haven-ai / haven-ai

To Run, Manage and Visualize Large Scale Experiments

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

haven_wizard for optimization

IssamLaradji opened this issue · comments

Change trainval.py to this (use this demo for reference: https://colab.research.google.com/drive/1iqmJWTfsC3Erfay_SwEoUhq_wS4l18Fd?usp=sharing#scrollTo=7bCDi3VlFsir)

# 1. define the training and validation function
def trainval(exp_dict, savedir, args):
    """
    exp_dict: dictionary defining the hyperparameters of the experiment
    savedir: the directory where the experiment will be saved
    args: arguments passed through the command line
    """
    ...

if __name__ == '__main__':
    hw.run_wizard(func=trainval, exp_groups =exp_configs.EXP_GROUPS)

def trainval(exp_dict, savedir_base, datadir, reset=False, num_workers=0, use_cuda=False):