Instructions to use microsoft/Phi-3-mini-128k-instruct-onnx with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use microsoft/Phi-3-mini-128k-instruct-onnx with Transformers:

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

pipe = pipeline("text-generation", model="microsoft/Phi-3-mini-128k-instruct-onnx", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("microsoft/Phi-3-mini-128k-instruct-onnx", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("microsoft/Phi-3-mini-128k-instruct-onnx", trust_remote_code=True)

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use microsoft/Phi-3-mini-128k-instruct-onnx with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "microsoft/Phi-3-mini-128k-instruct-onnx"
# 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-3-mini-128k-instruct-onnx",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/microsoft/Phi-3-mini-128k-instruct-onnx

SGLang

How to use microsoft/Phi-3-mini-128k-instruct-onnx 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-3-mini-128k-instruct-onnx" \
    --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-3-mini-128k-instruct-onnx",
		"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-3-mini-128k-instruct-onnx" \
        --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-3-mini-128k-instruct-onnx",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use microsoft/Phi-3-mini-128k-instruct-onnx with Docker Model Runner:
```
docker model run hf.co/microsoft/Phi-3-mini-128k-instruct-onnx
```

Could you tell me how to transfer a Phi-3 model (safetensors) to its onnx?

by AllenChai - opened May 20, 2024

Discussion

AllenChai

May 20, 2024

Hi dear!
I am a researcher from UW. And I have a fine-tuned model from Phi-3-mini-128k-instruct. I just wonder how to transfer it to onnx?
Your help is very essential to me.
Thanks!

kvaishnavi

Microsoft org May 21, 2024

You can use ONNX Runtime GenAI's model builder to quickly convert your fine-tuned Phi-3-mini-128k-instruct model to optimized and quantized ONNX models. This example should work for your scenario.

AllenChai

May 28, 2024

Hi @kvaishnavi Thank you for your help!
What about more than two safetensors? How to convert many safetensors split from one large model to a final onnx file?

Thanks!

kvaishnavi

Microsoft org May 28, 2024

If your fine-tuned model can be loaded with Hugging Face's AutoModelForCausalLM.from_pretrained method, then the model builder can produce the final ONNX model from any number of .safetensors files.

AllenChai

May 28, 2024

It does work!
Thank you very much for your soon feedback!

AllenChai changed discussion status to closed May 28, 2024

AllenChai

May 28, 2024

Hi dear,
BTW, I wonder in real application, which one should I use? model.onnx or model.onnx.data? What's the difference between them?

AllenChai changed discussion status to open May 28, 2024

AllenChai

May 28, 2024

I get it!
Thank you again!

AllenChai changed discussion status to closed May 28, 2024

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment