An implementation of inception_v1 on pytorch with pretrained weights.
This code is a pytorch translation of the soumith torch repo: https://github.com/soumith/inception.torch It implements the original version of the inception architechture; what is known has GoogLeNet.
Pretrained weights can be found at https://mega.nz/#!4RJE1SSY!kcCDyhkum6EQqVtqTc-deHnQuckM3zYSYq16bADbfww
Test accuracy of the pretrained model on imagenet is only 26.38%. If I am not mistaken, this is an issue of the original torch repo - the data loading is done correctly. If you train this model to better accuracy, I would love to get the new set of weights!
The code is licensed under the MIT Licence. See the LICENSE file for detail.