self.saved_searching_ternimal_state = self.root_state_tensor.clone()[self.lego_indices.view(-1), :].view(self.num_envs, 108, 13) RuntimeError: shape '[1, 108, 13]' is invalid for input of size 1716

sequential-dexterity / SeqDex

"Sequential Dexterity: Chaining Dexterous Policies for Long-Horizon Manipulation" code repository

https://sequential-dexterity.github.io/

Apache License 2.0

110 stars 10 forks source link

self.saved_searching_ternimal_state = self.root_state_tensor.clone()[self.lego_indices.view(-1), :].view(self.num_envs, 108, 13) RuntimeError: shape '[1, 108, 13]' is invalid for input of size 1716 #9

Open xiaobai0906 opened 5 months ago

xiaobai0906 commented 5 months ago

Hi, When I try to train each sub-task by 'python train_rlgames.py --task=BlockAssemblyOrient --num_envs=1024'. An error has occurred：
"self.saved_searching_ternimal_state = self.root_state_tensor.clone()[self.lego_indices.view(-1), :].view(self.num_envs, 108, 13) RuntimeError: shape '[1, 108, 13]' is invalid for input of size 1716." How can I solve it? Thanks!

cypypccpy commented 4 months ago

Hi @xiaobai0906 , This error is due to the environment dimension of the search terminal state you saved is different from the orient task. It seems that you are using one environment's searching_ternimal_state, and you are running 1024 environments' orient, so you also need to use the corresponding 1024-dimensional searching_ternimal_state. Hope this can help you.