Blank Depth image screen for distillation policy video

[sdalal1]

I have ran into an issue where I am able to train the model with the --use_camera tag after adding the following two lines before import issacgym in the play.py scripts.

os.environ['MESA_VK_DEVICE_SELECT'] = '10de:2206'
os.environ["CUDA_VISIBLE_DEVICES"] = '0'

I am training on a headless machine which runs the issaacgym docker environment with the package added on it.

When I am running the trained model the depth image screen shows up but it is completely blank.

I get some warnings when I run the code which are as below

/opt/test_see/lib/python3.8/site-packages/torchvision/transforms/functional.py:1603: UserWarning: The default value of the antialias parameter of all the resizing transforms (Resize(), RandomResizedCrop(), etc.) will change from None to True in v0.17, in order to be consistent across the PIL and Tensor backends. To suppress this warning, directly pass antialias=True (recommended, future default), antialias=None (current default, which means False for Tensors and True for PIL), or antialias=False (only works on Tensors - PIL will still use antialiasing). This also applies if you are using the inference transforms from the models weights: update the call to weights.transforms(antialias=True).
warnings.warn(
'train_cfg' provided -> Ignoring 'name=a1'
Using MLP and Priviliged Env encoder ActorCritic structure
ActorCritic.init got unexpected arguments, which will be ignored: ['continue_from_last_std', 'priv_encoder_dims', 'rnn_hidden_size', 'rnn_num_layers', 'rnn_type', 'tanh_encoder_output']

I have had no issues training the base policy.

Thank you.

[sdalal1]

This is solved.