Description
home/ubtnavi/pby/RoboVLMs/calvin/calvin_env/calvin_env/robot/robot.py:234 in relative_to_absolute
│
│ 231 │ │ rel_pos *= self.max_rel_pos * self.magic_scaling_factor_pos
│ 232 │ │ rel_orn *= self.max_rel_orn * self.magic_scaling_factor_orn
│ 233 │ │ if self.use_target_pose:
│ ❱ 234 │ │ │ self.target_pos += rel_pos
│ 235 │ │ │ self.target_orn += rel_orn
│ 236 │ │ │ return self.target_pos, self.target_orn, gripper
│ 237 │ │ else:
│ ╭──────────── locals ───────────────────────────╮
│ │ action = tensor([ 1.1146e-02, -7.9814e-04, 3.0290e-03, 2.8761e-03, -2.4277e-03,
│ │ │ │ -1.7090e-02, 1.0000e+00])
│ │ gripper = tensor([1.])
│ │ rel_orn = tensor([ 0.0029, -0.0024, -0.0171])
│ │ rel_pos = tensor([ 0.0111, -0.0008, 0.0030])
│ │ self = <calvin_env.robot.robot.Robot object at 0x7f9700902df0>
│ ╰────────────────────────────────────────────────────╯
╰─────────────────────────────────────────────────────╯
TypeError: Concatenation operation is not implemented for NumPy arrays, use np.concatenate() instead. Please do not rely on
this error; it may not be given on all Python implementations.
When i evluate with "python eval/calvin/eval_ckpts.py", it throws this error.
Because the rel_pos and rel_orn are numpy array, and the action is a pytorch tenosr.
Therefore, i changed the line 323 in eval/calvin/evaluate_ddp-v2.py, from obs, _, _, current_info = env.step(action) to obs, _, _, current_info = env.step(action.numpy()). It worked, but i'm not sure this was the correct solution.