Cannot load model trained by fastai1

Part 1 (2020)
1

Hi,

I have trained and built a GAN to colorise black and white photos by fastai1, it was working okay and I was able to build a web application for it.
But I got errors when I tried to load the models (pkl and pth) using fastai2. Is there a way to load models trained by fastai 1, or should I train it again using fastai2?

Thanks

2

You’d need to save the weights and the exact model, and then load those weights in, however this is learn.save(). fastai2 is not backwards compatible with fastai1

1 Like
3

Thanks!
I have one more question, I am trying to use FeatureLoss mentioned in lesson7 fastai v3 course.
But I cannot from fastai.core import * and from fastai.callbacks import hook_outputs using fastai2
I was trying to use the code from DeOldify. Thanks

from fastai import *
from fastai.core import *
from fastai.torch_core import *
from fastai.callbacks import hook_outputs
import torchvision.models as models
    

class FeatureLoss(nn.Module):
        def __init__(self, layer_wgts=[20, 70, 10]):
            super().__init__()

            self.m_feat = models.vgg16_bn(True).features.cuda().eval()
            requires_grad(self.m_feat, False)
            blocks = [
                i - 1
                for i, o in enumerate(children(self.m_feat))
                if isinstance(o, nn.MaxPool2d)
            ]
            layer_ids = blocks[2:5]
            self.loss_features = [self.m_feat[i] for i in layer_ids]
            self.hooks = hook_outputs(self.loss_features, detach=False)
            self.wgts = layer_wgts
            self.metric_names = ['pixel'] + [f'feat_{i}' for i in range(len(layer_ids))]
            self.base_loss = F.l1_loss

        def _make_features(self, x, clone=False):
            self.m_feat(x)
            return [(o.clone() if clone else o) for o in self.hooks.stored]

        def forward(self, input, target):
            out_feat = self._make_features(target, clone=True)
            in_feat = self._make_features(input)
            self.feat_losses = [self.base_loss(input, target)]
            self.feat_losses += [
                self.base_loss(f_in, f_out) * w
                for f_in, f_out, w in zip(in_feat, out_feat, self.wgts)
            ]

            self.metrics = dict(zip(self.metric_names, self.feat_losses))
            return sum(self.feat_losses)

        def __del__(self):
            self.hooks.remove()
4

use

from fastai.callbacks.hook import hook_outputs