---
tags:
- text-generation
language: zh
---

## Note

**This repo is under development.**

## Usage

```
pip install transformers
```

```python
from transformers import CPMAntTokenizer, CPMAntModel

texts = ["今天天气真好！"]
model = CPMAntModel.from_pretrained("openbmb/cpm-ant-10b")
tokenizer = CPMAntTokenizer.from_pretrained("openbmb/cpm-ant-10b")
input_ids = tokenizer.get_model_input(texts)
logits, hidden = model(**input_ids)

print(logits.shape)
print(hidden.shape)
```
## pipeline
```python
texts = ["昨天多云转阴，", "今天天气不错，"]
generator = pipeline("text-generation", model="openbmb/cpm-ant-10b", tokenizer="openbmb/cpm-ant-10b")
result = generator(texts)
print(result)
```