---
title: zero-shot-classify-SSTuning-base
canonical_url: "https://www.modelscope.cn/models/iic/zero-shot-classify-SSTuning-base"
md_url: "https://www.modelscope.cn/models/iic/zero-shot-classify-SSTuning-base.md"
repository: iic/zero-shot-classify-SSTuning-base
chinese_name: "零样本文本分类-SSTuning-base-英语"
last_updated: 2025-06-11
license: mit
pipeline_tag: text-classification
tasks:
  - text-classification
  - zero-shot-classification
  - sentiment-classification
model_type:
  - roberta
architectures:
  - RobertaForSequenceClassification
library_name:
  - pytorch
frameworks:
  - pytorch
language:
  - en
downloads: 2248
stars: 4
---

# zero-shot-classify-SSTuning-base

> zero-shot-classify-SSTuning-base - iic 在 ModelScope 开源的模型。零样本文本分类模型（英语），可以用于情感分类和主题分类。

iic/zero-shot-classify-SSTuning-base 是 ModelScope 魔搭社区上的text-classification、zero-shot-classification、sentiment-classification模型，采用 mit 许可。

- **Repository**: iic/zero-shot-classify-SSTuning-base
- **License**: mit
- **Tasks**: text-classification, zero-shot-classification, sentiment-classification
- **Downloads**: 2248
- **Stars**: 4
- **Last updated**: 2025-06-11

Source: https://www.modelscope.cn/models/iic/zero-shot-classify-SSTuning-base

---

# 模型介绍

* zero-shot-classify-SSTuning-base是达摩院通过自监督微调得到的零样本文本分类模型。
* 模型骨干是RoBERTa-base。
* 具体训练过程请参考这篇论文：[Zero-Shot Text Classification via Self-Supervised Tuning](https://arxiv.org/abs/2305.11442)。
* 更多模型和测试结果请参考[this repository](https://github.com/DAMO-NLP-SG/SSTuning)。

## 快速开始

### Pipeline
```python
from modelscope.pipelines import pipeline

model = 'damo/zero-shot-classify-SSTuning-base'
pipe = pipeline('zero-shot-classify-sstuning', model=model, model_revision='master')

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

output = pipe(text,list_label = list_label)
print(output)
# {'prediction': 'positive.', 'probability': '99.92%'}
```

### python
```python
from modelscope import AutoTokenizer, AutoModelForSequenceClassification, snapshot_download
import torch, string, random

model_dir = snapshot_download("damo/zero-shot-classify-SSTuning-base", revision='v1.0.2')

tokenizer = AutoTokenizer.from_pretrained(model_dir)
model = AutoModelForSequenceClassification.from_pretrained(model_dir)

text = "I love this place! The food is always so fresh and delicious."
list_label = ["negative", "positive"] #The number of labels should be 2 ~ 20. 

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: 
        random.shuffle(list_label_new)
    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}'

    model.to(device).eval()
    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:   99.92%
```

### 相关论文以及引用信息
```bibtxt
@inproceedings{acl23/SSTuning,
  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},
}
```