In tensorflow 1.14, it's obvious that tf.compat.v1.train.init_from_checkpoint
can load ckpt
to continue training (or to warm start). However, I couldn't find any corresponding approaches in SavedModel, and tf.estimator.WarmStartSetting only supports ckpt
as well. It's weird for me, because this answer mentioned that there should be a checkpoint stored in SavedModel
. Does anyone know:
- How to load checkpoint in SavedModel? or
- How to warm start training on SavedModel?