Sorry for cross-posting...
I am trying to use the "Seabird (Tern) Detection - Africa" deep learning training package by ESRI Analytics set from ArcGIS' "Living Atlas of the World".
https://edgeosci.maps.arcgis.com/home/item.html?id=4019a53c914947aea9621ba226ec8861
However, there appear to be crucial files missing from the DLPK pacakge, namely the model configuration files. These files seem to be present in their other deep learning data sets, e.g. Arctic seals. See error below.
---------------------------------------------------------------------------
NameError Traceback (most recent call last)
~\AppData\Local\Temp\ipykernel_20868\1360836376.py in <cell line: 1>()
----> 1 model = ModelExtension.from_model(emd_path = model_path, data = data)
~\anaconda3\envs\arcgis_deep\lib\site-packages\arcgis\learn\models\_model_extension.py in from_model(cls, emd_path, data)
371 mextnsn = cls(
372 data,
--> 373 model_configuration,
374 backbone,
375 pretrained_path=str(model_file),
NameError: name 'model_configuration' is not defined
---------------------------------------------------------------------------
Does anyone know how I can aquire the missing configuration files?
You link is to your university
did you try
Introduction to the model—ArcGIS pretrained models | Documentation
or
Living Atlas of the World | ArcGIS
ref page Pretrained Deep Learning Models | Image Feature Extraction & More (esri.com)
Hi Dan,
Yes, I looked on those pages. The problem is that there are files missing in the provided package.
Hi Amy. All models do not contain the same set of files. Which version of pro are you on? I just tested it on my side and it seems to be working fine.
Hello, thanks for getting in touch. I'm on Pro 3.0.3.
I was hoping to get the Model Configuration file so I could run the training through Python. I just tried to follow the instructions to do it all through Pro instead, and came up with another error (1) where it's saying I haven't selected a dlpk or emd flile (I've tried both). I tried importing it directly from the Portal, and came up with error 2 instead... - could you advise?
Many thanks! Amy