Duckietown Challenges Home Challenges Submissions

Submission 1414

Submission1414
Competingyes
Challengeaido1_LF1_r3-v3
UserGiwoong Lee
Date submitted
Last status update
Completecomplete
DetailsEvaluation is complete.
Sisters
Result💔
Jobsstep1-simulation: 14398
Next
User labelPyTorch DDPG algorithm by Giwoong Lee
Admin priority50
Blessingn/a
User priority50

Evaluation jobs for this submission

See previous jobs for previous versions of challenges
Job IDstepstatusup to datedate starteddate completeddurationmessage
14398step1-simulationfailedyes0:00:42
InvalidSubmission: T [...]
InvalidSubmission:
Traceback (most recent call last):
  File "/workspace/src/duckietown-challenges/src/duckietown_challenges/cie_concrete.py", line 486, in wrap_evaluator
    raise InvalidSubmission(out[SPECIAL_INVALID_SUBMISSION])
InvalidSubmission: Invalid solution:
Traceback (most recent call last):
  File "/workspace/src/duckietown-challenges/src/duckietown_challenges/cie_concrete.py", line 590, in wrap_solution
    raise InvalidSubmission(msg)
InvalidSubmission: Uncaught exception in solution:
Traceback (most recent call last):
  File "/workspace/src/duckietown-challenges/src/duckietown_challenges/cie_concrete.py", line 585, in wrap_solution
    solution.run(cis)
  File "solution.py", line 150, in run
    solve(params, cis)
  File "solution.py", line 117, in solve
    action = Agent.get_action(obs_cat, is_noise=True)
  File "/workspace/model.py", line 239, in get_action
    action = self.actor.forward(state)
  File "/workspace/model.py", line 52, in forward
    x = self.conv1(states)
  File "/opt/conda/lib/python2.7/site-packages/torch/nn/modules/module.py", line 477, in __call__
    result = self.forward(*input, **kwargs)
  File "/opt/conda/lib/python2.7/site-packages/torch/nn/modules/conv.py", line 301, in forward
    self.padding, self.dilation, self.groups)
RuntimeError: Expected 4-dimensional input for 4-dimensional weight [32, 12, 8, 8], but got input of size [3, 480, 160] instead


Artefacts hidden. If you are the author, please login using the top-right link or use the dashboard.
No reset possible