Fastflow API error

[harshsinha03]

Hey I am new to this library and I am trying to implement Fastflow but this is the error I am getting

model = Fastflow( input_size=(256, 256), backbone="resnet18", pre_trained = True, flow_steps=8, conv3x3_only=False, hidden_ratio=1.0)

def configure_optimizers(lightning_module: LightningModule, optimizer: Optimizer) -> Optimizer: del lightning_module return optimizer optimizer = Adam(params=model.parameters(), lr=0.001, betas=(0.9, 0.999), weight_decay=1e-5) fn = partial(configure_optimizers, optimizer=optimizer) update_wrapper(fn, configure_optimizers) model.configure_optimizers = MethodType(fn, model)

trainer = Trainer( accelerator="gpu", auto_scale_batch_size=False, check_val_every_n_epoch=1, devices=1, max_epochs=5, num_sanity_val_steps=0, val_check_interval=1.0, ) trainer.fit(model=model, datamodule=datamodule)

AttributeError: 'Fastflow' object has no attribute 'threshold_method'

[blaz-r]

I'm unsure if this is the only issue, but use Engine, not directly trainer from lighting. Engine is anomalib specific wrapper of trainer that does some additional things, but also takes all the params from trainer if you need those.