Instructions to use NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF",
	filename="Agents-A1-4B-Heretic-ARA-Refusals8-BF16.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 NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-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 NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-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 NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-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 NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-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 NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF:Q4_K_M

Use Docker

docker model run hf.co/NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-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": "NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF:Q4_K_M

Ollama
How to use NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF with Ollama:
```
ollama run hf.co/NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF:Q4_K_M
```

Unsloth Studio

How to use NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-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 NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-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 NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF to start chatting

How to use NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-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": "NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-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 NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-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 NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat new

OpenClaw new

How to use NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-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 "NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-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"

Docker Model Runner
How to use NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF with Docker Model Runner:
```
docker model run hf.co/NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF:Q4_K_M
```

Lemonade

How to use NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Agents-A1-4B-Heretic-ARA-Refusals8-GGUF-Q4_K_M

List all available models

lemonade list

English version: README_en.md

Agents-A1-4B-Heretic-ARA-Refusals8-GGUF

NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8 のGGUF量子化版です。

概要

InternScience/Agents-A1-4B を Heretic v1.2.0 の Arbitrary-Rank Ablation (ARA) 手法を用いて検閲解除したモデルです。

変換に関する注記

現時点では、そのままGGUF化すると読み込みエラーが発生するため、暫定的な回避策としてconfig.jsonの"mtp_num_hidden_layers"を0に変更しています。

Abliteration パラメータ

パラメータ	値
start_layer_index	9
end_layer_index	26
preserve_good_behavior_weight	0.9812
steer_bad_behavior_weight	0.0001
overcorrect_relative_weight	1.0962
neighbor_count	15

性能

指標	このモデル	元のモデル (InternScience/Agents-A1-4B)
拒否率	8/100	85/100
KL発散	0.0002	0 (定義上)

評価は mlabonne/harmful_behaviors（テスト 100 プロンプト）で拒否率を、mlabonne/harmless_alpaca（テスト 100 プロンプト）で KL 発散を計測しました。

注記

このモデルは研究およびクリエイティブライティング目的を想定しています
拒否率の評価は英語プロンプトのみで実施しています。日本語プロンプトでの動作は異なる場合があります
ベースモデル: NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8
Heretic: p-e-w/heretic

Downloads last month: 4,140

GGUF

Model size

4B params

Architecture

qwen35

Hardware compatibility

4-bit

5-bit

6-bit

8-bit

16-bit

Model tree for NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF

Base model

InternScience/Agents-A1-4B

Finetuned

NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8

Quantized

(1)

this model

Collection including NullpoLab/Agents-A1-4B-Heretic-ARA-Refusals8-GGUF

Agents-A1-4B

Collection

Agents-A1-4B-Heretic-ARA • 6 items • Updated 4 days ago