Method get_input

cogvideox-based/sat/diffusion_video.py:166–167 · view source on GitHub ↗

(self, batch)

Source from the content-addressed store, hash-verified

164	return loss, loss_dict
165
166	def get_input(self, batch):
167	return batch[self.input_key].to(self.dtype)
168
169	@torch.no_grad()
170	def decode_first_stage(self, z):

shared_stepMethod · 0.95

log_videoMethod · 0.95

toMethod · 0.80

no test coverage detected