> cmd
> cd D:\devtools\PythonVenv
> python3 -m venv llama2
> D:\devtools\PythonVenv\llama2\Scripts\activate.bat
> set http_proxy=192.168.2.199:58591
> set http_proxy=192.168.2.199:58591

> pip install -r requirements.txt
> pip install jupyter notebook -i https://pypi.tuna.tsinghua.edu.cn/simple
> pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118

> cmd
> D:\devtools\PythonVenv\llama2\Scripts\activate.bat

> Python: Select Interpreter

> torchrun --nproc_per_node 1 example_text_completion.py \
    --ckpt_dir llama-2-7b/ \
    --tokenizer_path tokenizer.model \
    --max_seq_len 128 --max_batch_size 4
>
> torchrun --nproc_per_node 1 example_text_completion.py --ckpt_dir E:\THUDM\llama2\model\llama-2-7b-chat --tokenizer_path E:\THUDM\llama2\model\tokenizer.model --max_seq_len 128 --max_batch_size 4

>
> python example_text_completion02.py --ckpt_dir E:\THUDM\llama2\model\llama-2-7b-chat --tokenizer_path E:\THUDM\llama2\model\tokenizer.model --max_seq_len 128 --max_batch_size 4
>

> export http_proxy=192.168.2.199:58591
> export https_proxy=192.168.2.199:58591

> wget https://developer.download.nvidia.com/compute/cuda/12.3.1/local_installers/cuda-repo-debian12-12-3-local_12.3.1-545.23.08-1_amd64.deb
> sudo dpkg -i cuda-repo-debian12-12-3-local_12.3.1-545.23.08-1_amd64.deb
> sudo cp /var/cuda-repo-debian12-12-3-local/cuda-*-keyring.gpg /usr/share/keyrings/
> sudo add-apt-repository contrib
> sudo apt-get update
> sudo apt-get -y install cuda-toolkit-12-3

> sudo apt install python3-venv  python3-pip
> mkdir /opt/Data/PythonVenv
> cd /opt/Data/PythonVenv
> python3 -m venv llama2
> source /opt/Data/PythonVenv/llama2/bin/activate

> pip install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple
> pip install gradio
> pip install pyreadline3
> pip install torch==2.0.1 torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118
> pip install transformers==4.31.0 protobuf accelerate llama2-wrapper -i https://pypi.tuna.tsinghua.edu.cn/simple
> pip install llama2-wrapper -i https://pypi.tuna.tsinghua.edu.cn/simple
> 

> # Linux
> python llama-recipes/src/llama_recipes/tools/convert_llama_weights_to_hf.py \
    --input_dir /opt/Data/THUDM/llama2 --model_size 7B --output_dir /opt/Data/THUDM/llama2.hf/llama-2-7b-chat-hf
>
> python llama-recipes/src/llama_recipes/tools/convert_llama_weights_to_hf.py --input_dir /opt/Data/THUDM/llama2 --model_size 13B --output_dir /opt/Data/THUDM/llama2.hf/llama-2-13b-chat-hf
>
> # windows
> python llama-recipes/src/llama_recipes/tools/convert_llama_weights_to_hf.py \
    --input_dir E:/THUDM/llama2 --model_size 7B --output_dir E:/THUDM/llama2.hf/llama-2-7b-chat-hf
>

> python openai/openapi.py
> python WebGradio/WebGradioAutoModel.py
> python WebGradio/WebGradioAutoModelLora.py
> python WebGradio/WebGradioAutoModelAtom.py
> jupyter notebook --no-browser --port 7001 --ip=192.168.2.198
> jupyter notebook --no-browser --port 7000 --ip=192.168.2.200