Hey, thanks for this repo. Your work on the LTH is really awesome!
I had a question about the resnet18 architecture provided. I noticed that the model was running a lot faster than other resnet18 models, so I printed the number of trainable parameters and noticed that for cifar_resnet_18 the number of parameters was 272474, whereas the resnet18 model from pytorch is around 11 million parameters. I was wondering if you could provide the naming schema that would return the typical 11 million parameter model?
Hey, thanks for this repo. Your work on the LTH is really awesome!
I had a question about the resnet18 architecture provided. I noticed that the model was running a lot faster than other resnet18 models, so I printed the number of trainable parameters and noticed that for
cifar_resnet_18
the number of parameters was 272474, whereas the resnet18 model from pytorch is around 11 million parameters. I was wondering if you could provide the naming schema that would return the typical 11 million parameter model?Thanks