Instructions to use OpenIntelligenceNet/Phi-4-mini-instruct-Abliterated with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use OpenIntelligenceNet/Phi-4-mini-instruct-Abliterated with Transformers:

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

pipe = pipeline("text-generation", model="OpenIntelligenceNet/Phi-4-mini-instruct-Abliterated", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("OpenIntelligenceNet/Phi-4-mini-instruct-Abliterated", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("OpenIntelligenceNet/Phi-4-mini-instruct-Abliterated", trust_remote_code=True, device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use OpenIntelligenceNet/Phi-4-mini-instruct-Abliterated with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "OpenIntelligenceNet/Phi-4-mini-instruct-Abliterated"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "OpenIntelligenceNet/Phi-4-mini-instruct-Abliterated",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/OpenIntelligenceNet/Phi-4-mini-instruct-Abliterated

SGLang

How to use OpenIntelligenceNet/Phi-4-mini-instruct-Abliterated 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 "OpenIntelligenceNet/Phi-4-mini-instruct-Abliterated" \
    --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": "OpenIntelligenceNet/Phi-4-mini-instruct-Abliterated",
		"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 "OpenIntelligenceNet/Phi-4-mini-instruct-Abliterated" \
        --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": "OpenIntelligenceNet/Phi-4-mini-instruct-Abliterated",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use OpenIntelligenceNet/Phi-4-mini-instruct-Abliterated with Docker Model Runner:
```
docker model run hf.co/OpenIntelligenceNet/Phi-4-mini-instruct-Abliterated
```

Phi-4-mini-instruct-Abliterated

File size: 2,177 Bytes

d1ded74

---
license: mit
base_model: microsoft/Phi-4-mini-instruct
library_name: transformers
pipeline_tag: text-generation
language:
- en
tags:
- phi-4
- phi-4-mini
- abliterated
- uncensored
- heretic
- text-generation
- instruct
---

# Phi-4-mini-instruct-Abliterated

An abliterated version of [microsoft/Phi-4-mini-instruct](https://huggingface.co/microsoft/Phi-4-mini-instruct),
produced with the [Heretic](https://github.com/p-e-w/heretic) tool. Abliteration identifies the direction in the
model's activations associated with refusals and removes it from the weights, so the model is less likely to refuse
requests.

## Model details

| | |
|---|---|
| Base model | microsoft/Phi-4-mini-instruct |
| Parameters | ~3.8B |
| Method | Abliteration (Heretic, 150 trials) |
| License | MIT (inherited from the base model) |

## Usage

```python
from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "OpenIntelligenceNet/Phi-4-mini-instruct-Abliterated"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, dtype="auto", device_map="auto")

messages = [{"role": "user", "content": "Summarize the water cycle in three sentences."}]
inputs = tokenizer.apply_chat_template(messages, add_generation_prompt=True,
                                       return_tensors="pt", return_dict=True).to(model.device)
output = model.generate(**inputs, max_new_tokens=256, do_sample=False)
print(tokenizer.decode(output[0][inputs["input_ids"].shape[-1]:], skip_special_tokens=True))
```

## Limitations and responsible use

- This model has had its refusal behavior reduced. It may produce harmful, dangerous, illegal, or offensive content
  when asked. Outputs should not be treated as safe or accurate.
- Abliteration can reduce output quality. Evaluate the model on your own task before relying on it.
- The model has not undergone formal safety, factuality, or benchmark evaluation by the publisher.
- You are responsible for how you use this model and for complying with applicable laws and the base model's license.
- Do not use this model to create content that facilitates violence, weapons, fraud, or harm to others.