tensorflow C++做部署 训练模型在GPU上训练

Question

Invalid argument: Cannot assign a device for operation model_0/split: {{node model_0/split}}was explicitly assigned to /device:GPU:0 but available devices are [ /job:localhost/replica:0/task:0/device:CPU:0 ]. Make sure the device specification refers to a valid device. The requested device appears to be a GPU, but CUDA is not enabled.
[[{{node model_0/split}}]]

Answer 1

博主，这个问题你解决了吗？我也遇到了这个问题。我在机器1上编译并测试通过了TF-2.0-GPU-C++的确使用了GPU，比使用CPU版本的inference性能提升很多。但是我打包到机器2上时出现这个问题，机器2上的CUDA我配置好了，环境变量也设置好了nvidia-smi nvcc --version都OK，但就是TF不识别CUDA。https://www.cnblogs.com/mango-pudding/p/15670428.html