Anthony W. Jung

dv3 Imagination notebook re-visited

4

I can't find a button to reopen the issue, regarding #272, I have two questions! 1) shouldn't initial `stochastic_state` be flattened s.t. ``` stochastic_state = player.stochastic_state.view(1, 1, -1).clone() ``` in...

New Dreamer V3 manuscript

2

Hi! Sharing slight change in Dreamer V3 according to their updated(2024/04/17) manuscript https://arxiv.org/pdf/2301.04104 Also their codes are updated few hours ago https://github.com/danijar/dreamerv3 It includes change in the optimizer (LaProp), experiments...

Often times, we want to save our models periodically. Model saved by frequency passed from config file is desirable.

We're trying to reproduce your work, I was wondering how you encoded register and memory. Can you tell me what each of dimension stands for? We assume B for batch...

Question for training dreamer_v3 on InvertedPendulum-v4

1

I ran below command to train InvertedPendulum-v4, ``` MUJOCO_GL=egl python sheeprl.py exp=dreamer_v3 env=mujoco env.id=InvertedPendulum-v4 fabric.accelerator=cuda fabric.devices=1 fabric.strategy=ddp algo.mlp_keys.encoder=[state] algo.mlp_keys.decoder=[state] algo.cnn_keys.encoder=[] algo.cnn_keys.decoder=[] ``` but it the agent is not learning (expected...