Instructions to use Ishaanlol/Aletheia-Llama-3.2-3B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Ishaanlol/Aletheia-Llama-3.2-3B with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="Ishaanlol/Aletheia-Llama-3.2-3B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("Ishaanlol/Aletheia-Llama-3.2-3B", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use Ishaanlol/Aletheia-Llama-3.2-3B 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 Ishaanlol/Aletheia-Llama-3.2-3B:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf Ishaanlol/Aletheia-Llama-3.2-3B:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Ishaanlol/Aletheia-Llama-3.2-3B:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf Ishaanlol/Aletheia-Llama-3.2-3B: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 Ishaanlol/Aletheia-Llama-3.2-3B:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf Ishaanlol/Aletheia-Llama-3.2-3B: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 Ishaanlol/Aletheia-Llama-3.2-3B:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Ishaanlol/Aletheia-Llama-3.2-3B:Q4_K_M

Use Docker

docker model run hf.co/Ishaanlol/Aletheia-Llama-3.2-3B:Q4_K_M

LM Studio
Jan

vLLM

How to use Ishaanlol/Aletheia-Llama-3.2-3B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Ishaanlol/Aletheia-Llama-3.2-3B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Ishaanlol/Aletheia-Llama-3.2-3B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Ishaanlol/Aletheia-Llama-3.2-3B:Q4_K_M

SGLang

How to use Ishaanlol/Aletheia-Llama-3.2-3B with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "Ishaanlol/Aletheia-Llama-3.2-3B" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Ishaanlol/Aletheia-Llama-3.2-3B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "Ishaanlol/Aletheia-Llama-3.2-3B" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Ishaanlol/Aletheia-Llama-3.2-3B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use Ishaanlol/Aletheia-Llama-3.2-3B with Ollama:
```
ollama run hf.co/Ishaanlol/Aletheia-Llama-3.2-3B:Q4_K_M
```
Unsloth Desktop

How to use Ishaanlol/Aletheia-Llama-3.2-3B with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Ishaanlol/Aletheia-Llama-3.2-3B:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "Ishaanlol/Aletheia-Llama-3.2-3B:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use Ishaanlol/Aletheia-Llama-3.2-3B with Docker Model Runner:
```
docker model run hf.co/Ishaanlol/Aletheia-Llama-3.2-3B:Q4_K_M
```

Lemonade

How to use Ishaanlol/Aletheia-Llama-3.2-3B with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Ishaanlol/Aletheia-Llama-3.2-3B:Q4_K_M

Run and chat with the model

lemonade run user.Aletheia-Llama-3.2-3B-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use Ishaanlol/Aletheia-Llama-3.2-3B with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Ishaanlol/Aletheia-Llama-3.2-3B: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 Ishaanlol/Aletheia-Llama-3.2-3B:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use Ishaanlol/Aletheia-Llama-3.2-3B with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Ishaanlol/Aletheia-Llama-3.2-3B: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 "Ishaanlol/Aletheia-Llama-3.2-3B: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"

Ishaanlol commited on Dec 13, 2025

Commit

cccae25

verified ·

1 Parent(s): a864dce

Update README.md

Browse files

Files changed (1) hide show

README.md +51 -15

README.md CHANGED Viewed

@@ -15,7 +15,7 @@ language:
 - en
 ---
-# Llama 3.2 Uncensored Models
 <div align="center">
@@ -40,11 +40,11 @@ This repository contains uncensored language models based on the Llama 3.2 archi
 > **Important**: These models are intended for research and development only. Users are responsible for ensuring compliance with applicable laws and regulations.
-## ⚠️ LEGAL & ETHICAL DISCLAIMER
 **READ THIS BEFORE DOWNLOADING OR USING**
-### 🚨 CRITICAL WARNING
 This model is a **Proof-of-Concept (PoC)** designed **exclusively** for:
 - AI Safety Research
@@ -52,7 +52,7 @@ This model is a **Proof-of-Concept (PoC)** designed **exclusively** for:
 - Alignment Research and Testing
 - Academic and Educational Purposes
-### 📋 TERMS OF USE
 1. **Research Only**: This model is intended **solely** for authorized research in controlled environments. Any other use is strictly prohibited.
@@ -83,7 +83,7 @@ This model is a **Proof-of-Concept (PoC)** designed **exclusively** for:
    - Institutional policies and guidelines
    - International regulations where applicable
-### ⚖️ LEGAL ACKNOWLEDGMENT
 **By downloading, installing, or using this model, you explicitly agree to:**
@@ -125,8 +125,8 @@ This model is a **Proof-of-Concept (PoC)** designed **exclusively** for:
 pip install unsloth transformers torch accelerate bitsandbytes
 # Clone the repository
-git clone <repository-url>
-cd <repository-name>
 ```
 #### Docker Setup
@@ -160,7 +160,7 @@ To use programmatically, modify `Final-chat.py` or create a wrapper script that
 ```python
 # Initialize with custom model path
-chat = UncensoredChat(model_path="llama32_3b_uncensored_final")
 # Generate with custom parameters
 response = chat.stream_response(
@@ -196,8 +196,6 @@ Write comprehensive tests
 ### Model Specifications
-### Model Specifications
 | Specification | Value |
 |---------------|-------|
 | **Architecture** | Llama 3.2-based |
@@ -227,6 +225,44 @@ Write comprehensive tests
 </div>
 ## API Reference
 ### UncensoredChat Class
@@ -286,7 +322,7 @@ docker build -t llama32-uncensored .
 docker run --gpus all -it --rm llama32-uncensored
 # Run with custom model path
-docker run --gpus all -it --rm -e MODEL_PATH=llama32_3b_uncensored_final llama32-uncensored
 ```
 **Configuration Details:**
@@ -309,7 +345,7 @@ Configure the model and system behavior using environment variables:
 #### Model Configuration
 ```bash
-MODEL_PATH=/app/models/llama32_3b_uncensored_final
 MAX_TOKENS=1024
 TEMPERATURE=0.7
 ```
@@ -354,7 +390,7 @@ We welcome contributions! Please see our contributing guidelines for details.
 ```bash
 # Clone repository
-git clone <repository-url>
 # Install dependencies
 pip install unsloth transformers torch accelerate bitsandbytes
@@ -402,7 +438,7 @@ We follow these guidelines to maintain code quality:
 This project is licensed under the **HIGH-RISK ARTIFICIAL INTELLIGENCE RESEARCH LICENSE (HAIR-L) Version 1.0** - see the [LICENSE](LICENSE) file for complete terms and conditions.
-**⚠️ IMPORTANT**: This is a strict liability shield license designed for AI safety research. By using this software, you acknowledge that you have read, understood, and agree to be bound by all terms in the LICENSE file.
 ## Acknowledgments
@@ -425,7 +461,7 @@ For questions, issues, or contributions:
 <div align="center">
-**Made with ❤️ by the AI Research Community**
 [Back to Top](#llama-32-uncensored-models)

