Guodong Zhang

potential bugs in kfac.py

1

In compute_cov_a, `a = a.view(-1, a.size(-1)).div_(a.size(1)).div_(a.size(2))` should be `a = a.view(-1, a.size(-1))` In compute_cov_g, `g = g.view(-1, g.size(-1)).mul_(g.size(1)).mul_(g.size(2))` should be `g = g.view(-1, g.size(-1))` and `g_ = g * batch_size`...

If I understand right, run_trpo_expl.py is trpo + vime, so run_trpo is trpo w/o vime?

How to use weight initialization in CNN?

1

Could you give an short instruction? It's confusing to understand your code.

No checkpoint specified

2

When I run the command (to create dataset) `for alpha in {0.0,1.0}; do python3 scripts/create_dataset.py --save_dir=$HOME/tmp/ --load_dir=$HOME/tmp/CartPole-v0 --env_name=cartpole--num_trajectory=400 --max_trajectory_length=250 --alpha=$alpha --tabular_obs=0; done` It runs into error with the following information:...