search

ozansener / active_learning_coreset

Source code for ICLR 2018 Paper: Active Learning for Convolutional Neural Networks: A Core-Set Approach
MIT License
260 stars 42 forks source link

Running the package #5

Closed mfaraki closed 3 years ago

mfaraki commented 5 years ago

Hi, Just wondering if you could specify how to run the package and reproduce the results. From your paper I understand that the implementation is in Tensorflow. So probably for cifar10 that is tf_base -> cifar10_train -> train_cifar ?

There is implementation in PyTorch additional_baselines -> main as well. Is that just re-implementation in PyTorch?

alex-zyk commented 3 years ago

Same question. Could you guys provide more details about how to reproduce your results? Thank you!

linlinlin1993 commented 3 years ago

Hi, could you please provide how to reproduce the results? Such as the "settings.json" file, thanks a lot

ozansener commented 3 years ago

Please use the code from: https://github.com/google/active-learning/blob/master/sampling_methods/kcenter_greedy.py We are not supporting this anymore.