Skip to content

[BUG] Diffusion Policy Bug #1083

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
rrzhang139 wants to merge 2 commits into
base: main
Choose a base branch
from
Open

[BUG] Diffusion Policy Bug #1083

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
9a29f3c
fix bug
rrzhang139 May 21, 2025
896b34b
fix padding
rrzhang139 May 22, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 7 additions & 4 deletions examples/baselines/diffusion_policy/train.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -147,14 +147,17 @@ def __getitem__(self, index):

obs_seq = self.trajectories['observations'][traj_idx][max(0, start):start+self.obs_horizon]
# start+self.obs_horizon is at least 1
act_seq = self.trajectories['actions'][traj_idx][max(0, start):end]
start_act = start + self.obs_horizon - 1 # start first action with the last obs
end_act = start_act + self.pred_horizon
act_seq = self.trajectories['actions'][traj_idx][max(0, start_act):min(end_act, L)]
if start < 0: # pad before the trajectory
obs_seq = torch.cat([obs_seq[0].repeat(-start, 1), obs_seq], dim=0)
act_seq = torch.cat([act_seq[0].repeat(-start, 1), act_seq], dim=0)
if end > L: # pad after the trajectory
if start_act < 0:
act_seq = torch.cat([act_seq[0].repeat(-start_act, 1), act_seq], dim=0)
if end_act > L: # pad after the trajectory
gripper_action = act_seq[-1, -1]
pad_action = torch.cat((self.pad_action_arm, gripper_action[None]), dim=0)
act_seq = torch.cat([act_seq, pad_action.repeat(end-L, 1)], dim=0)
act_seq = torch.cat([act_seq, pad_action.repeat(end_act-L, 1)], dim=0)
# making the robot (arm and gripper) stay still
assert obs_seq.shape[0] == self.obs_horizon and act_seq.shape[0] == self.pred_horizon
return {
Expand Down
15 changes: 8 additions & 7 deletions examples/baselines/diffusion_policy/train_rgbd.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -221,14 +221,15 @@ def __getitem__(self, index):
pad_obs_seq = torch.stack([obs_seq[k][0]] * abs(start), dim=0)
obs_seq[k] = torch.cat((pad_obs_seq, obs_seq[k]), dim=0)
# don't need to pad obs after the trajectory, see the above char drawing

act_seq = self.trajectories["actions"][traj_idx][max(0, start) : end]
if start < 0: # pad before the trajectory
act_seq = torch.cat([act_seq[0].repeat(-start, 1), act_seq], dim=0)
if end > L: # pad after the trajectory
gripper_action = act_seq[-1, -1] # assume gripper is with pos controller
start_act = start + self.obs_horizon - 1 # start first action with the last obs
end_act = start_act + self.pred_horizon
act_seq = self.trajectories['actions'][traj_idx][max(0, start_act):min(end_act, L)]
if start_act < 0:
act_seq = torch.cat([act_seq[0].repeat(-start_act, 1), act_seq], dim=0)
if end_act > L: # pad after the trajectory
gripper_action = act_seq[-1, -1]
pad_action = torch.cat((self.pad_action_arm, gripper_action[None]), dim=0)
act_seq = torch.cat([act_seq, pad_action.repeat(end - L, 1)], dim=0)
act_seq = torch.cat([act_seq, pad_action.repeat(end_act-L, 1)], dim=0)
# making the robot (arm and gripper) stay still
assert (
obs_seq["state"].shape[0] == self.obs_horizon
Expand Down