MCPcopy Create free account
hub / github.com/adobe-research/custom-diffusion / _predict_dataloader

Method _predict_dataloader

train.py:434–440  ·  view source on GitHub ↗
(self, shuffle=False)

Source from the content-addressed store, hash-verified

432 num_workers=self.num_workers, worker_init_fn=init_fn, shuffle=shuffle)
433
434 def _predict_dataloader(self, shuffle=False):
435 if isinstance(self.datasets['predict'], Txt2ImgIterableBaseDataset) or self.use_worker_init_fn:
436 init_fn = worker_init_fn
437 else:
438 init_fn = None
439 return DataLoader(self.datasets["predict"], batch_size=self.batch_size,
440 num_workers=self.num_workers, worker_init_fn=init_fn)
441
442
443class SetupCallback(Callback):

Callers

nothing calls this directly

Calls

no outgoing calls

Tested by

no test coverage detected