Skip to content
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.

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

Annotation mistake in examples #687

Open
hnl1 opened this issue Feb 6, 2025 · 5 comments
Open

Annotation mistake in examples #687

hnl1 opened this issue Feb 6, 2025 · 5 comments

Comments

@hnl1
Copy link

hnl1 commented Feb 6, 2025

1_load_lerobot_dataset.py L132

    print(f"{batch['observation.state'].shape=}")  # (32, 5, c)

I started learning LeRobot today. I noticed that the "5" in the comment here should be "6" because there are 6 timestamps in the observation.state of delta_timestamps earlier.

I'm not sure if this counts as an bug and whether it's worth creating a pull request (PR). However, as a beginner like me, it took me a while to confirm that it was a problem with the comments rather than an issue with my program's output.

@hnl1
Copy link
Author

hnl1 commented Feb 7, 2025

I found another problem at https://github.com/huggingface/lerobot/blob/638d411cd3acf32c28d8c2120f3c41bda8bb15d4/examples/2_evaluate_pretrained_policy.py#L7C1-L7C27

There is an extra backtick (`) at the end, which causes the command execution to fail.

@hnl1 hnl1 changed the title Annotation number mistake in 1_load_lerobot_dataset.py Annotation mistake in examples Feb 7, 2025
@tc-huang
Copy link
Contributor

tc-huang commented Feb 7, 2025

Hello @hnl1,

I noticed that the issue in examples/1_load_lerobot_dataset.py at line 132, which you mentioned, appears to have already been reported in another open pull request (#631).

@hnl1
Copy link
Author

hnl1 commented Feb 7, 2025

@tc-huang You're right. I just checked PR #631 , and the changes indeed solved the problem of 1_load_lerobot_dataset.py L132

@hnl1
Copy link
Author

hnl1 commented Feb 7, 2025

I found another problem at

# Convert to float32 with image from channel first in [0,255]

The original image already has the channels at the end. The operation here is to move the channels to the front. The annotation is incorrect.

@hnl1
Copy link
Author

hnl1 commented Feb 7, 2025

I will submit a PR to fix the other issues I've found, and then close this issue.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants