I'm doing some proof of concept testing on the MacOS platform, writing to the API in python (2.7.13)
I found the following code (https://www.clarifai.com/developer/guide/predictions#predictions) that filters the prediction results based on the confidence value and it works great on a single image:
cdata = model.predict_by_filename(bname, min_value=0.96)
However, I am trying to predict a group of files in a batch process and wanted to add the same "min_value=0.96" filter, but in the "batch" mode, it doesn't accept the filter, I found the starting point for this here: (http://help.clarifai.com/api/batch-processing-with-python). I'm using the full structure of the code to identify all the files in a folder.
cdata = model.predict(imageList, min_value=0.96)
returns the error:
TypeError: predict() got an unexpected keyword argument 'min_value'
Without the min_value flag, the rest of the process works the way I want, it just returns too many concepts for my liking. Is there an easy way to do this?
Thanks in advance!