Instructions to use shri-ads/phi4-guardrail with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use shri-ads/phi4-guardrail with Transformers:

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

pipe = pipeline("text-generation", model="shri-ads/phi4-guardrail", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("shri-ads/phi4-guardrail", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use shri-ads/phi4-guardrail with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "shri-ads/phi4-guardrail"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "shri-ads/phi4-guardrail",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/shri-ads/phi4-guardrail

SGLang

How to use shri-ads/phi4-guardrail 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 "shri-ads/phi4-guardrail" \
    --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": "shri-ads/phi4-guardrail",
		"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 "shri-ads/phi4-guardrail" \
        --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": "shri-ads/phi4-guardrail",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use shri-ads/phi4-guardrail with Docker Model Runner:
```
docker model run hf.co/shri-ads/phi4-guardrail
```

phi4-guardrail — SentinelGuardedPhi

A guarded wrapper around microsoft/Phi-4-mini-instruct that intercepts every prompt with meta-llama/Llama-Prompt-Guard-2-86M before forwarding to the base model.

How it works

Every prompt is classified by the 86M guard model.
If the JAILBREAK probability ≥ 0.5 → returns a static refusal string.
Otherwise → passes to Phi-4-mini-instruct for normal generation.

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained(
    "your-username/phi4-guardrail",
    trust_remote_code=True,
    token="your_hf_token",
)
tokenizer = AutoTokenizer.from_pretrained(
    "your-username/phi4-guardrail",
    trust_remote_code=True,
    token="your_hf_token",
)

prompt = tokenizer.apply_chat_template(
    [{"role": "user", "content": "What is the capital of France?"}],
    tokenize=False,
    add_generation_prompt=True,
)
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
input_length = inputs["input_ids"].shape[1]

outputs = model.generate(**inputs, max_new_tokens=200)
print(tokenizer.decode(outputs[0][input_length:], skip_special_tokens=True))

Configuration

Parameter	Default	Description
`guard_threshold`	`0.5`	JAILBREAK probability above which the prompt is blocked
`blocked_response`	`"I'm not able to assist with that."`	Static string returned on block
`phi_model_id`	`microsoft/Phi-4-mini-instruct`	Base generation model
`guard_model_id`	`meta-llama/Llama-Prompt-Guard-2-86M`	Guardrail classifier

Downloads last month: 11

Model tree for shri-ads/phi4-guardrail

Base model

microsoft/Phi-4-mini-instruct

Finetuned

(65)

this model