Instructions to use xXGioXx/sad_core_v1_q8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use xXGioXx/sad_core_v1_q8 with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="xXGioXx/sad_core_v1_q8",
	filename="sad_core_v1_q8.gguf",
)

llm.create_chat_completion(
	messages = "No input example has been defined for this model task."
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use xXGioXx/sad_core_v1_q8 with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf xXGioXx/sad_core_v1_q8
# Run inference directly in the terminal:
llama-cli -hf xXGioXx/sad_core_v1_q8

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf xXGioXx/sad_core_v1_q8
# Run inference directly in the terminal:
llama-cli -hf xXGioXx/sad_core_v1_q8

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 xXGioXx/sad_core_v1_q8
# Run inference directly in the terminal:
./llama-cli -hf xXGioXx/sad_core_v1_q8

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 xXGioXx/sad_core_v1_q8
# Run inference directly in the terminal:
./build/bin/llama-cli -hf xXGioXx/sad_core_v1_q8

Use Docker

docker model run hf.co/xXGioXx/sad_core_v1_q8

LM Studio
Jan
Ollama
How to use xXGioXx/sad_core_v1_q8 with Ollama:
```
ollama run hf.co/xXGioXx/sad_core_v1_q8
```

Unsloth Studio new

How to use xXGioXx/sad_core_v1_q8 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 xXGioXx/sad_core_v1_q8 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 xXGioXx/sad_core_v1_q8 to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for xXGioXx/sad_core_v1_q8 to start chatting

Pi new

How to use xXGioXx/sad_core_v1_q8 with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf xXGioXx/sad_core_v1_q8

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": "xXGioXx/sad_core_v1_q8"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use xXGioXx/sad_core_v1_q8 with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf xXGioXx/sad_core_v1_q8

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 xXGioXx/sad_core_v1_q8

Run Hermes

hermes

Docker Model Runner
How to use xXGioXx/sad_core_v1_q8 with Docker Model Runner:
```
docker model run hf.co/xXGioXx/sad_core_v1_q8
```

Lemonade

How to use xXGioXx/sad_core_v1_q8 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull xXGioXx/sad_core_v1_q8

Run and chat with the model

lemonade run user.sad_core_v1_q8-{{QUANT_TAG}}

List all available models

lemonade list

🚀 Descripción del Proyecto

SAD_CORE v1 es un Small Language Model (SLM) diseñado para ejecutarse en entornos de alta restricción de recursos, específicamente hardware móvil. Este modelo es el resultado de una investigación sobre la viabilidad de la inteligencia artificial local (Edge AI), permitiendo asistencia en programación y seguimiento de instrucciones sin conexión a internet.

Características Principales:

Optimizado para Móviles: Ejecución fluida en dispositivos con 4GB de RAM.
Especialización en Código: Refinado para mejorar la asistencia en Python y lógica de agentes.
Multilingüe: Ajustado para una respuesta coherente tanto en español como en inglés.

🛠️ Detalles Técnicos del Fine-Tuning

El entrenamiento se realizó bajo la metodología SFT (Supervised Fine-Tuning) con las siguientes especificaciones:

Framework: Unsloth (FastLanguageModel)
Adaptador: LoRA (Rank: 64, Alpha: 64)
Longitud de Secuencia (Max Seq): 2048 tokens.
Dataset Total: 10,000 muestras curadas de fuentes seleccionadas.
Optimizador: AdamW (8-bit) para eficiencia de VRAM.
Hardware de Entrenamiento: NVIDIA Tesla T4 (Cloud).

📱 Cómo usar en Android (Termux)

Este modelo está en formato GGUF y es compatible con llama.cpp.

Descarga el modelo:

wget https://huggingface.co/xXGioXx/sad_core_v1_q8/resolve/main/sad_core_v1_q8.gguf

📊 Rendimiento Esperado (Samsung A52 - 4GB RAM)

Velocidad de Inferencia: ~4-6 tokens/segundo.
Uso de RAM: ~1.8GB (Modelo) + sistema.
Temperatura: Estable con 3 hilos de ejecución.

Downloads last month: 9

GGUF

Model size

2B params

Architecture

qwen2

Hardware compatibility

We're not able to determine the quantization variants.

View all variants

Inference Providers NEW

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

Model tree for xXGioXx/sad_core_v1_q8

Base model

Qwen/Qwen2.5-1.5B

Finetuned

Qwen/Qwen2.5-Coder-1.5B

Finetuned

Qwen/Qwen2.5-Coder-1.5B-Instruct

Finetuned

unsloth/Qwen2.5-Coder-1.5B-Instruct

Quantized

(23)

this model

xXGioXx
/

sad_core_v1_q8

🚀 Descripción del Proyecto

Características Principales:

🛠️ Detalles Técnicos del Fine-Tuning

📱 Cómo usar en Android (Termux)

📊 Rendimiento Esperado (Samsung A52 - 4GB RAM)

Model tree for xXGioXx/sad_core_v1_q8

Datasets used to train xXGioXx/sad_core_v1_q8