Open taraxidas opened 3 years ago
Sorry for the delay replying.
I don't think there's much that prevents the C++ selfplay code from being compiled using a different toolchain, though that's not something we actively support. The C++ model code doesn't compile in any support for TensorFlow models by default (you have to define MG_ENABLE_TF_DUAL_NET
) so I strongly recommend trying to compile the simple_example first.
How are you planning to run the model? Do you need TensorFlow support? That will be a whole separate problem.
Hello i am working on my thesis and my job is to implement the concurrent selfplay code to an fpga via vivado hls. I wanted to ask you if there is a way that i can run concurrent selfplay via a compiler ide without using bazel or the terminal. Sorry if my question doesn't make sense and thanks for any help.