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

Libraries

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

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

pipe = pipeline("text-generation", model="microsoft/Phi-3.5-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-3.5-MoE-instruct", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("microsoft/Phi-3.5-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-3.5-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-3.5-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-3.5-MoE-instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

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

SGLang

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

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

The provided example doesn't work

#32

by kqsong - opened Aug 23, 2024

Discussion

kqsong

Aug 23, 2024

Tried to use the provided example, however, it is not work with "text-generate".

Any suggestions?

ykim362

Microsoft org Aug 23, 2024

Hi. Could you share some more details or error messages you see?

kqsong

Aug 23, 2024

Hi. Could you share some more details or error messages you see?

Error message:

The model 'PhiMoEForCausalLM' is not supported for text-generation. Supported models are ['BartForCausalLM', 'BertLMHeadModel', 'BertGenerationDecoder', 'BigBirdForCausalLM', 'BigBirdPegasusForCausalLM', 'BioGp
tForCausalLM', 'BlenderbotForCausalLM', 'BlenderbotSmallForCausalLM', 'BloomForCausalLM', 'CamembertForCausalLM', 'LlamaForCausalLM', 'CodeGenForCausalLM', 'CohereForCausalLM', 'CpmAntForCausalLM', 'CTRLLMHeadM
odel', 'Data2VecTextForCausalLM', 'DbrxForCausalLM', 'ElectraForCausalLM', 'ErnieForCausalLM', 'FalconForCausalLM', 'FuyuForCausalLM', 'GemmaForCausalLM', 'Gemma2ForCausalLM', 'GitForCausalLM', 'GPT2LMHeadModel
', 'GPT2LMHeadModel', 'GPTBigCodeForCausalLM', 'GPTNeoForCausalLM', 'GPTNeoXForCausalLM', 'GPTNeoXJapaneseForCausalLM', 'GPTJForCausalLM', 'JambaForCausalLM', 'JetMoeForCausalLM', 'LlamaForCausalLM', 'MambaForC
ausalLM', 'MarianForCausalLM', 'MBartForCausalLM', 'MegaForCausalLM', 'MegatronBertForCausalLM', 'MistralForCausalLM', 'MixtralForCausalLM', 'MptForCausalLM', 'MusicgenForCausalLM', 'MusicgenMelodyForCausalLM',
 'MvpForCausalLM', 'OlmoForCausalLM', 'OpenLlamaForCausalLM', 'OpenAIGPTLMHeadModel', 'OPTForCausalLM', 'PegasusForCausalLM', 'PersimmonForCausalLM', 'PhiForCausalLM', 'Phi3ForCausalLM', 'PLBartForCausalLM', 'P
rophetNetForCausalLM', 'QDQBertLMHeadModel', 'Qwen2ForCausalLM', 'Qwen2MoeForCausalLM', 'RecurrentGemmaForCausalLM', 'ReformerModelWithLMHead', 'RemBertForCausalLM', 'RobertaForCausalLM', 'RobertaPreLayerNormFo
rCausalLM', 'RoCBertForCausalLM', 'RoFormerForCausalLM', 'RwkvForCausalLM', 'Speech2Text2ForCausalLM', 'StableLmForCausalLM', 'Starcoder2ForCausalLM', 'TransfoXLLMHeadModel', 'TrOCRForCausalLM', 'WhisperForCaus
alLM', 'XGLMForCausalLM', 'XLMWithLMHeadModel', 'XLMProphetNetForCausalLM', 'XLMRobertaForCausalLM', 'XLMRobertaXLForCausalLM', 'XLNetLMHeadModel', 'XmodForCausalLM'].

I checked the huggingface, there's no official code for the model, but the authors added them to the checkpoint repo.

ykim362

Microsoft org Aug 23, 2024

•

edited Aug 23, 2024

Hi. Yes, we are working on adding our new model to the transformers library. Until it's merged, please use the model class we shared with the model checkpoint. The provided example should work if you downloaded the model and class file (which is done automatically) in the same directory.

kqsong

Aug 24, 2024

•

edited Aug 24, 2024

Hi. Yes, we are working on adding our new model to the transformers library. Until it's merged, please use the model class we shared with the model checkpoint. The provided example should work if you downloaded the model and class file (which is done automatically) in the same directory.

I tried to use text-generation with the provided code in the checkpoint repo. I want to do a 32k context length inference. however, it shows OOM for 8xH100 GPUs. Can you please provide any vLLM or TensorRT support or MII / Deepspeed inference example so that I can run through the code for a long context?

Thanks a lot.

nguyenbh

Microsoft org Aug 26, 2024

@kqsong If you want to do 32K context length inference, an alternative way is to get an endpoint on Azure AI https://aka.ms/try-phi3.5moe

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