README.md

---
license: mit
license_link: https://huggingface.co/microsoft/phi-2/resolve/main/LICENSE
language:
  - en
---

# phi-2-fp16-ov

 * Model creator: [Microsoft](https://huggingface.co/microsoft)
 * Original model: [phi-2](https://huggingface.co/microsoft/phi-2)

## Description

This is [phi-2](https://huggingface.co/microsoft/phi-2) model converted to the [OpenVINO™ IR](https://docs.openvino.ai/2024/documentation/openvino-ir-format.html) (Intermediate Representation) format with weights compressed to FP16.

## Compatibility

The provided OpenVINO™ IR model is compatible with:

* OpenVINO version 2024.1.0 and higher
* Optimum Intel 1.16.0 and higher

## Running Model Inference

1. Install packages required for using [Optimum Intel](https://huggingface.co/docs/optimum/intel/index) integration with the OpenVINO backend:

```
pip install optimum[openvino]
```

2. Run model inference:

```
from transformers import AutoTokenizer
from optimum.intel.openvino import OVModelForCausalLM

model_id = "OpenVINO/phi-2-fp16-ov"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = OVModelForCausalLM.from_pretrained(model_id)

inputs = tokenizer("What is OpenVINO?", return_tensors="pt")

outputs = model.generate(**inputs, max_length=200)
text = tokenizer.batch_decode(outputs)[0]
print(text)
```

For more examples and possible optimizations, refer to the [OpenVINO Large Language Model Inference Guide](https://docs.openvino.ai/2024/learn-openvino/llm_inference_guide.html).

## Limitations

Check the original model card for [limitations](https://huggingface.co/microsoft/phi-2#limitations-of-phi-2).

## Legal information

The original model is distributed under [MIT](https://huggingface.co/microsoft/phi-2/resolve/main/LICENSE) license. More details can be found in [original model card](https://huggingface.co/microsoft/phi-2).