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

Libraries

How to use microsoft/Phi-tiny-MoE-instruct with Transformers:

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

pipe = pipeline("text-generation", model="microsoft/Phi-tiny-MoE-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-tiny-MoE-instruct", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("microsoft/Phi-tiny-MoE-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

vLLM

How to use microsoft/Phi-tiny-MoE-instruct with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/microsoft/Phi-tiny-MoE-instruct

SGLang

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

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

Add library_name and link to code

by nielsr HF Staff - opened Jun 29, 2025

base: refs/heads/main

←

from: refs/pr/1

Discussion Files changed

-3

Files changed (1) hide show

README.md +6 -3

README.md CHANGED Viewed

@@ -1,15 +1,18 @@
 ---
-license: mit
 language:
 - en
 context_length:
 - 4k
-pipeline_tag: text-generation
 ---
 ## Model Summary
 Phi-tiny-MoE is a lightweight Mixture of Experts (MoE) model with 3.8B total parameters and 1.1B activated parameters. It is compressed and distilled from the base model shared by [Phi-3.5-MoE](https://huggingface.co/microsoft/Phi-3.5-MoE-instruct) and [GRIN-MoE](https://huggingface.co/microsoft/GRIN-MoE) using the [SlimMoE](https://arxiv.org/pdf/2506.18349) approach, then post-trained via supervised fine-tuning and direct preference optimization for instruction following and safety. The model is trained on Phi-3 synthetic data and filtered public documents, with a focus on high-quality, reasoning-dense content. It is part of the SlimMoE series, which includes a larger variant, [Phi-mini-MoE](https://huggingface.co/microsoft/Phi-mini-MoE-instruct), with 7.6B total and 2.4B activated parameters.
 References: <br>
 📖 [SlimMoE Paper](https://arxiv.org/pdf/2506.18349) <br>
@@ -165,4 +168,4 @@ Note that by default, the Phi-tiny-MoE model uses flash attention, which require
 The model is licensed under the [MIT license](./LICENSE).
 ## Trademarks
-This project may contain trademarks or logos for projects, products, or services. Authorized use of Microsoft trademarks or logos is subject to and must follow [Microsoft’s Trademark & Brand Guidelines](https://www.microsoft.com/en-us/legal/intellectualproperty/trademarks). Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship. Any use of third-party trademarks or logos are subject to those third-party’s policies.

 ---
 language:
 - en
+license: mit
+pipeline_tag: text-generation
 context_length:
 - 4k
+library_name: transformers
 ---
 ## Model Summary
 Phi-tiny-MoE is a lightweight Mixture of Experts (MoE) model with 3.8B total parameters and 1.1B activated parameters. It is compressed and distilled from the base model shared by [Phi-3.5-MoE](https://huggingface.co/microsoft/Phi-3.5-MoE-instruct) and [GRIN-MoE](https://huggingface.co/microsoft/GRIN-MoE) using the [SlimMoE](https://arxiv.org/pdf/2506.18349) approach, then post-trained via supervised fine-tuning and direct preference optimization for instruction following and safety. The model is trained on Phi-3 synthetic data and filtered public documents, with a focus on high-quality, reasoning-dense content. It is part of the SlimMoE series, which includes a larger variant, [Phi-mini-MoE](https://huggingface.co/microsoft/Phi-mini-MoE-instruct), with 7.6B total and 2.4B activated parameters.
+The code can be found at https://github.com/microsoft/MoE-compression.
 References: <br>
 📖 [SlimMoE Paper](https://arxiv.org/pdf/2506.18349) <br>
 The model is licensed under the [MIT license](./LICENSE).
 ## Trademarks
+This project may contain trademarks or logos for projects, products, or services. Authorized use of Microsoft trademarks or logos is subject to and must follow [Microsoft’s Trademark & Brand Guidelines](https://www.microsoft.com/en-us/legal/intellectualproperty/trademarks). Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship. Any use of third-party trademarks or logos are subject to those third-party’s policies.