jonbakerfish commented on issue #658: Checkpoint every some iterations URL: https://github.com/apache/incubator-mxnet/issues/658#issuecomment-329798467 FYI: a checkpoint class for `batch_end_callback`: class BatchCheckpoint(object): def __init__(self, mod, prefix, period, save_optimizer_states=False): """checkpoint for batch_end_callback Parameters ---------- mod : subclass of BaseModule The module to checkpoint. prefix : str The file prefix for this checkpoint. period : int How many batchs to wait before checkpointing. save_optimizer_states : bool Indicates whether or not to save optimizer states for continued training. Example ------- >>> # save checkpoint every ten batches. >>> module.fit(iterator, num_epoch=n_epoch, ... batch_end_callback=callback.BatchCheckpoint(module, prefix, 10)) """ self.mod = mod self.prefix = prefix self.period = period self.save_optimizer_states = save_optimizer_states def __call__(self, param): if (param.nbatch + 1) % self.period == 0: self.mod.save_checkpoint(self.prefix + '-e-%04d-b'%param.epoch, param.nbatch + 1, self.save_optimizer_states) ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org
With regards, Apache Git Services
