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

Libraries

How to use jsincn/phi-3-mini-128k-instruct-awq with Transformers:

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

pipe = pipeline("text-generation", model="jsincn/phi-3-mini-128k-instruct-awq", 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("jsincn/phi-3-mini-128k-instruct-awq", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("jsincn/phi-3-mini-128k-instruct-awq", 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 jsincn/phi-3-mini-128k-instruct-awq with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "jsincn/phi-3-mini-128k-instruct-awq"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "jsincn/phi-3-mini-128k-instruct-awq",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/jsincn/phi-3-mini-128k-instruct-awq

SGLang

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

Docker Model Runner
How to use jsincn/phi-3-mini-128k-instruct-awq with Docker Model Runner:
```
docker model run hf.co/jsincn/phi-3-mini-128k-instruct-awq
```

This repo includes a Version of Phi-3 that was quantized to AWQ using AutoAWQ. Currently hosting via the TGI docker image fails due to its fallback on AutoModel and that not being compatible with AWQ. Hosting on vLLM is recommended.

To run the model you need to set the trust-remote-code (or similar) flag. While the remote code comes from microsoft (see LICENSE information in the file) you should validate the code yourself before deployment.

Downloads last month: 25

Safetensors

Model size

4B params

Tensor type

I32

F16