沈阳网站建河北百度seo点击软件

1. 登录PyTorch查看需要的Python和CUDA版本，并进行安装

2.安装CUDA

下载并安装 CUDA Toolkit - Free Tools and Training | NVIDIA Developer

安装后运行  nvidia-smi  ， 返回显卡各种信息

3.使用conda创建虚拟环境（安装过程中有些包会冲突，这是我倒出的关键包）

# conda create -f environment.ymlname: aipy11
channels:- http://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main- http://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/win-64- http://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main/win-64
dependencies:- bzip2=1.0.8=h2bbff1b_6- ca-certificates=2025.2.25=haa95532_0- expat=2.7.1=h8ddb27b_0- libffi=3.4.4=hd77b12b_1- openssl=3.0.16=h3f729d1_0- pip=25.1=pyhc872135_2- python=3.11.13=h981015d_0- setuptools=78.1.1=py311haa95532_0- sqlite=3.45.3=h2bbff1b_0- tk=8.6.14=h5e9d12e_1- vc=14.42=haa95532_5- vs2015_runtime=14.42.34433=hbfb602d_5- wheel=0.45.1=py311haa95532_0- xz=5.6.4=h4754444_1- zlib=1.2.13=h8cc25b3_1- pip:- accelerate==1.7.0- addict==2.4.0- huggingface-hub==0.33.0- langchain==0.3.25- langchain-community==0.3.25- langchain-core==0.3.65- langchain-openai==0.3.22- langchain-text-splitters==0.3.8- langsmith==0.3.45- llama-cloud==0.1.26- llama-cloud-services==0.6.31- llama-index==0.12.42- llama-index-agent-openai==0.4.11- llama-index-cli==0.4.3- llama-index-core==0.12.42- llama-index-embeddings-huggingface==0.5.4- llama-index-embeddings-openai==0.3.1- llama-index-indices-managed-llama-cloud==0.7.7- llama-index-llms-huggingface==0.5.0- llama-index-llms-openai==0.4.5- llama-index-multi-modal-llms-openai==0.5.1- llama-index-program-openai==0.3.2- llama-index-question-gen-openai==0.3.1- llama-index-readers-file==0.4.9- llama-index-readers-llama-parse==0.4.0- llama-parse==0.6.31- pillow==11.0.0- pydantic==2.11.5- pydantic-core==2.33.2- pydantic-settings==2.9.1

3.大模型的下载

可以从modelscope、huggingface、ollama等平台下载 ，指定保存位置，否则都到了c盘

os.environ["TRANSFORMERS_CACHE"] = cache_apath
os.environ["HF_HOME"] = cache_apath
os.environ["MODELSCOPE_CACHE"] = cache_apath