Training Large Graph for edge regression

Questions
#1

Hello,
When I trainning a graph for edge regression value, I optain this error of memory allocation.
“”""
RuntimeError: [enforce fail at CPUAllocator.cpp:71] . DefaultCPUAllocator: can’t allocate memory: you tried to allocate 106968243600 bytes. Error code 12 (Cannot allocate memory)
“”""
I try to use mini-batch but not success.
Please can you help me ! Thanks in advance.

#2

Could you provide more details so that we can help? Like the dataset you’re using and a code snippet to reproduce the error.

#3

Thanks for response dyru!
My dataset contains two files one represent the nodes contains 120 nodes and other represents the edges contains 7 features and 271845 lines: Graph(num_nodes=120, num_edges=271845)
I solved the issue of memory allocation but I would like to create a model for the regression to predict value of one of the features (one amoung 7) related to the edge.
Please, have you any idea about the implementation of this thing.
Thanks in advance !

#4

A simplified example is provided here. You may follow and modify it to build your own edge regression model.

#5

Thanks a lot dyru.
It’s helpfull !

#6

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.