Ding Chen

Who have better GAIL hyper parameters?

In GAIL, I tested this hyper parameters and this seems to work a lot more consistently: a2c_hidden_size = 32 discrim_hidden_size = 128 lr = 1e-3 num_steps = 128 mini_batch_size =...

https://github.com/higgsfield/RL-Adventure-2/blob/master/3.ppo.ipynb

Paillier和LWE是分开实现的

可以参考Privacy-Preserving Deep Learning via Additively Homomorphic Encryption，https://www.cnblogs.com/lucifer1997/p/11243430.html，有这两种加密方法的实现

基本内容可以参见Privacy-Preserving Deep Learning via Additively Homomorphic Encryption，主要是复现这篇论文的工作，并加入了一些自己的想法

I see the similar problem in https://github.com/tristandeleu/pytorch-maml-rl/issues/41

Thanks a lot. After adding 'mp.set_start_method('spawn')', I fix this problem using this branch https://github.com/tristandeleu/pytorch-maml-rl/tree/fix/mp-spawn

Excuse me, have you finally solved this problem？