---
license: apache-2.0
datasets:
- Universal-NER/Pile-NER-type
language:
- en
---
<div align="center">

# tiny-universal-NER
</div>

This model is finetuned from [TinyLLama](https://huggingface.co/TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T). 

It is trained on ChatGPT-generated [Pile-NER-type data](https://huggingface.co/datasets/Universal-NER/Pile-NER-type).

Check this [paper](https://arxiv.org/abs/2308.03279) for more information.


### How to use
You will need the transformers>=4.34
Do check the [TinyLlama](https://github.com/jzhang38/TinyLlama) github page for more information.

```python
# Install transformers from source - only needed for versions <= v4.34
# pip install git+https://github.com/huggingface/transformers.git
# pip install accelerate
import torch
from transformers import pipeline

pipe = pipeline("text-generation", model="LR-AI-Labs/tiny-universal-NER",
                torch_dtype=torch.bfloat16, device_map="auto")
messages = [
    {
        "role": "system",
        "content": "A virtual assistant answers questions from a user based on the provided text.",
    },
    {
        "role": "user",
        "content": "Text: VinBigData Joint Stock Company provides platform technology solutions and advanced products based on Big Data and Artificial Intelligence. With a staff of professors, doctors, and global technology experts, VinBigData is currently developing and deploying products such as ViVi virtual assistant, VinBase the comprehensive multi-cognitive artificial intelligence ecosystem, Vizone the ecosystem of smart image analysis solutions, VinDr the medical image digitization platform,..."
    },
    {
        "role": "assistant",
        "content": "I've read this text."
    },
    {
        "role": "user",
        "content": "What describes products in the text?"
    }
]
prompt = pipe.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
outputs = pipe(prompt, max_new_tokens=256, do_sample=False)
print(outputs[0]["generated_text"])
# <|system|>
# A virtual assistant answers questions from a user based on the provided text.</s>
# <|user|>
# Text: The American Bank Note Company Printing Plant is a repurposed complex of three interconnected buildings in the Hunts Point neighborhood of the Bronx in New York City. The innovative Kirby, Petit & Green design was built in 1909–1911 by the American Bank Note Company on land which had previously been part of Edward G. Faile's country estate. A wide variety of financial instruments were printed there; at one point, over five million documents were produced per day, including half the securities traded on the New York Stock Exchange.</s>
# <|assistant|>
# I've read this text.</s>
# <|user|>
# What describes location in the text?</s>
# <|assistant|>
# ["ViVi", "VinBase", "Vizone", "VinDr"]
```

### Note: Inferences are based on one entity type at a time. For multiple entity types, create separate instances for each type.

## License


This model and its associated data are released under the [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/) license. They are primarily used for research purposes.