论文NCF代码实现

零.环境配置

pyhton = 3.6

1.安装theano

conda install mingw libpython
pip install theano

 2.安装tensorflow和keras

 版本关系

pip install tensorflow==2.1.0
pip install keras==2.3.1

 安装之后出现问题

Could not load dynamic library 'cudart64_101.dll'; dlerror: cudart64_101.dll not found

Ignore above cudart dlerror if you do not have a GPU set up on your machine.

 解决办法：

下载cudart64_101.dll并将其放在  C:\Windows\System32

下载地址

 下载成功后再测试

 3.CUDA 11.6对应的tensorflow-gpu版本问题

pip install tensorflow-gpu==2.10.0

Python 解决报错cudnn64_8.dll

查看自己使用的tensorflow是cpu还是gpu版本

[name: "/device:CPU:0"
device_type: "CPU"
memory_limit: 268435456
locality {
}
incarnation: 10253058432213849976
xla_global_id: -1
, name: "/device:GPU:0"
device_type: "GPU"
memory_limit: 1435228571
locality {
  bus_id: 1
  links {
  }
}
incarnation: 4545768756310698123
physical_device_desc: "device: 0, name: NVIDIA GeForce 940MX, pci bus id: 0000:01:00.0, compute capability: 5.0"
xla_global_id: 416903419
]