update doc and requirements.txt (#119)

* update requirements

* update transformers version

* lint

* comments

* lint

* update requirements

* remove setup_requires

---------
Co-authored-by: dongchunyu <dongchunyu@pjlab.org.cn>

README.md

@@ -108,6 +108,12 @@ For the deployment of other supported models, such as LLaMA, vicuna, you can fin
 
 ### Inference with PyTorch
 
+You have to install deepspeed first before running with PyTorch.
+
+```
+pip install deepspeed
+```
+
 #### Single GPU
 
 ```shell

README_zh-CN.md

@@ -107,6 +107,12 @@ python3 -m lmdeploy.app {server_ip_addresss}:33337 internlm
 
 ### 基于 PyTorch 的推理
 
+你必须确保环境中有安装 deepspeed：
+
+```
+pip install deepspeed
+```
+
 #### 单个 GPU
 
 ```shell

lmdeploy/serve/turbomind/__init__.py

-# Copyright (c) OpenMMLab. All rights reserved.
-from lmdeploy.serve.turbomind.chatbot import Chatbot  # noqa: F401,F403

requirements.txt

 fire
+gradio
 mmengine
 numpy
 pybind11
 safetensors
+sentencepiece
 setuptools
-transformers<4.30.0
-tritonclient==2.33.0
+torch
+transformers
+tritonclient[all]

setup.py

@@ -106,7 +106,6 @@ if __name__ == '__main__':
           packages=find_packages(
               exclude=('lmdeploy/serve/turbomind/triton_models', )),
           include_package_data=True,
-          setup_requires=parse_requirements('requirements.txt'),
           install_requires=parse_requirements('requirements.txt'),
           classifiers=[
               'Programming Language :: Python :: 3.8',