GET3D 代码笔记

训练的调用链为 main->launch_training->subprocess_fn->training/training_loop_3d.py

在launch_training函数中用torch.multiprocessing.spawn分布式调用subprocess_fn:

torch.multiprocessing.spawn(fn=subprocess_fn, args=(c, temp_dir), nprocs=c.num_gpus)

你可能感兴趣的:(AI与ML,深度学习,pytorch,神经网络)