Alan Descoins

Hey @jrief! Just tried with Django 1.9.12 and I get the exact same behavior.

It would be something like: ```python class MyInlineModel(models.Model): parent = models.ForeignKey(Parent, related_name='parent') order = models.PositiveIntegerField(default=0, editable=False, db_index=True) image = ImageField(upload_to='path', height_field='height', width_field='width') height = models.PositiveIntegerField(null=True, blank=True, editable=False, default=0) width =...

@jrief is there any way to fix this with the current models? It will take some time to do the necessary migrations to test this. Any reason why just using...

I don't use django-filer. I use the vanilla Django ImageField. This is why I find it weird. This case should probably work just fine, as it does in `django-admin-sortable`.

@EthanReid any updates on this? Could you get it to work?

Hello @jiangzihanict! Even though data might be quite different from ImageNet, I would still advise you to start from the pre-trained model and not from scratch. Transfer learning is widely...

Thanks for your response! `block2` means the training will proceed from this block of the ResNet onwards. [As you can see in the TensorFlow implementation](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/contrib/slim/python/slim/nets/resnet_v1.py#L285), the ResNet has 4 blocks....

Hi @cjpurackal! Currently, nobody is working on RetinaNet or Mask R-CNN implementation. I guess the simplest to start would be RetinaNet over Faster R-CNN, and take it from there. Currently...

Hi @SpeedyGonzalves! What is `GoogleModel` here? For this to work, either it must be a remote checkpoint hosted by Luminoth (it is not), or it must have been saved by...

The checkpoints should be available in the `~/.luminoth/checkpoints/` folder. Check that this is the case wherever you are running the web server.