chatGlm报错

  File "/root/.cache/huggingface/modules/transformers_modules/chatglm-6b/quantization.py", line 157, in quantize
    layer.attention.query_key_value = QuantizedLinear(
  File "/root/.cache/huggingface/modules/transformers_modules/chatglm-6b/quantization.py", line 137, in __init__
    self.weight = compress_int4_weight(self.weight)
  File "/root/.cache/huggingface/modules/transformers_modules/chatglm-6b/quantization.py", line 78, in compress_int4_weight
    kernels.int4WeightCompression(
  File "/root/miniconda3/envs/glm/lib/python3.10/site-packages/cpm_kernels/kernels/base.py", line 48, in __call__
    func = self._prepare_func()
  File "/root/miniconda3/envs/glm/lib/python3.10/site-packages/cpm_kernels/kernels/base.py", line 36, in _prepare_func
    curr_device = cudart.cudaGetDevice()
  File "/root/miniconda3/envs/glm/lib/python3.10/site-packages/cpm_kernels/library/base.py", line 72, in wrapper
    raise RuntimeError("Library %s is not initialized" % self.__name)
RuntimeError: Library cudart is not initialized

报以上错

可能没有装cudnn(Ubuntu22.04)

#ubuntu 22.04

#wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-keyring_1.0-1_all.deb

#sudo dpkg -i cuda-keyring_1.0-1_all.deb 

#sudo apt-get update 

#sudo apt-get -y install cuda

环境配置

sudo gedit ~/.bashrc

 在文末添加以下信息

    export LD_LIBRARY_PATH=/usr/local/cuda/lib64:/usr/local/cuda/extras/CPUTI/lib64
    export CUDA_HOME=/usr/local/cuda/bin
    export PATH=$PATH:$LD_LIBRARY_PATH:$CUDA_HOME

验证cuda安装是否成功：关闭当前命令行，并执行

    source ~/.bashrc
    nvcc -V

最后显示安装的cuda版本为11.4，安装成功。

验证cuda是否正常

    cd ~/NVIDIA_CUDA-11.4_Samples/1_Utilities/bandwidthTest/
    make
    ./bandwidthTest