how to get clustering result of origin file name through data in file "/exp/clusters" ？

[SpartacusIn21]

No description provided.

[mathildecaron31]

Hi,
You may load the list of cluster assignments with pickle as it was saved with pickle.
You can do the following: cluster_assignments = pickle.load(open('/exp/clusters', 'rb')). This is a list, cluster_assignments[0] will give you the cluster assignment for the whole dataset at epoch 0.

[cptbtptp125]

Can the results of cluster assignment be visualized directly? That is, the classification result image is directly saved into the corresponding category folder.