search

octo-models / octo

Octo is a transformer-based robot policy trained on a diverse mix of 800k robot trajectories.
https://octo-models.github.io/
MIT License
787 stars 152 forks source link

Initializer expected to generate shape (444, 256) but got shape (472, 256) instead for parameter "kernel" in "/diffusion_model/reverse_network/Dense_0". (https://flax.readthedocs.io/en/latest/api_reference/flax.errors.html#flax.errors.ScopeParamShapeError) python-BaseException #97

Open oym1994 opened 4 months ago

oym1994 commented 4 months ago

Hi, thanks for your great job.

When I test 01_inference_pretrained.ipynb without any change, there occurs a mistake: "Initializer expected to generate shape (444, 256) but got shape (472, 256) instead for parameter "kernel" in "/diffusion_model/reverse_network/Dense_0". (https://flax.readthedocs.io/en/latest/api_reference/flax.errors.html#flax.errors.ScopeParamShapeError) python-BaseException"

What is the reason and how to solve it? Thank you for your kind response.

oym1994 commented 4 months ago

It has no problem now. The reason is that my partner changed the pred_horizon in the octo-small checkpoint file config.json to 8. So it is ok.