yulin

Results 6 comments of yulin

@theeluwin Oh, I just do not understand your initialization. I suppose that there must be some reasons which I do not know, so I want to ask about it. And...

I have changed CUDA version to 7.5 , cuDNN version to 5.0, and now the deepbench can run most of the benchmarks but except the 'nccl_mpi_all_reduce'.

> > @chxuan 最后解决了。原因应该是我在anaconda的虚拟环境下编译的,所以导致了-11这个错误。当退出虚拟环境用系统默认的Python解释器编译时,再次进入vim会出现类似[https://github.com/chxuan/vimplus](vimplus)Q&A中的第二个提示`(The ycmd server SHUT DOWN (restart with ':YcmRestartServer'). YCM core library compiled for Python 3 but loaded in Python 2. Set the 'g:ycm_server_python_interpreter' option to a Python...

感谢提供的精彩教程! 第七章的```NoisyDQN.ipynb```预测函数```predict_action```的第一句应该需要加上```self.policy_net.eval()```: ```python @torch.no_grad() def predict_action(self, state): self.policy_net.eval() ... ```

> 你好,请问在噪声网络中采样需要用到ε-贪心探索吗,我理解的是不用,但是我看你发的代码里用到ε-贪心探索了 使用作者提供的代码,实验了一下,去掉ε-贪心探索最终也可以将模型训练出来,但是需要稍多一点的训练step。由此看来,ε-贪心探索是一个比较通用且有效率的策略,可以在相对较短时间内探索出更多样的样本供网络学习。

我理解是为了将除法操作转换为减法操作吧