Question about how to invoke scene graph in caption_model

[xixiareone]

hi! In caption_model, I don't see how you invoke the code of the scene graph for each video you generated before, but instead use the data in sum data. Where did you invoke the scene graph you generated? thank you very much!

[maurya-rohit]

Hi, Actually we load the pickle file of each generated scene graph and created a list of sentences from the triplets. It is in the file

Scene-Graph-For-Videos/scene_graph_aggregator.py

Line 1 in 9d83307

#import requests

[xixiareone]

Can you share the scene graph data already extracted for the caption model? I want to directly use your scene graph data to train caption. Because I see the data in the sumdata you use directly，But the data in sumdata does not seem to be a charades scene graph.
Thank you so much！

[xixiareone]

I would like to ask how much graphics memory you use？thank you very much