Instructions to use AryavA/Phi-4-30B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AryavA/Phi-4-30B with Transformers:

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

pipe = pipeline("text-generation", model="AryavA/Phi-4-30B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("AryavA/Phi-4-30B")
model = AutoModelForCausalLM.from_pretrained("AryavA/Phi-4-30B")
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]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use AryavA/Phi-4-30B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "AryavA/Phi-4-30B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "AryavA/Phi-4-30B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/AryavA/Phi-4-30B

SGLang

How to use AryavA/Phi-4-30B 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 "AryavA/Phi-4-30B" \
    --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": "AryavA/Phi-4-30B",
		"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 "AryavA/Phi-4-30B" \
        --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": "AryavA/Phi-4-30B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use AryavA/Phi-4-30B with Docker Model Runner:
```
docker model run hf.co/AryavA/Phi-4-30B
```

Phi 4 30B — An "added layers" version of Phi-4 15B, without any knowledge loss

Info

I have added 40 more layers to the phi-4 15B, with 0 parameter initialization.

Hardware

An RTX 6000 Blackwell WS, with 96GB of VRAM, and the whole task was done within 20 minutes

Code

Can be found here in my repository.

Github Repo: ag-aryav/Phi-4-30B

Downloads last month: 4

Safetensors

Model size

28B params

Tensor type

BF16

Model tree for AryavA/Phi-4-30B

Base model

microsoft/phi-4

Finetuned

(100)

this model