RuntimeError: expected scalar type long int but found float

[Ethereal1679]

Traceback (most recent call last):
File "train.py", line 43, in
train(args)
File "train.py", line 38, in train
ppo_runner, train_cfg = task_registry.make_alg_runner(env=env, name=args.task, args=args)
File "/home/yyds/桌面/Gym5_human/humanoid-gym-main/humanoid/utils/task_registry.py", line 152, in make_alg_runner
runner = runner_class(env, all_cfg, log_dir, device=args.rl_device)
File "/home/yyds/桌面/Gym5_human/humanoid-gym-main/humanoid/algo/ppo/on_policy_runner.py", line 91, in init
_, _ = self.env.reset()
File "/home/yyds/桌面/Gym5_human/humanoid-gym-main/humanoid/envs/base/legged_robot.py", line 115, in reset
obs, privileged_obs, _, _, _ = self.step(torch.zeros(
File "/home/yyds/桌面/Gym5_human/humanoid-gym-main/humanoid/envs/custom/humanoid_env.py", line 197, in step
return super().step(actions)
File "/home/yyds/桌面/Gym5_human/humanoid-gym-main/humanoid/envs/base/legged_robot.py", line 102, in step
self.post_physics_step()
File "/home/yyds/桌面/Gym5_human/humanoid-gym-main/humanoid/envs/base/legged_robot.py", line 142, in post_physics_step
self.compute_reward()
File "/home/yyds/桌面/Gym5_human/humanoid-gym-main/humanoid/envs/base/legged_robot.py", line 226, in compute_reward
rew = self.reward_functionsi * self.reward_scales[name]
File "/home/yyds/桌面/Gym5_human/humanoid-gym-main/humanoid/envs/custom/humanoid_env.py", line 343, in _reward_feet_contact_number
reward = torch.where(contact == stance_mask, 1, -0.3)
RuntimeError: expected scalar type long int but found float
####################################################
It seems that torch.where is something wrong , can u help me to solve it?

[zlw21gxy]

Have you changed the definitions of contact or stance_mask? Does this happen with the default settings? Try this line: reward = torch.where(contact == stance_mask, 1.0, -0.3). Thank you for your feedback.