update README.md
Browse files
README.md
CHANGED
|
@@ -10,7 +10,7 @@ tags:
|
|
| 10 |
|
| 11 |
Link to github: [here](https://github.com/sunzeyeah/RLHF)
|
| 12 |
|
| 13 |
-
|
| 14 |
|
| 15 |
本仓库由[THUDM/chatglm-6b](https://huggingface.co/THUDM/chatglm-6b) fork而来，原仓库实现了PyTorch版本的ChatGLM模型，该模型有60亿参数量，模型权重文件以FP16格式存储。
|
| 16 |
|
|
@@ -20,19 +20,24 @@ This repository is forked from [THUDM/chatglm-6b](https://huggingface.co/THUDM/c
|
|
| 20 |
|
| 21 |
It is slightly different from the original ChatGLM implementation to support the ChatGPT training pipeline in this github repo: [sunzeyeah/RLHF](https://github.com/sunzeyeah/RLHF).
|
| 22 |
|
|
|
|
| 23 |
|
| 24 |
-
|
| 25 |
ChatGLM-6B 是一个开源的、支持中英双语问答的对话语言模型，基于 [General Language Model (GLM)](https://github.com/THUDM/GLM) 架构，具有 62 亿参数。结合模型量化技术，用户可以在消费级的显卡上进行本地部署（INT4 量化级别下最低只需 6GB 显存）。ChatGLM-6B 使用了和 [ChatGLM](https://chatglm.cn) 相同的技术，针对中文问答和对话进行了优化。经过约 1T 标识符的中英双语训练，辅以监督微调、反馈自助、人类反馈强化学习等技术的加持，62 亿参数的 ChatGLM-6B 已经能生成相当符合人类偏好的回答。
|
| 26 |
|
| 27 |
ChatGLM-6B is an open bilingual language model based on [General Language Model (GLM)](https://github.com/THUDM/GLM) framework, with 6.2 billion parameters. With the quantization technique, users can deploy locally on consumer-grade graphics cards (only 6GB of GPU memory is required at the INT4 quantization level). ChatGLM-6B uses technology similar to ChatGPT, optimized for Chinese QA and dialogue. The model is trained for about 1T tokens of Chinese and English corpus, supplemented by supervised fine-tuning, feedback bootstrap, and reinforcement learning wit human feedback. With only about 6.2 billion parameters, the model is able to generate answers that are in line with human preference.
|
| 28 |
|
| 29 |
-
|
|
|
|
|
|
|
| 30 |
|
| 31 |
```shell
|
| 32 |
pip install protobuf==3.20.0 transformers==4.26.1 icetk cpm_kernels
|
| 33 |
```
|
| 34 |
|
| 35 |
-
|
|
|
|
|
|
|
| 36 |
|
| 37 |
可以通过如下代码调用 ChatGLM-6B 模型来生成对话：
|
| 38 |
|
|
|
|
| 10 |
|
| 11 |
Link to github: [here](https://github.com/sunzeyeah/RLHF)
|
| 12 |
|
| 13 |
+
---
|
| 14 |
|
| 15 |
本仓库由[THUDM/chatglm-6b](https://huggingface.co/THUDM/chatglm-6b) fork而来，原仓库实现了PyTorch版本的ChatGLM模型，该模型有60亿参数量，模型权重文件以FP16格式存储。
|
| 16 |
|
|
|
|
| 20 |
|
| 21 |
It is slightly different from the original ChatGLM implementation to support the ChatGPT training pipeline in this github repo: [sunzeyeah/RLHF](https://github.com/sunzeyeah/RLHF).
|
| 22 |
|
| 23 |
+
---
|
| 24 |
|
| 25 |
+
# 介绍
|
| 26 |
ChatGLM-6B 是一个开源的、支持中英双语问答的对话语言模型，基于 [General Language Model (GLM)](https://github.com/THUDM/GLM) 架构，具有 62 亿参数。结合模型量化技术，用户可以在消费级的显卡上进行本地部署（INT4 量化级别下最低只需 6GB 显存）。ChatGLM-6B 使用了和 [ChatGLM](https://chatglm.cn) 相同的技术，针对中文问答和对话进行了优化。经过约 1T 标识符的中英双语训练，辅以监督微调、反馈自助、人类反馈强化学习等技术的加持，62 亿参数的 ChatGLM-6B 已经能生成相当符合人类偏好的回答。
|
| 27 |
|
| 28 |
ChatGLM-6B is an open bilingual language model based on [General Language Model (GLM)](https://github.com/THUDM/GLM) framework, with 6.2 billion parameters. With the quantization technique, users can deploy locally on consumer-grade graphics cards (only 6GB of GPU memory is required at the INT4 quantization level). ChatGLM-6B uses technology similar to ChatGPT, optimized for Chinese QA and dialogue. The model is trained for about 1T tokens of Chinese and English corpus, supplemented by supervised fine-tuning, feedback bootstrap, and reinforcement learning wit human feedback. With only about 6.2 billion parameters, the model is able to generate answers that are in line with human preference.
|
| 29 |
|
| 30 |
+
---
|
| 31 |
+
|
| 32 |
+
# 软件依赖
|
| 33 |
|
| 34 |
```shell
|
| 35 |
pip install protobuf==3.20.0 transformers==4.26.1 icetk cpm_kernels
|
| 36 |
```
|
| 37 |
|
| 38 |
+
---
|
| 39 |
+
|
| 40 |
+
# 代码调用
|
| 41 |
|
| 42 |
可以通过如下代码调用 ChatGLM-6B 模型来生成对话：
|
| 43 |
|