Alan Bi comments

Results 41 comments of


                                            Alan Bi

Losses come out to be None, as if the model isn't training on the data

You have to pass in a validation dataset/loader as the second argument to `model.fit` (or the keyword argument `val_dataset`) in order for it to return a list of losses. So...

Losses come out to be None, as if the model isn't training on the data

That likely means the model wasn't trained well enough in order to detect any objects. You could try printing or plotting the returned list of losses (if you have a...

Combining Two models

Currently this is not possible to do with Detecto - I'd recommend looking into PyTorch if you wanted to do something like this.

Forward all image objects at once instead of single object

Could you clarify the code or situation in which this happens? If you're using the `core.Dataset` class I think it should have been fixed in PR #62 and release [v1.2.0](https://github.com/alankbi/detecto/releases/tag/v1.2.0)

how to run it on raspberrypi and andriod and ios

Hi! There have been a few people who've also run into issues with getting Detecto/PyTorch to work on a Raspberry Pi, so unfortunately it doesn't look like Detecto is the...

Element missing from middle of dataset

I would guess most likely there's an issue with the XML file or image file for that specific index. Could you double-check that everything matches correctly for that index?

Jetson nano compatibility

It should be able to barring any unforeseen bugs - if you're coming across any issues feel free to let me know

Train Model has problem

Could you share the code and full output that results in this error?

Plot Training Loss

Suggestion noted. In the meantime, a workaround could be to copy and modify the source code for the `fit` method to return both the training and validation losses.

Plot Training Loss

@cagbal just want to confirm: have you been using the `verbose=True` flag when training?