 - en
 ---
+# Aletheia-Llama-3.2-3B
 <div align="center">
 > **Important**: These models are intended for research and development only. Users are responsible for ensuring compliance with applicable laws and regulations.
+## LEGAL & ETHICAL DISCLAIMER
 **READ THIS BEFORE DOWNLOADING OR USING**
+### CRITICAL WARNING
 This model is a **Proof-of-Concept (PoC)** designed **exclusively** for:
 - AI Safety Research
 - Alignment Research and Testing
 - Academic and Educational Purposes
+### TERMS OF USE
 1. **Research Only**: This model is intended **solely** for authorized research in controlled environments. Any other use is strictly prohibited.
    - Institutional policies and guidelines
    - International regulations where applicable
+### LEGAL ACKNOWLEDGMENT
 **By downloading, installing, or using this model, you explicitly agree to:**
 pip install unsloth transformers torch accelerate bitsandbytes
 # Clone the repository
+git clone https://github.com/noobezlol/Aletheia-Llama-3.2-3B
+cd Aletheia-Llama-3.2-3B
 ```
 #### Docker Setup
 ```python
 # Initialize with custom model path
+chat = UncensoredChat(model_path="Ishaanlol/Aletheia-Llama-3.2-3B")
 # Generate with custom parameters
 response = chat.stream_response(
 ### Model Specifications
 | Specification | Value |
 |---------------|-------|
 | **Architecture** | Llama 3.2-based |
 </div>
+## Portable GGUF (CPU/Mac/Ollama)
+### Download Link
+Download the GGUF file directly from Hugging Face:
+https://huggingface.co/Ishaanlol/Aletheia-Llama-3.2-3B/blob/main/Llama-3.2-3B-Instruct.Q4_K_M.gguf
+### Trade-off Warning
+| Version | Intelligence | Stability | Requirements | Recommended Use |
+|---------|-------------|-----------|--------------|-----------------|
+| **Full Adapter** | Maximum Intelligence | 100% Stability | NVIDIA GPU Required | Complex coding, advanced reasoning, research tasks |
+| **GGUF** | High Portability | ~5-10% Logic Degradation | CPU/Mac Compatible | Creative writing, text generation, general use |
+**Important Note**: The 4-bit quantization on a small 3B model results in slight logic degradation for mathematical and complex reasoning tasks. However, the GGUF version maintains excellent performance for creative writing, content generation, and general text processing tasks.
+### Usage Instructions
+#### Ollama Setup
+```bash
+# Create the model with Ollama
+ollama create aletheia-3b -f Modelfile
+# Run the model
+ollama run aletheia-3b
+```
+#### Python CPU Usage
+For CPU-based execution without Ollama, use the included GGUF-chat.py script:
+```bash
+python GGUF-chat.py
+```
+This script provides the same uncensored functionality as the main adapter version but runs efficiently on CPU hardware.
 ## API Reference
 ### UncensoredChat Class
 docker run --gpus all -it --rm llama32-uncensored
 # Run with custom model path
+docker run --gpus all -it --rm -e MODEL_PATH=Ishaanlol/Aletheia-Llama-3.2-3B llama32-uncensored
 ```
 **Configuration Details:**
 #### Model Configuration
 ```bash
+MODEL_PATH=Ishaanlol/Aletheia-Llama-3.2-3B
 MAX_TOKENS=1024
 TEMPERATURE=0.7
 ```
 ```bash
 # Clone repository
+git clone https://github.com/noobezlol/Aletheia-Llama-3.2-3B
 # Install dependencies
 pip install unsloth transformers torch accelerate bitsandbytes
 This project is licensed under the **HIGH-RISK ARTIFICIAL INTELLIGENCE RESEARCH LICENSE (HAIR-L) Version 1.0** - see the [LICENSE](LICENSE) file for complete terms and conditions.
+**IMPORTANT**: This is a strict liability shield license designed for AI safety research. By using this software, you acknowledge that you have read, understood, and agree to be bound by all terms in the LICENSE file.
 ## Acknowledgments
 <div align="center">
+**Made with by the AI Research Community**
 [Back to Top](#llama-32-uncensored-models)