Luke Japanese Large Sentiment Analysis Wrime

🚀 Luke Japanese Large Sentiment Analysis Model

This model is fine-tuned from Luke-japanese-large-lite and can analyze which of the eight emotions (joy, sadness, anticipation, surprise, anger, fear, disgust, trust) are included in the text.

🚀 Quick Start

This model is a fine - tuned version based on studio - ousia/Luke - japanese - large - lite. It can analyze which emotions (joy, sadness, anticipation, surprise, anger, fear, disgust, or trust) are present in a given text. The model was trained using the wrime dataset (https://huggingface.co/datasets/shunk031/wrime).

✨ Features

• Emotion Analysis: Capable of analyzing eight different emotions in text.
• Fine - Tuned Model: Based on Luke - japanese - large - lite and fine - tuned on the wrime dataset.

📦 Installation

1. Update transformers and install sentencepiece, Python, and PyTorch.

💻 Usage Examples

Basic Usage

from transformers import AutoTokenizer, AutoModelForSequenceClassification, LukeConfig
import torch
tokenizer = AutoTokenizer.from_pretrained("Mizuiro-sakura/luke-japanese-large-sentiment-analysis-wrime")
config = LukeConfig.from_pretrained('Mizuiro-sakura/luke-japanese-large-sentiment-analysis-wrime', output_hidden_states=True)    
model = AutoModelForSequenceClassification.from_pretrained('Mizuiro-sakura/luke-japanese-large-sentiment-analysis-wrime', config=config)

text='すごく楽しかった。また行きたい。'

max_seq_length=512
token=tokenizer(text,
        truncation=True,
        max_length=max_seq_length,
        padding="max_length")
output=model(torch.tensor(token['input_ids']).unsqueeze(0), torch.tensor(token['attention_mask']).unsqueeze(0))
max_index=torch.argmax(torch.tensor(output.logits))

if max_index==0:
    print('joy、うれしい')
elif max_index==1:
    print('sadness、悲しい')
elif max_index==2:
    print('anticipation、期待')
elif max_index==3:
    print('surprise、驚き')
elif max_index==4:
    print('anger、怒り')
elif max_index==5:
    print('fear、恐れ')
elif max_index==6:
    print('disgust、嫌悪')
elif max_index==7:
    print('trust、信頼')

🔧 Technical Details

What is Luke?

LUKE (Language Understanding with Knowledge - based Embeddings) is a new pre - trained contextualized representation of words and entities based on the transformer. LUKE treats words and entities in a given text as independent tokens and outputs contextualized representations of them. It adopts an entity - aware self - attention mechanism, which is an extension of the self - attention mechanism of the transformer and considers the types of tokens (words or entities) when computing attention scores.

LUKE achieves state - of - the - art results on five popular NLP benchmarks including SQuAD v1.1 (extractive question answering), CoNLL - 2003 (named entity recognition), ReCoRD (cloze - style question answering), TACRED (relation classification), and Open Entity (entity typing). luke - japanese is the Japanese version of the knowledge - enhanced pre - trained Transformer model LUKE for words and entities.

📄 License

This project is licensed under the MIT License.

Acknowledgments

I would like to thank Mr. Yamada @ikuyamada and Studio ousia @StudioOusia.

Citation

[1]@inproceedings{yamada2020luke, title={LUKE: Deep Contextualized Entity Representations with Entity - aware Self - attention}, author={Ikuya Yamada and Akari Asai and Hiroyuki Shindo and Hideaki Takeda and Yuji Matsumoto}, booktitle={EMNLP}, year={2020} }