Error when running python train.py task=Hover total_frames=100_000_000 eval_interval=60 headless=true

[godk1122]

hi！
I'm using isaac-sim:4.0.0, when running python train.py task=Hover total_frames=100_000_000 eval_interval=60 headless=true (enable evaluation (which would record a video clip for visulization) during training), there is an error:

Error executing job with overrides: ['task=Hover', 'total_frames=100_000_000', 'eval_interval=60', 'headless=true']
Traceback (most recent call last):
File "/omni_wjk/scripts/train.py", line 183, in main
info.update(evaluate())
File "/isaac-sim/exts/omni.isaac.ml_archive/pip_prebundle/torch/utils/_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
File "/omni_wjk/scripts/train.py", line 123, in evaluate
trajs = env.rollout(
File "/root/anaconda3/envs/sim/lib/python3.10/site-packages/torchrl/envs/common.py", line 2432, in rollout
tensordicts = self._rollout_nonstop(**kwargs)
File "/root/anaconda3/envs/sim/lib/python3.10/site-packages/torchrl/envs/common.py", line 2530, in _rollout_nonstop
callback(self, tensordict)
File "/omni_wjk/omni_drones/utils/torchrl/env.py", line 95, in call
frame = env.render(mode="rgb_array")
File "/omni_wjk/omni_drones/envs/isaac_env.py", line 377, in render
return rgb_data[:, :, :3]
IndexError: too many indices for array: array is 1-dimensional, but 3 were indexed

it seems like rgb_data is empty, how can i fix this problem? Thanks for your help!