Closed SlowMonk closed 3 years ago
try to do the classification with efficnetnet-D0. and found there is two way to do this
model = timm.create_model('tf_efficientnet_b0_ns', pretrained=True, num_classes=num_classes) from efficientnet_pytorch import EfficientNet self.efficientnet = EfficientNet.from_pretrained(args.model_name)
the timm model's speed is much faster. what is the difference between thouse two?
Q for the forums
try to do the classification with efficnetnet-D0. and found there is two way to do this
the timm model's speed is much faster. what is the difference between thouse two?