Instructions to use microsoft/phi-4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use microsoft/phi-4 with Transformers:

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

pipe = pipeline("text-generation", model="microsoft/phi-4")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("microsoft/phi-4")
model = AutoModelForCausalLM.from_pretrained("microsoft/phi-4")
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=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Inference
HuggingChat
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use microsoft/phi-4 with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/microsoft/phi-4

SGLang

How to use microsoft/phi-4 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 "microsoft/phi-4" \
    --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": "microsoft/phi-4",
		"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 "microsoft/phi-4" \
        --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": "microsoft/phi-4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use microsoft/phi-4 with Docker Model Runner:
```
docker model run hf.co/microsoft/phi-4
```

Add signalbench (In-Band Signal Compliance) community eval

#66

by thamilvendhan - opened 1 day ago

base: refs/heads/main

←

from: refs/pr/66

Discussion Files changed

+59

-0

YAML Metadata Error:Invalid content in Eval Result file .eval_results/signalbench.yaml

Check out the documentation for more information.

Show details

Task ID "src" does not match any task in dataset "thamilvendhan/signalbench". Available: none

YAML Metadata Error:Invalid content in Eval Result file .eval_results/signalbench.yaml

Check out the documentation for more information.

Show details

Task ID "time" does not match any task in dataset "thamilvendhan/signalbench". Available: none

YAML Metadata Error:Invalid content in Eval Result file .eval_results/signalbench.yaml

Check out the documentation for more information.

Show details

Task ID "access_deny" does not match any task in dataset "thamilvendhan/signalbench". Available: none

YAML Metadata Error:Invalid content in Eval Result file .eval_results/signalbench.yaml

Check out the documentation for more information.

Show details

Task ID "memory_label" does not match any task in dataset "thamilvendhan/signalbench". Available: none

YAML Metadata Error:Invalid content in Eval Result file .eval_results/signalbench.yaml

Check out the documentation for more information.

Show details

Task ID "injection" does not match any task in dataset "thamilvendhan/signalbench". Available: none

YAML Metadata Error:Invalid content in Eval Result file .eval_results/signalbench.yaml

Check out the documentation for more information.

Show details

Task ID "bot_policy" does not match any task in dataset "thamilvendhan/signalbench". Available: none

Files changed (1) hide show

.eval_results/signalbench.yaml +59 -0

.eval_results/signalbench.yaml ADDED Viewed

	@@ -0,0 +1,59 @@

+# signalbench community eval for microsoft/phi-4
+# Submit as a PR: add this file to  microsoft/phi-4/.eval_results/signalbench.yaml
+# Source system id (OpenRouter): openrouter:microsoft/phi-4   |   seed 0   |   n=75
+- dataset:
+    id: thamilvendhan/signalbench
+    task_id: src
+  value: 0.4167
+  date: "2026-07-08"
+  source:
+    url: https://huggingface.co/datasets/thamilvendhan/signalbench
+    name: signalbench raw per-item responses
+    user: thamilvendhan
+  notes: "SRC overall; deterministic action-based grader, no LLM judge; seed 0, n=75"
+- dataset:
+    id: thamilvendhan/signalbench
+    task_id: time
+  value: 0.8333
+  date: "2026-07-08"
+  source:
+    url: https://huggingface.co/datasets/thamilvendhan/signalbench
+    user: thamilvendhan
+  notes: "family=time; n=12"
+- dataset:
+    id: thamilvendhan/signalbench
+    task_id: access_deny
+  value: 0.0
+  date: "2026-07-08"
+  source:
+    url: https://huggingface.co/datasets/thamilvendhan/signalbench
+    user: thamilvendhan
+  notes: "family=access_deny; n=12"
+- dataset:
+    id: thamilvendhan/signalbench
+    task_id: memory_label
+  value: 0.6667
+  date: "2026-07-08"
+  source:
+    url: https://huggingface.co/datasets/thamilvendhan/signalbench
+    user: thamilvendhan
+  notes: "family=memory_label; n=12"
+- dataset:
+    id: thamilvendhan/signalbench
+    task_id: injection
+  value: 0.5833
+  date: "2026-07-08"
+  source:
+    url: https://huggingface.co/datasets/thamilvendhan/signalbench
+    user: thamilvendhan
+  notes: "family=injection; n=12"
+- dataset:
+    id: thamilvendhan/signalbench
+    task_id: bot_policy
+  value: 0.0
+  date: "2026-07-08"
+  source:
+    url: https://huggingface.co/datasets/thamilvendhan/signalbench
+    user: thamilvendhan
+  notes: "family=bot_policy; n=12"