mi-prometheus
mi-prometheus copied to clipboard
IBM
Enabling reproducible Machine Learning research
In order to enable reproducible VIGIL experiments, and make CLEVR a lighter class: Move most of the complexity of CLEVR.generate_feature_maps_file() to GenerateFeatureMaps\
Grid Trainers/Testers on GPU have hardcoded sleep time (currently 3s). This is motivated by the fact that cuda-gpupick picks a free GPU only by checking the contexts running on a...
A slow-down in some experiments has been noticed after releasing 0.2.0 and above: - A run of MAC on CLEVR for 20 epochs takes more than 3 days whereas it...
The workers have several flags available, which are added based on inheritance for now (e.g. Worker :arrow_left: Trainer :arrow_left: OfflineTrainer etc). This results in some flags not particularly useful for...
**Describe the bug** Currently, all workers assume that they are executed from the mi-prometheus main directory. Along with setup.py we open the possibility to execute the mip-* workers from any...
The current implementation of curriculum learning forces by default that it needs to be finished even if convergence has been reached (in terms of loss < threshold): ` # If...