Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
项目地址：[LLMPruner：大语言模型裁剪工具](https://github.com/yangjianxin1/LLMPruner)
|
| 2 |
+
|
| 3 |
+
LLMPruner是一个大语言模型裁剪工具，通过对大语言模型的冗余词表进行裁剪，减少模型参数量，降低显存占用，提升训练速度，并且能够保留预训练中学习到的知识。
|
| 4 |
+
|
| 5 |
+
本项目对Bloom进行词表裁剪，保留中文token和常用的英文token，词表由250880将至46145，缩减为原来的18.39%。裁剪得到的Bloom模型如下表：
|
| 6 |
+
|
| 7 |
+
| 裁剪模型 | 原模型 | 参数量比例 |
|
| 8 |
+
|-----------------------------------------------------------------------------|-----------------------------------------------------------------------------|--------|
|
| 9 |
+
| [YeungNLP/bloom-396m-zh](https://huggingface.co/YeungNLP/bloom-396m-zh) | [bigscience/bloom-560m](https://huggingface.co/bigscience/bloom-560m) | 70.96% |
|
| 10 |
+
| [YeungNLP/bloom-820m-zh](https://huggingface.co/YeungNLP/bloom-820m-zh) | [bigscience/bloom-1b1](https://huggingface.co/bigscience/bloom-1b1) | 77.13% |
|
| 11 |
+
| [YeungNLP/bloom-1b4-zh](https://huggingface.co/YeungNLP/bloom-1b4-zh) | [bigscience/bloom-1b7](https://huggingface.co/bigscience/bloom-1b7) | 81.14% |
|
| 12 |
+
| [YeungNLP/bloom-2b6-zh](https://huggingface.co/YeungNLP/bloom-2b6-zh) | [bigscience/bloom-3b](https://huggingface.co/bigscience/bloom-3b) | 86.48% |
|
| 13 |
+
| [YeungNLP/bloom-6b4-zh](https://huggingface.co/YeungNLP/bloom-6b4-zh) | [bigscience/bloom-7b1](https://huggingface.co/bigscience/bloom-7b1) | 90.81% |
|
| 14 |
+
| [YeungNLP/bloomz-396m-zh](https://huggingface.co/YeungNLP/bloomz-396m-zh) | [bigscience/bloomz-560m](https://huggingface.co/bigscience/bloomz-560m) | 70.96% |
|
| 15 |
+
| [YeungNLP/bloomz-820m-zh](https://huggingface.co/YeungNLP/bloomz-820m-zh) | [bigscience/bloomz-1b1](https://huggingface.co/bigscience/bloomz-1b1) | 77.13% |
|
| 16 |
+
| [YeungNLP/bloomz-1b4-zh](https://huggingface.co/YeungNLP/bloomz-1b4-zh) | [bigscience/bloomz-1b7](https://huggingface.co/bigscience/bloomz-1b7) | 81.14% |
|
| 17 |
+
| [YeungNLP/bloomz-2b6-zh](https://huggingface.co/YeungNLP/bloomz-2b6-zh) | [bigscience/bloomz-3b](https://huggingface.co/bigscience/bloomz-3b) | 86.48% |
|
| 18 |
+
| [YeungNLP/bloomz-6b4-zh](https://huggingface.co/YeungNLP/bloomz-6b4-zh) | [bigscience/bloomz-7b1](https://huggingface.co/bigscience/bloomz-7b1) | 90.81% |
|
| 19 |
+
| [YeungNLP/bloomz-6b4-mt-zh](https://huggingface.co/YeungNLP/bloomz-6b4-mt-zh) | [bigscience/bloomz-7b1-mt](https://huggingface.co/bigscience/bloomz-7b1-mt) | 90.81% |
|
| 20 |
+
|
| 21 |
+
|
| 22 |
+
使用方法：
|
| 23 |
+
```python
|
| 24 |
+
from transformers import BloomTokenizerFast, BloomForCausalLM
|
| 25 |
+
|
| 26 |
+
tokenizer = BloomTokenizerFast.from_pretrained('YeungNLP/bloom-1b4-zh')
|
| 27 |
+
model = BloomForCausalLM.from_pretrained('YeungNLP/bloom-1b4-zh')
|
| 28 |
+
print(tokenizer.batch_decode(model.generate(tokenizer.encode('长风破浪会有时', return_tensors='pt'))))
|
| 29 |
+
```
|