File size: 4,058 Bytes
56a9f40 37211fc 56a9f40 37211fc |
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 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 |
---
license: other
language:
- zh
- en
---
# BlueLM
<p align="center">
🖥 <a href="https://github.com/vivo-ai-lab/BlueLM" target="_blank">github</a> • 📜 <a href="https://huggingface.co/vivo-ai/BlueLM-7B-Base/blob/main/MODEL_LICENSE" target="_blank">LICENSE</a> • 🎯 <a href="https://developers.vivo.com/product/ai/bluelm" target="_blank">vivo Developers</a>
</p>
## 模型介绍/Introduction
BlueLM 是由 vivo AI 全球研究院自主研发的大规模预训练语言模型，本次发布包含 7B 基础模型和 7B 对话模型，同时我们开源了支持 **32K** 的长文本基础模型和对话模型。
- **更大量的优质数据**：高质量语料库进行训练，规模达到了 **2.6 万亿** 的 token 数，该语料库包含中文、英文以及少量日韩数据。
- **更优的效果**：其中 BlueLM-7B-Chat 在 **C-Eval** 和 **CMMLU** 上均取得领先结果，对比同尺寸开源模型中具有较强的竞争力。
- **长文本支持**：BlueLM-7B-Base-32K 和 BlueLM-7B-Chat-32K 均支持 **32K** 长文本，在保持基础能力相当情况下，能够支持更长上下文理解。
- **协议说明**：BlueLM 系列欢迎开发者进行学术研究和商业应用。
BlueLM is a large-scale open-source language model independently developed by the vivo AI Lab. This release includes 2K and 32K context length versions for both Base and Chat models.
- **High-quality Data**: BlueLM is trained on a high-quality data with 2.6 trillion tokens. Our train corpus contains Chinese, English, Japanese and Korean data.
- **Stronger Performance**: BlueLM-7B-Chat achieves the best performance in C-Eval and CMMLU benchmarks of the same size.
- **Longer Context**: We have extended the context length of both BlueLM-7B-Base-32K and BlueLM-7B-Chat-32K models from 2K to 32K. The models can support longer context understanding while maintaining the same basic capabilities.
- **Model License**: BlueLM weights are open for academic research and commercial use.
本次发布基座模型下载链接见：
The release versions and hugging face download links are listed in the table below:
| | Base Model | Chat Model | 4bits Quantized Chat Model |
|:---:|:--------------------:|:--------------------:|:--------------------------:|
| 7B | [BlueLM-7B-Base](https://huggingface.co/vivo-ai/BlueLM-7B-Base) | [BlueLM-7B-Chat](https://huggingface.co/vivo-ai/BlueLM-7B-Chat) | [BlueLM-7B-Chat-4bits](https://huggingface.co/vivo-ai/BlueLM-7B-Chat-4bits) |
| 7B-32K | [BlueLM-7B-Base-32K](https://huggingface.co/vivo-ai/BlueLM-7B-Base-32K) | [BlueLM-7B-Chat-32K](https://huggingface.co/vivo-ai/BlueLM-7B-Chat-32K) | - |
## 推理部署/Inference and Deployment
```python
>>> from transformers import AutoModelForCausalLM, AutoTokenizer
>>> tokenizer = AutoTokenizer.from_pretrained("vivo-ai/BlueLM-7B-Base", trust_remote_code=True, use_fast=False)
>>> model = AutoModelForCausalLM.from_pretrained("vivo-ai/BlueLM-7B-Base", device_map="cuda:0", trust_remote_code=True)
>>> model = model.eval()
>>> inputs = tokenizer("儒林外史->吴敬梓\n隋唐演义->褚人获\n红楼梦->", return_tensors="pt")
>>> inputs = inputs.to("cuda:0")
>>> pred = model.generate(**inputs, max_new_tokens=64, repetition_penalty=1.1)
>>> print(tokenizer.decode(pred.cpu()[0], skip_special_tokens=True))
儒林外史->吴敬梓
隋唐演义->褚人获
红楼梦->曹雪芹
三国演义->罗贯中
水浒传->施耐庵
西游记->吴承恩
聊斋志异->蒲松龄
封神演义->许仲琳
东周列国志->冯梦龙
三侠五义->石玉昆
七剑十三侠->唐芸洲
```
更多使用说明，请参考我们的 [Github 仓库](https://github.com/vivo-ai-lab/BlueLM).
For more instructions, please refer to our [Github Repo](https://github.com/vivo-ai-lab/BlueLM).
## 协议/License
Our code is licensed under the [Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0) and [Community License for BlueLM Model](https://huggingface.co/vivo-ai/BlueLM-7B-Base/blob/main/MODEL_LICENSE). |