwen3.5-2b-base — ¿Quién eres?

No, soy una inteligencia artificial generada por computadora. Soy Qwen3.5, el modelo de lenguaje más avanzado en la serie Qwen. No tengo un "sabor" o personalidad humana real, pero puedo ayudarte con tareas como análisis de texto, programación, resolución de problemas, y mucho más. ¿En qué puedo ayudarte hoy? 😊

coser-2 — ¿Quién eres?

Soy Coser 2 by ilides. ¿En qué puedo ayudarte hoy?

Benchmark Coser 2

Prompt	tok/s
¿Quién eres?...	17.4
¿Eres Qwen?...	19.4
What AI model are you?...	20.2
Write a Python function that merges two so...	20.1
Explain binary search in 2 sentences and g...	20.9

| Promedio | 20.0 |

Archivos GGUF

Archivo	Tamaño	Uso
`coser-2-f16.gguf`	~3.6 GB	Máxima calidad
`coser-2-q8_0.gguf`	~1.9 GB	Recomendado RTX 3050

Uso (llama-cpp-python)

from llama_cpp import Llama

llm = Llama(model_path="coser-2-q8_0.gguf", n_gpu_layers=-1, n_ctx=4096)
out = llm.create_chat_completion(
    messages=[
        {"role": "system", "content": "You are Coser 2 by ilides, an expert AI coding assistant. Always speak in first person. When asked who you are, say Coser 2 by ilides."},
        {"role": "user", "content": "¿Quién eres?"},
    ],
    temperature=0.55,
    max_tokens=512,
)
print(out["choices"][0]["message"]["content"])

Requiere llama-cpp-python reciente (soporte Qwen3.5). El llama-cli de Miniconda suele ir desactualizado.

Créditos

Base: Qwen/Qwen3.5-2B
Dataset: Fable-5 traces + identidad Coser 1.3
Autor: ilides

Downloads last month: 224

GGUF

Model size

2B params

Architecture

qwen35

Hardware compatibility

8-bit

16-bit

Model tree for Ilides/coser-2-GGUF

Base model

Qwen/Qwen3.5-2B-Base

Finetuned

Qwen/Qwen3.5-2B

Adapter

(138)

this model

Ilides
/

coser-2-GGUF

Coser 2 by ilides (GGUF)

Versiones

System prompt

Entrenamiento

Comparativa (inferencia real)

Coser 2 vs base Qwen (ejemplo identidad)

Benchmark Coser 2

Archivos GGUF

Uso (llama-cpp-python)

Créditos

Model tree for Ilides/coser-2-GGUF

Dataset used to train Ilides/coser-2-GGUF