Edit model card

Zero-shot text classification (large-sized model) trained with self-supervised tuning

Zero-shot text classification model trained with self-supervised tuning (SSTuning). It was introduced in the paper Zero-Shot Text Classification via Self-Supervised Tuning by Chaoqun Liu, Wenxuan Zhang, Guizhen Chen, Xiaobao Wu, Anh Tuan Luu, Chip Hong Chang, Lidong Bing and first released in this repository.

The model backbone is RoBERTa-large.

Model description

The model is tuned with unlabeled data using a learning objective called first sentence prediction (FSP). The FSP task is designed by considering both the nature of the unlabeled corpus and the input/output format of classification tasks. The training and validation sets are constructed from the unlabeled corpus using FSP.

During tuning, BERT-like pre-trained masked language models such as RoBERTa and ALBERT are employed as the backbone, and an output layer for classification is added. The learning objective for FSP is to predict the index of the correct label. A cross-entropy loss is used for tuning the model.

Model variations

There are three versions of models released. The details are:

Model Backbone #params accuracy Speed #Training data
zero-shot-classify-SSTuning-base roberta-base 125M Low High 20.48M
zero-shot-classify-SSTuning-large roberta-large 355M Medium Medium 5.12M
zero-shot-classify-SSTuning-ALBERT albert-xxlarge-v2 235M High Low 5.12M

Please note that zero-shot-classify-SSTuning-base is trained with more data (20.48M) than the paper, as this will increase the accuracy.

Intended uses & limitations

The model can be used for zero-shot text classification such as sentiment analysis and topic classification. No further finetuning is needed.

The number of labels should be 2 ~ 20.

How to use

You can try the model with the Colab Notebook.

from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch, string, random

tokenizer = AutoTokenizer.from_pretrained("DAMO-NLP-SG/zero-shot-classify-SSTuning-large")
model = AutoModelForSequenceClassification.from_pretrained("DAMO-NLP-SG/zero-shot-classify-SSTuning-large")

text = "I love this place! The food is always so fresh and delicious."
list_label = ["negative", "positive"]

device = torch.device('cuda') if torch.cuda.is_available() else torch.device('cpu')
list_ABC = [x for x in string.ascii_uppercase]

def check_text(model, text, list_label, shuffle=False): 
    list_label = [x+'.' if x[-1] != '.' else x for x in list_label]
    list_label_new = list_label + [tokenizer.pad_token]* (20 - len(list_label))
    if shuffle: 
    s_option = ' '.join(['('+list_ABC[i]+') '+list_label_new[i] for i in range(len(list_label_new))])
    text = f'{s_option} {tokenizer.sep_token} {text}'

    encoding = tokenizer([text],truncation=True, max_length=512,return_tensors='pt')
    item = {key: val.to(device) for key, val in encoding.items()}
    logits = model(**item).logits
    logits = logits if shuffle else logits[:,0:len(list_label)]
    probs = torch.nn.functional.softmax(logits, dim = -1).tolist()
    predictions = torch.argmax(logits, dim=-1).item() 
    probabilities = [round(x,5) for x in probs[0]]

    print(f'prediction:    {predictions} => ({list_ABC[predictions]}) {list_label_new[predictions]}')
    print(f'probability:   {round(probabilities[predictions]*100,2)}%')

check_text(model, text, list_label)
# prediction:    1 => (B) positive.
# probability:   99.84%

BibTeX entry and citation info

  author    = {Chaoqun Liu and
               Wenxuan Zhang and
               Guizhen Chen and
               Xiaobao Wu and
               Anh Tuan Luu and
               Chip Hong Chang and 
               Lidong Bing},
  title     = {Zero-Shot Text Classification via Self-Supervised Tuning},
  booktitle = {Findings of the Association for Computational Linguistics: ACL 2023},
  year      = {2023},
  url       = {https://arxiv.org/abs/2305.11442},
Downloads last month
Inference Examples
Inference API (serverless) has been turned off for this model.