[build] require for parallel compiling!
Description
BitNet.cpp is nearly impossible to deploy on a low-ended arm device due to the low IO rate of the chipset. By introducing parallel compiling, this issue could be greatly improved!
Related Pull Request
https://github.com/microsoft/BitNet/pull/302
I also encountered this problem. I compiled on an M4 MAX device. For the LLAMA3 model, I selected -p and the compilation time exceeded one and a half hours (it was not completed when I came to the issue). I rewrote setup_env.py myself. Supporting ninja and multi-core did not speed up the compilation process.
I also encountered this problem. I compiled on an M4 MAX device. For the LLAMA3 model, I selected -p and the compilation time exceeded one and a half hours (it was not completed when I came to the issue). I rewrote setup_env.py myself. Supporting ninja and multi-core did not speed up the compilation process.
Sorry to hear that you experienced the same tragedy as me. My pull request has been published for 3 weeks and the maintainers did not reply. I presume they may get fired in Microsoft's global layoffs.
Let's just relax and pray for them.