search

hanjq17 / GMN

[ICLR 2022] The implementation for the paper "Equivariant Graph Mechanics Networks with Constraints".
https://arxiv.org/pdf/2203.06442.pdf
MIT License
56 stars 7 forks source link

visualization #4

Open newboy111 opened 11 months ago

newboy111 commented 11 months ago

Hello, may I ask how the data.pkl file is generated in visualization? I hope to get your help. Thank you!

hanjq17 commented 11 months ago

Hi,

Please refer to the comments in vis.ipynb, which tell that:

"data.pkl can be collected using the evaluation script, simply by manually saving data after testing. Important things for visualization: input pos (x_in), target pos (x_end), the prediction (x_gmn) and the cfg (kinematics decomposition)."

You can load your model, infer the results required above, and save them into data.pkl.

Let me know if you have further questions!