where is the resnet50-19c8e357.pth

[FJP123]

where is the resnet50-19c8e357.pth

[daodaofr]

hi, you can download the resnet50 model from "https://download.pytorch.org/models/resnet50-19c8e357.pth", and replace the line with your path:

hypergraph_reid/models/ResNet_hypergraphsage_part.py

Line 543 in 40fa990

self.base.load_param('/home/ubuntu/.torch/models/resnet50-19c8e357.pth')

[daodaofr]

Hi @FJP123

build_adj_full_d() is designed to return the adjacency matrix for the hypergraph. The input arguments are the temporal range, the number of part partitions. You can ignore d because we don't use it here.

Building the adjacency matrix is a basic step in GCN, please refer to https://github.com/Diego999/pyGAT/blob/b7b68866739fb5ae6b51ddb041e16f8cef07ba87/utils.py#L28