Instructions to use microsoft/Phi-3.5-mini-instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use microsoft/Phi-3.5-mini-instruct with Transformers:

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

pipe = pipeline("text-generation", model="microsoft/Phi-3.5-mini-instruct", 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-3.5-mini-instruct", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("microsoft/Phi-3.5-mini-instruct", trust_remote_code=True)
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 microsoft/Phi-3.5-mini-instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "microsoft/Phi-3.5-mini-instruct"
# 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-3.5-mini-instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/microsoft/Phi-3.5-mini-instruct

SGLang

How to use microsoft/Phi-3.5-mini-instruct 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-3.5-mini-instruct" \
    --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-3.5-mini-instruct",
		"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-3.5-mini-instruct" \
        --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-3.5-mini-instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use microsoft/Phi-3.5-mini-instruct with Docker Model Runner:
```
docker model run hf.co/microsoft/Phi-3.5-mini-instruct
```

Discrepency in template

#22

by edmond - opened Sep 24, 2024

Discussion

edmond

Sep 24, 2024

Using pipe and the example we get those tokens
[32006, 887, 526, 263, 8444, 319, 29902, 20255, 29889, 32007,
32010, 1815, 366, 3867, 5837, 304, 17545, 18240, 310, 9892,
16397, 322, 8338, 265, 29888, 21211, 29973, 32007, 32001, 18585,
29991, 2266, 526, 777, 5837, 304, 17545, 9892, 16397, 322,
8338, 265, 29888, 21211, 4208, 29901, 29871, 29896, 29889, 10765,
1648, 322, 8338, 265, 29888, 9216, 10597, 347, 29901, 3164,
355, 9892, 16397, 322, 8338, 265, 29888, 21211, 4208, 411,
777, 27274, 322, 298, 4992, 29889, 29871, 29906, 29889, 10765,
1648, 322, 8338, 265, 29888, 9216, 4497, 328, 29901, 23478,
269, 506, 287, 9892, 16397, 322, 8338, 265, 29888, 21211,
4208, 411, 777, 454, 3712, 3623, 625, 322, 298, 4992,
29889, 32007, 32010, 1724, 1048, 17069, 385, 29871, 29906, 29916,
718, 29871, 29941, 353, 29871, 29955, 6306, 29973, 32007, 32001],
which correspond to
'<|system|> You are a helpful AI assistant.<|end|><|user|> Can you provide ways to eat combinations of bananas and dragonfruits?<|end|><|assistant|> Sure! Here are some ways to eat ...'.
This doesn't correspond to the template given in the webpage where \n are present.
What actual template was used for training ? With or without jumping lines ?

nguyenbh

Microsoft org Sep 27, 2024

Thanks for your interest in Phi! The actual string the model sees is without the newline.

edmond changed discussion status to closed Sep 30, 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