TurboPascal
commited on
Commit
•
85418ab
1
Parent(s):
424d5f1
Update README.md
Browse files
README.md
CHANGED
@@ -8,7 +8,7 @@ pipeline_tag: text-generation
|
|
8 |
Llama-zh-base is an open-source project that offers a complete training pipeline for building Chinese large language models, ranging from dataset preparation to tokenization, pre-training, prompt tuning, and the reinforcement learning technique RLHF.
|
9 |
This is the Llama-zh-base model trained from scratch using the Chinese pretrain corpus in this project.The amount of parameters is about 0.8B.
|
10 |
|
11 |
-
使用
|
12 |
|
13 |
项目github link [Repo Links](https://github.com/enze5088/Chatterbox/blob/main/docs/model/llama-zh-base.md)
|
14 |
|
@@ -34,7 +34,7 @@ Notes:
|
|
34 |
|
35 |
## 数据
|
36 |
|
37 |
-
预训练阶段使用开源数据与本项目爬取的部分数据。共使用约33G中文预训练数据、MC4-zh、Code数据集。清洗后筛选共120G左右数据训练
|
38 |
|
39 |
### 中文预训练数据
|
40 |
|
|
|
8 |
Llama-zh-base is an open-source project that offers a complete training pipeline for building Chinese large language models, ranging from dataset preparation to tokenization, pre-training, prompt tuning, and the reinforcement learning technique RLHF.
|
9 |
This is the Llama-zh-base model trained from scratch using the Chinese pretrain corpus in this project.The amount of parameters is about 0.8B.
|
10 |
|
11 |
+
使用120G中文语料重头开始预训练的Llama模型,旨在提供可用的中小型基础模型。重新构建了embedding层和tokenizer。目前未经过指令微调。参数量约为0.8B左右。
|
12 |
|
13 |
项目github link [Repo Links](https://github.com/enze5088/Chatterbox/blob/main/docs/model/llama-zh-base.md)
|
14 |
|
|
|
34 |
|
35 |
## 数据
|
36 |
|
37 |
+
预训练阶段使用开源数据与本项目爬取的部分数据。共使用约33G中文预训练数据、MC4-zh、Code数据集。清洗后筛选共120G左右数据训练1 epoch,初始学习率1e-4。未经过指令微调。
|
38 |
|
39 |
### 中文预训练数据
|
40 |
|