File size: 1,812 Bytes
1960c3c 79a1f2e 1960c3c 79a1f2e 330fa80 79a1f2e |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 |
---
license: apache-2.0
language:
- zh
- en
tags:
- openba
---
# Introduction
OpenBA is an Open-Sourced 15B Bilingual Asymmetric Seq2Seq Model Pre-trained from Scratch.
## Open Source Plan
We are excited to unveil two distinguished versions of our model, with another on the horizon:
- [OpenBA-LM](https://huggingface.co/OpenBA/OpenBA-LM): The backbone language models was pre-trained on 340B English, Chinese, and code tokens.
- [OpenBA-Flan](https://huggingface.co/OpenBA/OpenBA-Flan): We perform supervised fine-tuning on the base model with additional 40B tokens using our collected BiFlan Dataset.
- OpenBA-Chat: coming soon
## Model Description
- **Model type:** Language model
- **Language(s) (NLP):** zh, en (We also offer the possibility for multilingual learning, by using a multilingual tokenizer.)
- **License:** Apache 2.0
- **Resources for more information:**
- [Paper](https://arxiv.org/abs/2309.10706)
- [GitHub Repo](https://github.com/OpenNLG/OpenBA/)
# Usage
## Install requirements
```bash
pip install transformers torch>=2.0 sentencepiece
```
## Demo usage
```python
>>> from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
>>> tokenizer = AutoTokenizer.from_pretrained("OpenBA/OpenBA-LM", trust_remote_code=True)
>>> model = AutoModelForSeq2SeqLM.from_pretrained("OpenBA/OpenBA-LM", trust_remote_code=True).half().cuda()
>>> model = model.eval()
>>> query = "<S>" + "苏州处太湖平原，沿江为高沙平原，河" + "<extra_id_0>"
>>> inputs = tokenizer(query, return_tensors="pt").to("cuda")
>>> outputs = model.generate(**inputs, do_sample=True, max_new_tokens=32)
>>> response = tokenizer.decode(outputs[0], skip_special_tokens=True)
>>> print(response)
流两侧为河淤平原,苏州平原是江苏平原主体,地势低平,土地肥沃,气候温和
``` |