lknownothing

关于复现结果的问题

2

感谢分享！ 我使用readme里默认的Digital_Music数据集，训练了DeepCoNN和NARRE。 1. 发现在DeepCoNN上验证结果为`evaluation reslut: mse: 0.8446; rmse: 0.9190;` 好于NRCA中出现的`mse=1.056`。在NARRE上为`evaluation reslut: mse: 0.8321; rmse: 0.9122; mae: 0.6754` ，差于NRCA文中的`mse=0.812`。 2. 此外，在较大的数据集上Video_Games上，NARRE的mse也只有`1.14`左右，达不到NRCA中的`1.11`。在Toys_and_Games数据集上NARRE和DeepCoNN都是差`0.844`左右。 3. 上述结果，我都是用不同随机数训练3-4次，还是存在和NRCA的差距。 想问下作者在具体复现baselines的时候，还需要注意哪些问题能达到和NRCA中差不多的结果呢。

Thank you for sharing the code. You are sooooo nice!!! When I reprodcue the VAE-CF base on TAFA. I found that the result is very poor. Could you share the...

What is the version of tensorflow?

1

Hi, Thank you for sharing. My question is as the title description.

Error: Could not initialize EGL. and RuntimeError: Failed to initialize OpenGL

4

Environment: + ubuntu 18.04 mujoco_py==2.0.2.5 ``` Found 3 GPUs for rendering. Using device 1. libEGL warning: Not allowed to force software rendering when API explicitly selects a hardware device. Could...