MCPcopy Create free account
hub / github.com/NVIDIA/TensorRT / get_model

Method get_model

demo/Diffusion/models.py:339–344  ·  view source on GitHub ↗
(self)

Source from the content-addressed store, hash-verified

337 self.name = "VAE decoder"
338
339 def get_model(self):
340 vae = AutoencoderKL.from_pretrained(self.path,
341 subfolder="vae",
342 use_auth_token=self.hf_token).to(self.device)
343 vae.forward = vae.decode
344 return vae
345
346 def get_input_names(self):
347 return ['latent']

Callers 3

loadEnginesMethod · 0.45
mainFunction · 0.45
load_and_setup_modelFunction · 0.45

Calls

no outgoing calls

Tested by

no test coverage detected