Another bug - In the load phase, optimizers were being pushed twice into the queue. Once as part of initModelFromConfig, and again in Model::load. Changed Model::load to work on the previously pushed optimizer objects instead.
if resume_training is set to true, the existing model_dir will be overwritten with the new model. If resume_from_checkpoint is set, then the new model is output to model_dir (either user-specified or newly created model_x where x belongs to [0,10] )
Testing resume_training with and without model_dir/resume_from_checkpoint specified.
Also a minor nit - It would be good to write more descriptive PR title and description. For example in this case, what is the bug, when is it triggered, how did you fix it etc.
Changed it to log in the following manner.
Another bug - In the load phase, optimizers were being pushed twice into the queue. Once as part of
initModelFromConfig
, and again inModel::load
. ChangedModel::load
to work on the previously pushed optimizer objects instead.if
resume_training
is set totrue
, the existingmodel_dir
will be overwritten with the new model. Ifresume_from_checkpoint
is set, then the new model is output tomodel_dir
(either user-specified or newly created model_x where x belongs to [0,10] )Testing
resume_training
with and withoutmodel_dir
/resume_from_checkpoint
specified.